An Adaptable Image Retrieval System With Relevance
Feedback Using Kernel Machines and Selective Sampling
Abstract: This paper presents an adaptable content-based image retrieval
(CBIR) system developed using regularization theory, kernel-based machines, and
Fisher information measure. The system consists of a retrieval subsystem that
carries out similarity matching using image-dependant information, multiple
mapping subsystems that adaptively modify the similarity measures, and a
relevance feedback mechanism that incorporates user information. The adaptation
process drives the retrieval error to zero in order to exactly meet either an existing
multiclass classification model or the user high-level concepts using referencemodel
or relevance feedback learning, respectively. To facilitate the selection of
the most informative query images during relevance feedback learning a new
method based upon the Fisher information is introduced. Model-reference and
relevance feedback learning mechanisms are thoroughly tested on a domainspecific
image database that encompasses a wide range of underwater objects
captured using an electro-optical sensor. Benchmarking results with two other
relevance feedback learning methods are also provided.
Morphological Background Detection
andEnhancement of Images With Poor
Lighting
Abstract: In this paper, some morphological transformations are used to detect
the background in images characterized by poor lighting. Lately, contrast image
enhancement has been carried out by the application of two operators based on the
Weber’s law notion. The first operator employs information from block analysis,
while the second transformation utilizes the opening by reconstruction, which is
employed to define the multibackground notion. The objective of contrast
operators consists in normalizing the grey level of the input image with the
purpose of avoiding abrupt changes in intensity among the different regions.
Finally, the performance of the proposed operators is illustrated through the
processing of images with different backgrounds, the majority of them with poor
lighting conditions.
Variational Bayesian Sparse Kernel-Based Blind Image
Deconvolution With Student’s-t Priors
Abstract: In this paper, we present a new Bayesian model for the blind image
deconvolution (BID) problem. The main novelty of this model is the use of a
sparse kernel-based model for the point spread function (PSF) that allows
estimation of both PSF shape and support. In the herein proposed approach, a
robust model of the BID errors and an image prior that preserves edges of the
reconstructed image are also used. Sparseness, robustness, and preservation of
edges are achieved by using priors that are based on the Student’s-t probability
density function (PDF). This PDF, in addition to having heavy tails, is closely
related to the Gaussian and, thus, yields tractable inference algorithms. The
approximate variational inference methodology is used to solve the corresponding
Bayesian model. Numerical experiments are presented that compare this BID
methodology to previous ones using both simulated and real data.
Independent Component Analysis-Based Background
Subtraction for Indoor Surveillance
Abstract: In video surveillance, detection of moving objects from an image
sequence is very important for target tracking, activity recognition, and behavior
understanding. Background subtraction is a very popular approach for foreground
segmentation in a still scene image. In order to compensate for illumination
changes, a background model updating process is generally adopted, and leads to
extra computation time. In this paper, we propose a fast background subtraction
scheme using independent component analysis (ICA) and, particularly, aims at
indoor surveillance for possible applications in home-care and health-care
monitoring, where moving and motionless persons must be reliably detected. The
proposed method is as computationally fast as the simple image difference
method, and yet is highly tolerable to changes in room lighting. The proposed
background subtraction scheme involves two stages, one for training and the other
for detection. In the training stage, an ICA model that directly measures the
statistical independency based on the estimations of joint and marginal probability
density functions from relative frequency distributions is first proposed. The
proposed ICA model can well separate two highly-correlated images. In the
detection stage, the trained de-mixing vector is used to separate the foreground in
a scene image with respect to the reference background image. Two sets of indoor
examples that involve switching on/off room lights and opening/closing a door are
demonstrated in the experiments. The performance of the proposed ICA model for
background subtraction is also compared with that of the well-known FastICA
algorithm.
Self-Similarity Driven Color Demosaicking
Abstract: Demosaicking is the process by which from a matrix of colored pixels
measuring only one color component per pixel, red, green, or blue, one can infer
whole color information at each pixel. This inference requires a deep
understanding of the interaction between colors, and the involvement of image
local geometry. Although quite successful in making such inferences with very
small relative error, state-of-the-art demosaicking methods fail when the local
geometry cannot be inferred from the neighboring pixels. In such a case, which
occurs when thin structures or fine periodic patterns were present in the original,
state-of-the-art methods can create disturbing artifacts, known as zipper effect,
blur, and color spots. The aim of this paper is to show that these artifacts can be
avoided by involving the image self-similarity to infer missing colors. Detailed
experiments show that a satisfactory solution can be found, even for the most
critical cases. Extensive comparisons with state-of-the-art algorithms will be
performed on two different classic image databases.
Mat lab
A Shear let Approach to Edge Analysis and Detection
Abstract: It is well known that the wavelet transform provides a very effective
framework for analysis of multi scale edges. In this paper, we propose a novel
approach based on the shear let transform: a multi scale directional transform with
a greater ability to localize distributed discontinuities such as edges. Indeed,
unlike traditional wavelets, shear lets are theoretically optimal in representing
images with edges and, in particular, have the ability to fully capture directional
and other geometrical features. Numerical examples demonstrate that the shear let
approach is highly effective at detecting both the location and orientation of
edges, and outperforms methods based on wavelets as well as other standard
methods. Furthermore, the shear let approach is useful to design simple and
effective algorithms for the detection of corners and junctions.
Color Texture Segmentation Based on the Modal Energy of
Deformable Surfaces
Abstract: This paper presents a new approach for the segmentation of color
textured images, which is based on a novel energy function. The proposed energy
function, which expresses the local smoothness of an image area, is derived by
exploiting an intermediate step of modal analysis that is utilized in order to
describe and analyze the deformations of a 3-D deformable surface model. The
external forces that attract the 3-D deformable surface model combine the
intensity of the image pixels with the spatial information of local image regions.
The proposed image segmentation algorithm has two steps. First, a color
quantization scheme, which is based on the node displacements of the deformable
surface model, is utilized in order to decrease the number of colors in the image.
Then, the proposed energy function is used as a criterion for a region growing
algorithm. The final segmentation of the image is derived by a region merge
approach. The proposed method was applied to the Berkeley segmentation
database. The obtained results show good segmentation robustness, when
compared to other state of the art image segmentation algorithms.
