Abstract: One of the main issues in Computer Vision is to extract the movement of one or several points or objects of interest in an image or video sequence to conduct any kind of study or control process. Different techniques to solve this problem have been applied in numerous areas such as surveillance systems, analysis of traffic, motion capture, image compression, navigation systems and others, where the specific characteristics of each scenario determine the approximation to the problem. This paper puts forward a Computer Vision based algorithm to analyze fish trajectories in high turbulence conditions in artificial structures called vertical slot fishways, designed to allow the upstream migration of fish through obstructions in rivers. The suggested algorithm calculates the position of the fish at every instant starting from images recorded with a camera and using neural networks to execute fish detection on images. Different laboratory tests have been carried out in a full scale fishway model and with living fishes, allowing the reconstruction of the fish trajectory and the measurement of velocities and accelerations of the fish. These data can provide useful information to design more effective vertical slot fishways.
Abstract: Segmentation of Magnetic Resonance Imaging (MRI) images is the most challenging problems in medical imaging. This paper compares the performances of Seed-Based Region Growing (SBRG), Adaptive Network-Based Fuzzy Inference System (ANFIS) and Fuzzy c-Means (FCM) in brain abnormalities segmentation. Controlled experimental data is used, which designed in such a way that prior knowledge of the size of the abnormalities are known. This is done by cutting various sizes of abnormalities and pasting it onto normal brain tissues. The normal tissues or the background are divided into three different categories. The segmentation is done with fifty seven data of each category. The knowledge of the size of the abnormalities by the number of pixels are then compared with segmentation results of three techniques proposed. It was proven that the ANFIS returns the best segmentation performances in light abnormalities, whereas the SBRG on the other hand performed well in dark abnormalities segmentation.
Abstract: As the use of registration packages spreads, the number of the aligned image pairs in image databases (either by manual or automatic methods) increases dramatically. These image pairs can serve as a set of training data. Correspondingly, the images that are to be registered serve as testing data. In this paper, a novel medical image registration method is proposed which is based on the a priori knowledge of the expected joint intensity distribution estimated from pre-aligned training images. The goal of the registration is to find the optimal transformation such that the distance between the observed joint intensity distribution obtained from the testing image pair and the expected joint intensity distribution obtained from the corresponding training image pair is minimized. The distance is measured using the divergence measure based on Tsallis entropy. Experimental results show that, compared with the widely-used Shannon mutual information as well as Tsallis mutual information, the proposed method is computationally more efficient without sacrificing registration accuracy.
Abstract: Effective estimation of just noticeable distortion (JND) for images is helpful to increase the efficiency of a compression algorithm in which both the statistical redundancy and the perceptual redundancy should be accurately removed. In this paper, we design a DCT-based model for estimating JND profiles of color images. Based on a mathematical model of measuring the base detection threshold for each DCT coefficient in the color component of color images, the luminance masking adjustment, the contrast masking adjustment, and the cross masking adjustment are utilized for luminance component, and the variance-based masking adjustment based on the coefficient variation in the block is proposed for chrominance components. In order to verify the proposed model, the JND estimator is incorporated into the conventional JPEG coder to improve the compression performance. A subjective and fair viewing test is designed to evaluate the visual quality of the coding image under the specified viewing condition. The simulation results show that the JPEG coder integrated with the proposed DCT-based JND model gives better coding bit rates at visually lossless quality for a variety of color images.
Abstract: We constructed a method of noise reduction for
JPEG-compressed image based on Bayesian inference using the
maximizer of the posterior marginal (MPM) estimate. In this method,
we tried the MPM estimate using two kinds of likelihood, both of
which enhance grayscale images converted into the JPEG-compressed
image through the lossy JPEG image compression. One is the
deterministic model of the likelihood and the other is the probabilistic
one expressed by the Gaussian distribution. Then, using the Monte
Carlo simulation