Composition of a Dewarped and Enhanced Document Image
From Two View Images
Abstract: In this paper, we propose an algorithm to compose a geometrically
dewarped and visually enhanced image from two document images taken by a
digital camera at different angles. Unlike the conventional works that require
special equipments or assumptions on the contents of books or complicated image
acquisition steps, we estimate the unfolded book or document surface from the
corresponding points between two images. For this purpose, the surface and
camera matrices are estimated using structure reconstruction, 3-D projection
analysis, and random sample consensus based curve fitting with the cylindrical
surface model. Because we do not need any assumption on the contents of books,
the proposed method can be applied not only to optical character recognition
(OCR), but also to the high-quality digitization of pictures in documents. In
addition to the Dewar ping for a structurally better image, image mosaic is also
performed for further improving the visual quality. By finding better parts of
images (with less out of focus blur and/or without secular reflections) from either
of views, we compose a better image by stitching and blending them. These
processes are formulated as energy minimization problems that can be solved
using a graph cut method. Experiments on many kinds of book or document
images show that the proposed algorithm robustly works and yields visually
pleasing results. Also, the OCR rate of the resulting image is comparable to that of
document images from a flatbed scanner.
Estimation of Motions in Color Image Sequences Using
Hyper complex Fourier Transforms
Abstract: Although the motion estimation problem has been extensively studied,
most of the proposed estimation approaches deal mainly with monochrome
videos. The most usual way to apply them also in color image sequences is to
process each color channel separately. A different, more sophisticated approach is
to process the color channels in a “holistic” manner using quaternions, as
proposed by Ell and Sangwine. In this paper, we extend standard spatiotemporal
Fourier-based approaches to handle color image sequences, using the hyper
complex Fourier transform. We show that translational motions are manifested as
energy concentration along planes in the hyper complex 3-D Fourier domain and
we describe a methodology to estimate the motions, based on this property.
Furthermore, we compare the three-channels-separately approach with our
approach and we show that the computational effort can be reduced by a factor of
1/3, using the hyper complex Fourier transforms. Also, we propose a simple,
accompanying method to extract the moving objects in the hyper complex Fourier
domain. Our experimental results on synthetic and natural images verify our
arguments throughout the paper.
Facial Recognition Using Multi sensor Images Based on
Localized Kernel Eigen Spaces
Abstract: A feature selection technique along with an information fusion
procedure for improving the recognition accuracy of a visual and thermal imagebased
facial recognition system is presented in this paper. A novel modular kernel
eigen spaces approach is developed and implemented on the phase congruency
feature maps extracted from the visual and thermal images individually. Smaller
sub-regions from a predefined neighborhood.
Within the phase congruency images of the training samples are merged to obtain
a large set of features. These features are then projected into higher dimensional
spaces using kernel methods. The proposed localized nonlinear feature selection
procedure helps to overcome the bottlenecks of illumination variations, partial
occlusions, expression variations and variations due to temperature changes that
affect the visual and thermal face recognition techniques. AR and Equinox
databases are used for experimentation and evaluation of the proposed technique.
The proposed feature selection procedure has greatly improved the recognition
accuracy for both the visual and thermal images when compared to conventional
techniques. Also, a decision level fusion methodology is presented which along
with the feature selection procedure has outperformed various other face
recognition techniques in terms of recognition accuracy
Study on Huber Fractal Image Compression
Abstract: In this paper, a new similarity measure for fractal image compression
(FIC) is introduced. In the proposed Huber fractal image compression (HFIC), the
linear Huber regression technique from robust statistics is embedded into the
encoding procedure of the fractal image compression. When the original image is
corrupted by noises, we argue that the fractal image compression scheme should
be insensitive to those noises presented in the corrupted image. This leads to a
new concept of robust fractal image compression. The proposed HFIC is one of
our attempts toward the design of robust fractal image compression. The main
disadvantage of HFIC is the high computational cost. To overcome this drawback,
particle swarm optimization (PSO) technique is utilized to reduce the searching
time. Simulation results show that the proposed HFIC is robust against outliers in
the image. Also, the PSO method can effectively reduce the encoding time while
retaining the quality of the retrieved image.
Fuzzy Peer Groups for Reducing Mixed Gaussian-Impulse
Noise From Color Images
Abstract: The peer group of an image pixel is a pixel similarity based concept
which has been successfully used to devise image denoising methods. However,
since it is difficult to define the pixel similarity in a crisp way, we propose to
represent this similarity in fuzzy terms. In this paper, we introduce the fuzzy peer
group concept, which extends the peer group concept in the fuzzy setting. A fuzzy
peer group will be defined as a fuzzy set that takes a peer group as support set and
where the membership degree of each peer group member will be given by its
fuzzy similarity with respect to the pixel under processing. The fuzzy peer group
of each image pixel will be determined by means of a novel fuzzy logic-based
procedure. We use the fuzzy peer group concept to design a two-step color image
filter cascading a fuzzy rule-based switching impulse noise filter by a fuzzy
average filtering over the fuzzy peer group. Both steps use the same fuzzy peer
group, which leads to computational savings. The proposed filter is able to
efficiently suppress Gaussian noise and impulse noise, as well as mixed Gaussianimpulse
noise. Experimental results are provided to show that the proposed filter
achieves a promising performance.
High-Fidelity Data Embedding for Image Annotation
Abstract: High fidelity is a demanding requirement for data hiding, especially
for images with artistic or medical value. This correspondence proposes a highfidelity
image watermarking for annotation with robustness to moderate
distortion. To achieve the high fidelity of the embedded image, we introduce a
visual perception model that aims at quantifying the local tolerance to noise for
arbitrary imagery. Based on this model, we embed two kinds of watermarks: a
pilot watermark that indicates the existence of the watermark and an information
watermark that conveys a payload of several dozen bits. The objective is to embed
32 bits of metadata into a single image in such a way that it is robust to JPEG
compression and cropping. We demonstrate the effectiveness of the visual model
and the application of the proposed annotation technology using a database of
challenging photographic and medical images that contain a large amount of
smooth regions.
Mat lab
A New Look to Multi channel Blind Image Deconvolution
Abstract: The aim of this paper is to propose a new look to MBID, examine
some known approaches, and provide a new MC method for restoring blurred and
noisy images. First, the direct image restoration problem is briefly revisited. Then
a new method based on inverse filtering for perfect image restoration in the
noiseless case is proposed. The noisy case is addressed by introducing a
regularization term into the objective function in order to avoid noise
amplification. Second, the filter identification problem is considered in the MC
context. A new robust solution to estimate the degradation matrix filter is then
derived and used in conjunction with a total variation approach to restore the
original image. Simulation results and performance evaluations using recent
image quality metrics are provided to assess the effectiveness of the proposed
methods.
Image Segmentation Using Information BottleneckMethod
Abstract: In image processing, segmentation algorithms constitute one of the
main focuses of research. In this paper, new image segmentation algorithms based
on a hard version of the information bottleneck method are presented. The
objective of this method is to extract a compact representation of a variable,
considered the input, with minimal loss of mutual information with respect to
another variable, considered the output. First, we introduce a split-and-merge
algorithm based on the definition of an information channel between a set of
regions (input) of the image and the intensity histogram bins (output). From this
channel, the maximization of the mutual information gain is used to optimize the
image partitioning. Then, the merging process of the regions obtained in the
previous phase is carried out by minimizing the loss of mutual information. From
the inversion of the above channel, we also present a new histogram clustering
algorithm based on the minimization of the mutual information loss, where now
the input variable represents the histogram bins and the output is given by the set
of regions obtained from the above split-and-merge algorithm.
Finally, we introduce two new clustering algorithms which show how the
information bottleneck method can be applied to the registration channel obtained
when two multimodal images are correctly aligned. Different experiments on 2-D
and 3-D images show the behavior of the proposed algorithms.
Reconstructing Orientation Field From Fingerprint
Minutiae to Improve Minutiae-Matching Accuracy
Abstract: Minutiae are very important features for fingerprint representation,
and most practical fingerprint recognition systems only store the minutiae
template in the database for further usage. The conventional methods to utilize
minutiae information are treating it as a point set and finding the matched points
from different minutiae sets. In this paper, we propose a novel algorithm to use
minutiae for fingerprint recognition, in which the fingerprint’s orientation field is
reconstructed from minutiae and further utilized in the matching stage to enhance
the system’s performance. First, we produce “virtual” minutiae by using
interpolation in the sparse area, and then use an orientation model to reconstruct
the orientation field from all “real” and “virtual” minutiae. A decision fusion
scheme is used to combine the reconstructed orientation field matching with
conventional minutiae-based matching. Since orientation field is an important
global feature of fingerprints, the proposed method can obtain better results than
conventional methods. Experimental results illustrate its effectiveness.
A Fast Optimization Transfer Algorithm for Image In
painting in Wavelet Domains
Abstract: A wavelet in painting problem refers to the problem of filling in
missing wavelet coefficients in an image. A variational approach was used by
Chan et al. The resulting functional was minimized by the gradient descent
method. In this paper, we use an optimization transfer technique which involves
replacing their univariate functional by a bivariate functional by adding an
auxiliary variable. Our bivariate functional can be minimized easily by alternating
minimization: for the auxiliary variable, the minimum has a closed form solution,
and for the original variable, the Minimization problem can be formulated as a
classical total variation (TV) denoising problem and, hence, can be solved
efficiently using a dual formulation. We show that our bivariate functional is
equivalent to the original univariate functional. We also show that our alternating
minimization is convergent. Numerical results show that the proposed algorithm
is very efficient and outperforms that of Chan et al.
Mat lab
Interpolation Artifacts in Sub-Pixel Image Registration
Abstract: We consider the problem of registering (aligning) two images to subpixel
accuracy by optimization of objective functions constructed from the
images’ intensity values. We show that some widely used interpolation methods
can introduce multiple local optima in the energy of the interpolated image which,
if not counterbalanced by other terms, can cause local optima in registration
objective functions including the sum of squared differences, cross correlation,
and mutual information. We discuss different solutions to address the problem
based on high degree B-spline interpolation, low pass filtering the images, and
stochastic integration. Numerical examples using synthetic and real signals and
images are shown.
High-Dimensional Statistical Measure for Region-of-Interest
Tracking
Abstract: This paper deals with region-of-interest (ROI) tracking in video
sequences. The goal is to determine in successive frames the region which best
matches, in terms of a similarity measure, a ROI defined in a reference frame.
Some tracking methods define similarity measures which efficiently combine
several visual features into a probability density function (PDF) representation,
thus building a discriminative model of the ROI. This approach implies dealing
with PDFS with domains of definition of high dimension. To overcome this
obstacle, a standard solution is to assume independence between the different
features in order to bring out low-dimension marginal laws and/or to make some
parametric assumptions on the PDFS at the cost of generality. We discard these
assumptions by proposing to compute the Kullback–Leibler divergence between
high-dimensional PDFS using the nearest neighbor framework. In consequence,
the divergence is expressed directly from the samples, i.e., without explicit
estimation of the underlying PDFS. As an application, we defined 5, 7, and 13-
dimensional feature vectors containing color information (including pixel-based,
gradient-based and patch-based) and spatial layout. The proposed procedure
performs tracking allowing for translation and scaling of the ROI. Experiments
show its efficiency on a movie excerpt and standard test sequences selected for
the specific conditions they exhibit: partial occlusions, variations of luminance,
noise, and complex motion
Improved Resolution Scalability for Bilevel Image Data in
JPEG2000
Abstract: In this paper, we address issues concerning bilevel image compression
using JPEG2000. While JPEG2000 is designed to compress both bilevel and
continuous tone image data using a single unified framework, there exist
significant limitations with respect to its use in the lossless compression of bilevel
imagery. In particular, substantial degradation in image quality at low resolutions
severely limits the resolution scalable features of the JPEG2000 code-stream. We
examine these effects and present two efficient methods to improve resolution
scalability for bilevel imagery in JPEG2000. By analyzing the sequence of
rounding operations performed in the JPEG2000 lossless compression pathway,
we introduce a simple pixel assignment scheme that improves image quality for
commonly occurring types of bilevel imagery. Additionally, we develop a more
general strategy based on the JPIP protocol, which enables efficient interactive
access of compressed bilevel imagery. It may be noted that both proposed
methods are fully compliant with Part 1 of the JPEG2000 standard.
Joint Optimization of Run-Length Coding, Huffman
Coding, and Quantization Table With Complete Baseline
JPEG Decoder Compatibility
Abstract: To maximize rate distortion performance while remaining faithful to
the JPEG syntax, the joint optimization of the Huffman tables, quantization step
sizes, and DCT indices of a JPEG encoder is investigated. Given Huffman tables
and quantization step sizes, an efficient graph-based algorithm is first proposed to
find the optimal DCT indices in the form of run-size pairs. Based on this graphbased
algorithm, an iterative algorithm is then presented to jointly optimize runlength
coding, Huffman coding, and quantization table selection. The proposed
iterative algorithm not only results in a compressed bit stream completely
compatible with existing JPEG and MPEG decoders, but is also computationally
efficient. Furthermore, when tested over standard test images, it achieves the best
JPEG compression results, to the extent that its own JPEG compression
performance even exceeds the quoted PSNR results of some state-of-the-art
Wavelet-based image coders such as Shapiro’s embedded zero tree wavelet
algorithms at the common bit rates under comparison. Both the graph-based
algorithm and the iterative algorithm can be applied to application areas such as
web image acceleration, digital camera image compression, MPEG frame
optimization, and Transco ding, etc.
Semi-Automatically Labeling Objects in Images
A bstract:Labeling objects in images plays a crucial role in many visual learning
and recognition applications that need training data, such as image retrieval,
object detection and recognition. Manually creating object labels in images is time
consuming and, thus, becomes impossible for labeling a large image dataset. In
this paper, we present a family of semi-automatic methods based on a graph-based
semi-supervised learning algorithm for labeling objects in images. We first
present Smart- Label that proposes to label images with reduced human input by
iteratively computing the harmonic solutions to minimize a quadratic energy
function on the Gaussian fields. Smart Label tackles the problem of lacking
negative data in the learning by embedding relevance feedback after the first
iteration, which also leads to one limitation of Smart Label—needing additional
human supervision. To overcome the limitation and enhance Smart Label, we
propose SmartLabel-2 that utilizes a novel scheme to sample negative examples
automatically, replace regular patch partitioning in Smart Label by quad tree
partitioning and applies image over-segmentation (superpixels) to extract smooth
object contours. Evaluations on six diverse object categories have indicated that
SmartLabel-2 can achieve promising results with a small amount of labeled data
(e.g., 1%–5% of image size) and obtain close-to-fine extraction of object contours
on different kinds of objects.
Phase Information and Space Filling Curves in Noisy Motion
Estimation
Abstract: This correspondence presents a novel approach for translational motion
estimation based on the phase of the Fourier transform. It exploits the equality
between the averaging of a group of successive frames and the convolution of the
reference one with an impulse train function. The use of suitable space filling
curves allows reducing the error in motion estimation making the proposed
approach robust under noise. Experimental results show that the proposed
approach outperforms available techniques in terms of objective (PSNR) and
subjective quality with a lower computational effort. Mat lab
A Dynamic Hierarchical Clustering Method for Trajectory-
Based Unusual Video Event Detection
Abstract: The proposed unusual video event detection method is based on
unsupervised clustering of object trajectories, which are modeled by hidden
Markov models (HMM). The novelty of the method includes a dynamic
hierarchical process incorporated in the trajectory clustering algorithm to prevent
model over fitting and a 2-depth greedy search strategy for efficient clustering.
Rate-Invariant Recognition of Humans and Their Activities
Abstract: Pattern recognition in video is a challenging task because of the
multitude of spatio-temporal variations that occur in different videos capturing the
exact same event. While traditional pattern-theoretic approaches account for the
spatial changes that occur due to lighting and pose, very little has been done to
address the effect of temporal rate changes in the executions of an event. In this
paper, we provide a systematic model-based approach to learn the nature of such
temporal variations (time warps) while simultaneously allowing for the spatial
variations in the descriptors. We illustrate our approach for the problem of action
recognition and provide experimental justification for the importance of
accounting for rate variations in action recognition. The model is composed of a
nominal activity trajectory and a function space capturing the probability
distribution of activity-specific time warping transformations. We use the squareroot
parameterization of time warps to derive geodesics, distance measures, and
probability distributions on the space of time warping functions. We then design a
Bayesian algorithm which treats the execution rate function as a nuisance variable
and integrates it out using Monte Carlo sampling, to generate estimates of class
posteriors. This approach allows us to learn the space of time warps for each
activity while simultaneously capturing other intra- and interclass variations.
Next, we discuss a special case of this approach which assumes a uniform
distribution on the space of time warping functions and show how
computationally efficient inference algorithms may be derived for this special
case. We discuss the relative advantages and disadvantages of both approaches
and show their efficacy using experiments on gait-based person identification and
activity recognition
A Unified Relevance Feedback Framework for Web Image
Retrieval
Abstract: Although relevance feedback (RF) has been extensively studied in the
content-based image retrieval community, no commercial Web image search
engines support RF because of scalability, efficiency, and effectiveness issues. In
this paper, we propose a unified relevance feedback framework for Web image
retrieval. Our framework shows advantage over traditional RF mechanisms in the
following three aspects. First, during the RF process, both textual feature and
visual feature are used in a sequential way. To seamlessly combine textual
feature-based RF and visual feature-based RF, a query concept-dependent fusion
strategy is automatically learned. Second, the textual feature-based RF mechanism
employs an effective search result clustering (SRC) algorithm to obtain salient
phrases, based on which we could construct an accurate and low-dimensional
textual space for the resulting Web images. Thus, we could integrate RF into Web
image retrieval in a practical way. Last, a new user interface (UI) is proposed to
support implicit RF. On the one hand, unlike traditional RF UI which enforces
users to make explicit judgment on the results, the new UI regards the users’
click-through data as implicit relevance feedback in order to release burden from
the users. On the other hand, unlike traditional RF UI which hardily substitutes
subsequent results for previous ones, a recommendation scheme is used to help
the users better understand the feedback process and to mitigate the possible
waiting caused by RF. Experimental results on a database consisting of nearly
three million Web images show that the proposed framework is wieldy, scalable,
and effective.
Sparse Image Reconstruction for Molecular Imaging
Abstract: The application that motivates this paper is molecular imaging at the
atomic level. When discretized at subatomic distances, the volume is inherently
sparse. Noiseless measurements from an imaging technology can be modeled by
convolution of the image with the system point spread function (PSF). Such is the
case with magnetic resonance force microscopy (MRFM), an emerging
technology where imaging of an individual tobacco mosaic virus was recently
demonstrated with nanometer resolution. We also consider additive white
Gaussian noise (AWGN) in the measurements.
Many prior works of sparse estimators have focused on the case when H has low
coherence; however, the system matrix H in our application is the convolution
matrix for the system PSF. A typical convolution matrix has high coherence. This
paper, therefore, does not assume a low coherence. A discrete-continuous form of
the Laplacian and atom at zero (LAZE) PDF used by Johnstone and Silverman is
formulated, and two sparse estimators derived by maximizing the joint PDF of the
observation and image conditioned on the hyper parameters. A threshold rule that
generalizes the hard and soft thresholding rule appears in the course of the
derivation. This so-called hybrid thresholding rule, when used in the iterative
thresholding framework, gives rise to the hybrid estimator, a generalization of the
lasso. Estimates of the hyper parameters for the lasso and hybrid estimator are
obtained via Stein’s unbiased risk estimate (SURE). A numerical study with a
Gaussian PSF and two sparse images shows that the hybrid estimator outperforms
the lasso.
Optimal Spread Spectrum Watermark Embedding via a
Multi step Feasibility Formulation
Abstract: We consider optimal formulations of spread spectrum watermark
embedding where the common requirements of watermarking, such as perceptual
closeness of the watermarked image to the cover and detect ability of the
watermark in the presence of noise and compression, are posed as constraints
while one metric pertaining to these requirements is optimized. We propose an
algorithmic framework for solving these optimal embedding problems via a multi
step feasibility approach that combines projections onto convex sets (POCS)
based feasibility watermarking with a bisection parameter search for determining
the optimum value of the objective function and the optimum watermarked image.
The framework is general and can handle optimal watermark embedding problems
with convex and quasi-convex formulations of watermark requirements with
assured convergence to the global optimum. The proposed scheme is a natural
extension of set-theoretic watermark design and provides a link between convex
feasibility and optimization formulations for watermark embedding. We
demonstrate a number of optimal watermark embeddings in the proposed
framework corresponding to maximal robustness to additive noise, maximal
robustness to compression, minimal frequency weighted perceptual distortion, and
minimal watermark texture visibility. Experimental results demonstrate that the
framework is effective in optimizing the desired characteristic while meeting the
constraints. The results also highlight both anticipated and unanticipated
competition between the common requirements for watermark embedding.
Curve let-Based Feature Extraction with B-LDA for Face
Recognition
Abstract: In this paper, we propose a novel feature extraction scheme based on
the multi-resolution curve let transform for face recognition. The obtained curve
let coefficients act as the feature set for classification, and is used to train the
ensemble-based discriminate learning approach, capable of taking advantage of
both the boosting and LDA (BLDA) techniques. The proposed method CV-BLDA
has been extensively assessed using different databases: the ATT, YALE and
FERET, Tests indicate that using curve let-based features significantly improves
the accuracy compared to standard face recognition algorithms and other multiresolution
based approaches.
Mat lab
Watermarking of Co lour Images in the DCT Domain Using
Y Channel
Abstract: In this paper an image authentication technique that embeds a binary
watermark into a host color image is proposed. In this scheme the co lour image is
first transformed from the RGB to the YCBCR color space. The watermark is
embedded into the Y channel of the host image by selectively modifying the very
low frequency parts of the DCT transformation. It is shown that the proposed
technique can resist classical attacks such as JPEG Compression, low pass
filtering, median filtering, cropping, and geometrical scaling attack. Moreover, the
recovery method is blind and doesn’t need the original host image for extraction.
Face Recognition Based on Facial Feature Training
Abstract: The detection and extraction of characteristic features of the human
face are primordial tasks in any given approach for face recognition. Within this
context, the present paper will present a comprehensive list of steps that can help
obtain the major discriminative features of the human face with the abstraction of
the fuzzy data that are influenced by several external factors, notably light
conditions, which often disrupt the results obtained by classifiers for human face
recognition.The approach proposed in this paper can be considered a potentially
strong candidate for use in a variety of commercial and industrial applications,
particularly those related to security. In fact, in addition to its usefulness in
identification processes, this face recognition system is also of particular
importance for those who are interested in search and navigation processes in
online video masses.In essence, our process was based on a corpus that contained
a huge number of faces acquired in different positions and various lighting
conditions. The identification and classification of faces was achieved through the
use of a Neural Network called the Multi-Layer Perceptron (MLP), which is one
of the most common networks used in this context.
Mat lab
Character Recognition using Geometrical Features of
Alphabet: A Novel Technique
Abstract: In this paper, we have shown that recognition of machine printed
characters can be modeled around the geometry of an alphabet. This geometry is
used as a basis for recognizing various characters. The present technique can be
implemented for any language, provided, the alphabets of the respective language
are made up of geometrical shapes, which is typically true.
An Improved Eye Localization Algorithm using Multi-cue
Facial Information
Abstract: In this paper, we propose a four step method to detect center of eyes
robustly and accurately. The method doesn’t put any restrictions on the
background. The method is based on Ada Boost algorithm for face and eye
candidate points detection. Candidate points are tuned such that two candidate
points are exactly in centers of irises. Mean crossing function and convolution
template are proposed to filter out candidate points and select iris pair. The
advantage of using this kind of hybrid method is that Ada Boost is robust to
different illumination conditions and backgrounds. The tuning step improves the
precision of iris localization while the convolution filter and mean crossing
function reliably filter out candidate points and select iris pair. The proposed
structure is evaluated on three public databases, Bern, Yale and Bio ID. Extensive
experimental results verified the robustness and accuracy of the proposed method.
Using the Bern database, the performance of the proposed algorithm is also
compared with some of the existing methods.
Mat lab
Improvement of Image Zooming Using Least Directional
Differences based on Linear and Cubic Interpolation
Abstract: There are many interpolation methods, among them; bilinear (BL) and
bicubic (BC) are more popular. However, these methods suffer from low quality
edge blurring and aliasing effect. In the other hand, if high resolution images are
not available, it is impossible to produce high quality display images and prints.
To overcome this drawback, in this paper, we proposed a new method that uses
least directional differences of neighbor pixels, based on proceeding bilinear and
bicubic interpolation methods for images. The qualitative and quantitative results
of proposed technique show that this method improves bilinear and bicubic
interpolations. The proposed algorithm can also be applied both to RGB and gray
level images.
Intelligent Feature-guided Multi-object Tracking Using
Kalman Filter
Abstract: Kalman filtering, a recursive state estimation filter is a robust method
for tracking objects. It has been proven that Kalman filter gives a good estimation
when tested on various tracking systems. However, unsatisfying tracking results
may be produced due to different real-time conditions. These conditions include:
inter-object occulusion and separation which are observed when objects are being
tracked in real-time. Thus, it is challenging to handle for the classical Kalman
filter. In this paper, we proposed an idea of intelligent feature-guided tracking
using Kalman filtering. A new method is developed named Correlation-Weighted
Histogram Intersection (CWHI), in which correlation weights are applied to
Histogram Intersection (HI) method. We focus on multi-object tracking in traffic
sequences and our aim is to achieve efficient tracking of multiple moving objects
under the confusing situations. The proposed algorithm achieves robust tracking
with 97.3% accuracy and 0.07% covariance error in different real-time scenarios.
Mat lab
USE OF LIP SYNCHRONIZATION BY HEARING
IMPAIRED USING DIGITAL IMAGE PROCESSING
FOR ENHANCED PERCEPTION OF SPEECH
Abstract: There are various means of electronic communication however only
limited inconvenient options are available to the hearing impaired subject. We
propose to teach the hearing impaired the correlation of lip movements with
information, in speech. Therefore, we propose a technique to help learn hearing
impaired primary school children to communicate using visual lip movement
cues, deploying MAT LAB software. Requisite programming have written in
MAT LAB, convenient software for image processing of visual lip movement
cues. Short video sequences availability, has been found to a major learning
resources for this techniques)
A Robust Image Encryption Scheme using State-Feedback
Control
Abstract: In recent years, the security of multimedia data transmitted over
insecure networks like the Internet has gained a lot of attention. Applying
standard cryptographic encryption and authentication algorithms to multimedia
data is not feasible because multimedia data may undergo certain content
preserving operations that although may change the data values but keep the
semantic of the data intact. In this paper, we propose a robust encryption
algorithm for digital images that not only provides confidentiality to the image
being transmitted or stored, but it can also tolerate JPEG compression. The
proposed algorithm is private key i.e. the encryption and decryption process
requires the same key. This paper presents an interesting approach by adapting
techniques from modern control theory to devise the proposed encryption and
decryption algorithms. The results reported in this paper are extremely
encouraging and reflects the point that how multi-disciplinary fields can be
combined to solve complex problems.
Mat lab
An effective Edge Detection Methodology for medical images
based on texture discrimination
Abstract: As medical images are fuzzy, edge detection based on texture
characteristics is comparatively effective than intensity based techniques. A new
methodology is described for texture edge detection in medical images that is
applicable across modalities. We use a multi-scale filter to capture texture edge
information. An experimental prototype based on the proposed methodology
provides a test bed for comparison with a popular edge detection technique.
Adaptive Objects Tracking by using Statistical Features
Shape Modeling and Histogram Analysis
Abstract: We propose a novel method for object tracking using an adaptive
algorithm based on statistical analysis of objects shape. To track objects in video
sequence, we use a system that combines two algorithms: a histogram analysis
algorithm and a statistical shape features modeling algorithm. The main
improvement of the proposed system with respect to the others present in
literature is that we do not use any a priori knowledge about how objects look
like. This no a-priori model has been carried out by computing a model that takes
into account the statistical behavior of the most important objects features over the
whole video frames. Moreover, an adaptive mechanism allows us to reset the
statistical model creation when such a model is too much dissimilar from the real
blobs features. Experiments on some real-world difficult scenarios of low
resolution videos and in unconstrained environments demonstrate the very
promising results achieved.
Mat lab
Color Image Retrieval Using M-Band Wavelet Transform
Based Color-Texture Feature
Abstract: Feature Extraction algorithm is a very important component of any
retrieval scheme. We propose M-band Wavelet Transform based feature
extraction algorithm in this paper. The MXM sub-bands are used as primitive
features, over which energies computed in a neighborhood are taken as the
features for each pixel of the image. These features are clustered using FCM to
obtain image signature for similarity matching using the Earth Mover’s Distance.
The results obtained were compared with MPEG-7 content descriptor based
system and found to be superior.
Bangla Speech Recognition System using LPC and ANN
Abstract: This paper presents the Bangla speech recognition system. Bangla
speech recognition system is divided mainly into two major parts. The first part is
speech signal processing and the second part is speech pattern recognition
technique. The speech processing stage consists of speech starting and end point
detection, windowing, filtering, calculating the Linear Predictive Coding (LPC)
and Cepstral Coefficients and finally constructing the codebook by vector
quantization. The second part consists of pattern recognition system using
Artificial Neural Network (ANN). Speech signals are recorded using an audio
wave recorder in the normal room environment. The recorded speech signal is
passed through the speech starting and end-point detection algorithm to detect the
presence of the speech signal and remove the silence and pauses portions of the
signals. The resulting signal is then filtered for the removal of unwanted
background noise from the speech signals. The filtered signal is then windowed
ensuring half frame overlap. After windowing, the speech signal is then subjected
to calculate the LPC coefficient and Cepstral coefficient. The feature extractor
uses a standard LPC Cepstrum coder, which converts the incoming speech signal
into LPC Cepstrum feature space. The Self Organizing Map (SOM) Neural
Network makes each variable length LPC trajectory of an isolated word into a
fixed length LPC trajectory and thereby making the fixed length feature vector, to
be fed into to the recognizer. The structures of the neural network is designed with
Multi Layer Perceptron approach and tested with 3, 4, 5 hidden layers using the
Transfer functions of Tanh Sigmoid for the Bangla speech recognition system.
Comparison among different structures of Neural Networks conducted here for a
better understanding of the problem and its possible solutions.
Cast Shadow Removal Using Time and Exposure Varying
Images
Abstract: Shadows are the natural accomplice of objects. As such, they have
affected various algorithms dealing with image segmentation, object tracking and
recognition. A lot of research has been focused on removing shadows from
images while preserving the information available in the shadow region. In this
paper, we present a simple yet robust algorithm for cast shadow removal utilizing
images taken at different times with different exposures. While different
exposures allow good recovery of shadow regions, the time-varying feature is
used to suppress the shadow edges.
Mat lab
Combining features for Shape and Motion Trajectory of
Video Objects for efficient Content based Video Retrieval
Abstract: This paper proposes a system for content based video retrieval based
on shape and motion features of the video object. We have used Curvature scale
space for shape representation and Polynomial curve fitting for trajectory
representation and retrieval. The shape representation is invariant to translation,
rotation and scaling and robust with respect to noise. Trajectory matching
incorporates visual distance, velocity dissimilarity and size dissimilarity for
retrieval. The cost of matching two video objects is based on shape and motion
features, to retrieve similar video shots. We have tested our system on standard
synthetic databases. We have also tested our system on real world databases.
Experimental results have shown good performance.
Detecting and tracking People in a Homogeneous
Environment using Skin Color Model
Abstract: The task of correctly identifying and tracking people in a shadow
environment for understanding the group dynamics is of paramount importance in
many vision systems. This work presents a real time system for detecting and
tracking people, in an environment where, people have similar attire. The
proposed frame work contains shadow removal in HSV color space, detection
through occlusion, person identification by developing skin color model and
tracking by extracting image features. Experimental results illustrate that the
proposed approach works robustly in homogeneous environment. Mat lab
Document Clustering for Event Identification and Trend
Analysis in Market News
Abstract: In this paper we have proposed a stock market analysis system that
analyzes financial news items to identify and characterize major events that
impact the market. The events have been identified using Latent Dirichlet
Allocation (LDA) based topic extraction mechanism. The topic-document data is
then clustered using kernel k means algorithm. The clusters are analyzed jointly
with the SENSEX raw data to extract major events and their effects. The system
has been implemented on capital market news about the Indian share market of
the past three years.
A Novel Face Recognition Method using Facial Landmarks
Abstract: In this paper we have presented a novel approach for face recognition.
The proposed method is based on the facial landmarks such as eyes, nose, lips etc.
In this approach, first the probable position of these landmarks is located from the
gradient image. Secondly, the template matching is employed over a region
around the probable positions to detect exact location of the landmarks. Then,
statistical and geometric features are extracted from these regions. To reduce the
dimension of the feature vector PCA is employed. In this experiment to classify
the images Mahalanobis distance is employed. The performance of the proposed
method is tested on ORL database. Mat lab
Facial Expression Recognition with Multi-Channel
Deconvolution
Abstract: Facial expression recognition is an important task in human computer
interaction systems to include emotion processing. In this work we present a
Multi-Channel Deconvolution method for post processing of face expression data
derived from video sequences. Photogrammetric techniques are applied to
determine real world geometric measures and to build the feature vector. SVM
classification is used to classify a limited number of emotions from the feature
vector. A Multi-Channel Deconvolution removes ambiguities at the transitions
between different classified emotions. This way, typical temporal behavior of
facial expression change is considered.
Feature Extraction Using Gabor-Filter and Recursive Fisher
Linear Discriminant with Application in Fingerprint
Identification
Abstract: Fingerprint is widely used in identification and verification systems.
In this paper, we present a novel feature extraction method based on Gabor filter
and Recursive Fisher Linear Discriminate (RFLD) algorithm, which is used for
fingerprint identification. Our proposed method is assessed on images from the
bio lab database. Experimental results show that applying RFLD to a Gabor filter
in four orientations, in comparison with Gabor filter and PCA transform, increases
the identification accuracy from 85.2% to 95.2% by nearest cluster center point
classifier with Leave-One-Out method. Also, it has shown that applying RFLD to
a Gabor filter in four orientations, in comparison with Gabor filter and PCA
transform, increases the identification accuracy from 81.9% to 100% by 3NN
classifier. The proposed method has lower computational complexity and higher
accuracy rates than conventional methods based on texture features.
Mat lab
Image denoising using edge model-based representation of
Laplacian sub bands
Abstract: This paper presents a novel method of removing unstructured,
spurious artifacts (more popularly called noise) from images. This method uses an
edge model-based representation of Laplacian sub bands and deals with noise at
Laplacian subband levels to reduce it effectively. As the prominent edges are
retained in their original form in the denoised images, the proposed method can be
classified as an edge preserving denoising scheme. Laplacian sub bands are
represented using a Primitive Set (PS) consisting of 7 × 7 sub images of sharp and
blurred Laplacian edge elements. The choice of edge model-based representation
provides greater flexibility in removing characteristic artifacts from noise sources
Multi sensor Biometric Evidence Fusion for Person
Authentication using Wavelet Decomposition and
Monotonic-Decreasing Graph
Abstract: This paper presents a novel biometric sensor generated evidence
fusion of face and palm print images using wavelet decomposition for personnel
identity verification. The approach of biometric image fusion at sensor level refers
to a process that fuses multi spectral images captured at different resolutions and
by different biometric sensors to acquire richer and complementary information to
produce a new fused image in spatially enhanced form. When the fused image is
ready for further processing, SIFT operator are then used for feature extraction
and the recognition is performed by adjustable structural graph matching between
a pair of fused images by searching corresponding points using recursive descent
tree traversal approach. The experimental result shows the efficacy of the
proposed method with 98.19% accuracy, outperforms other methods when it is
compared with uni-modal face and palm print authentication results with
recognition rates 89.04% and 92.17%, respectively and when all the methods are
processed in the same feature space.
Mat lab
Object Recognition in Presence of Blur from Fourier
Transform Spectrum
Abstract: A new technique for recognition of objects in presence of blur is
presented where the object features are chosen from the Fourier power spectrum
on the locus of equal intensity points around the zero order spectrums. Here the
property of 2D Fourier slice theorem and Fraunhofer distribution for 1D aperture
is used to show that the spread of Fourier transform spectrum along any direction
is related to the spread of the object along that direction. Computation is done on
different intensity profiles to show the validity of the technique.
On-line Signature Verification: An Approach based on
Cluster Representations of Global Features
Abstract: In this paper, we propose a new method of representation of on-line
signatures by clustering of signatures. Our idea is to provide better representation
by clustering of signatures based on global features. Global features of signatures
of each cluster are used to form an interval valued feature vector which is a
symbolic representation for a cluster. Based on cluster representation, we propose
methods of signature verification. We compare the feasibility of the proposed
representation scheme for signature verification on a large MCYT_ signature
database [1] of 16500 signatures. Unlike other signature verification methods, the
proposed method is simple and efficient and in addition, shows a remarkable
reduction in EER.
Mat lab
Online Character Recognition using Elastic Curvature
Matching
Abstract: An efficient method for online character recognition is suggested. It
consists of two steps: curvature extraction and curvature matching. The online
signal with a single stroke is a sequence of two-dimensional positional vectors
whereas its curvature is one-dimensional. Elastic curvature matching is basically a
1D-to-1D matching problem between curvatures of reference and test characters,
and one-dimensionality of curvature makes the matching problem more quick and
easy than 2D-to-2D matching. We show the results obtained from applying it to
online digit recognition and discuss them.
Region Based Image Fusion for Detection of Ewing Sarcoma
Abstract: In the medical image processing different sources of images are
providing complementary information so fusion of different source images will
give more details for diagnosis of patients. In this paper an automatic region based
image fusion algorithm is proposed which is applied on the registered Magnetic
Resonance (MR) image of human brain. The aim of this paper is to detect all the
information required for accurate diagnosis of a brain tumor namely, Ewing
sarcoma which is simultaneously not available in individual MR images. The
proposed region based image fusion method is applied on two types of MR
sequence images to extract useful information which is than compared with
different pixel based algorithm and the performance of these fusion schemes are
evaluated using standard quality assessment parameters. From the analysis of
quality assessment parameters we found that our scheme provides better result
compared to pixel based fusion scheme. The resultant fused image is assessed and
validated by radiologist.
Mat lab
Relevant and Redundant Feature Analysis with Ensemble
Classification
Abstract: Feature selection and ensemble classification increase system
efficiency and accuracy in machine learning, data mining and biomedical
informatics. This research presents an analysis of the effect of removing irrelevant
and redundant features with ensemble classifiers using two datasets from UCI
machine learning repository. Accuracy and computational time were evaluated by
four base classifiers; NaiveBayes, Multilayer Perceptron, Support Vector
Machines and Decision Tree. Eliminating irrelevant features improves accuracy
and reduces computational time while removing redundant features reduces
computational time and reduces accuracy of the ensemble.
A Secure Steganography Technique for Blind Steganalysis
Resistance
Abstract: A simple yet effective tactic for secure Steganography is proposed in
this paper that can resist the blind steganalysis. In this method author derives a
matrix based on the image content and thus providing the security. This matrix is
used by Quantization Index Modulation (QIM) based encoder and decoder. The
embedding location of data is also randomized so as to immobilize the self
calibration process. It is shown that detection rate of steganalysis scheme to
proposed method is close to arbitrary speculation.
Mat lab
Separation of Foreground Text from Complex Background
in Color Document Images
Abstract: Reading of the foreground text is difficult in documents having multi
colored complex background. Automatic foreground text separation in such
document images is very much essential for smooth reading of the document
contents. In this paper we propose a hybrid approach which combines connected
component analysis and an unsupervised thresholding for separation of text from
the complex background. The proposed approach identifies the candidate text
regions based on edge detection followed by a connected component analysis.
Because of background complexity it is also possible that a non text region may
be identified as a text region. To overcome this problem we extract texture
features of connected components and analyze the feature values. Finally the
threshold value for each detected text region is derived automatically from the
data of corresponding image region to perform foreground separation. The
proposed approach can handle document images with varying background of
multiple colors. Also it can handle foreground text of any color, font and size.
Experimental results show that the proposed algorithm detects on an average
97.8% of text regions in the source document. Readability of the extracted
foreground text is illustrated through OC Ring.
A Skin-Color and Template Based Technique for Automatic
Ear Detection
Abstract: This paper proposes an efficient skin-color and template based
technique for automatic ear detection in a side face image. The technique first
separates skin regions from non skin regions and then searches for the ear within
skin regions. Ear detection process involves three major steps. First, Skin
Segmentation to eliminate all non-skin pixels from the image, second Ear
Localization to perform ear detection using template matching approach, and third
Ear Verification to validate the ear detection using the Zernike moments based
shape descriptor. To handle the detection of ears of various shapes and sizes, an
ear template is created considering the ears of various shapes (triangular, round,
oval and rectangular) and resized automatically to a size suitable for the detection.
Proposed technique is tested on the IIT Kanpur ear database consisting of 150 side
face images and gives 94% accuracy.
Mat lab
Unsupervised Change Detection of Remotely Sensed Images
using Fuzzy Clustering
Abstract: In this paper two fuzzy clustering algorithms, namely Fuzzy C-Means
(FCM) and Gustafson Kessel Clustering (GKC), have been used for detecting
changes in multi temporal remote sensing images. Change detection maps are
obtained by separating the pixel-patterns of the difference image into two groups.
To show the effectiveness of the proposed technique, experiments are conducted
on three multi spectral and multi temporal images. Results are compared with
those of existing Markov Random Field (MRF) & neural network based
algorithms and found to be superior. The proposed technique is less time
consuming and unlike MRF do not need any a priori knowledge of distribution of
changed and unchanged pixels (as required by MRF).
Unsupervised Satellite Image Segmentation by Combining
SA based Fuzzy Clustering with Support Vector Machine
Abstract: Fuzzy clustering is an important tool for unsupervised pixel
classification in remotely sensed satellite images. In this article, a Simulated
Annealing (SA) based fuzzy clustering method is developed and combined with
popular Support vector Machine (SVM) classifier to fine tune the clustering
produced by SA for obtaining an improved clustering performance. The
performance of the proposed technique has been compared with that of some
other well-known algorithms for an IRS satellite image of the city of Kolkata and
its superiority has been demonstrated quantitatively and visually. Mat lab
A Modified Fuzzy C-Means Algorithm with Adaptive
Spatial Information for Color Image Segmentation
Abstract: Though FCM has long been widely used in image segmentation, it yet
faces several challenges. Traditional FCM needs a laborious process to decide
cluster center number by repetitive tests. Moreover, random initialization of
cluster centers can let the algorithm easily fall onto local minimum, causing the
segmentation results to be suboptimal. Traditional FCM is also sensitive to noise
due to the reason that the pixel partitioning process goes completely in the feature
space, ignoring some necessary spatial information. In this paper we introduce a
modified FCM algorithm for color image segmentation. The proposed algorithm
adopts an adaptive and robust initialization method which automatically decides
initial cluster center values and center number according to the input image. In
addition, by deciding the window size of pixel neighbor and the weights of
neighbor memberships according to local color variance, the proposed approach
adaptively incorporates spatial information to the clustering process and increases
the algorithm robustness to noise pixels and drastic color variance. Experimental
results have shown the superiority of modified FCM over traditional FCM
algorithm.
No comments:
Post a Comment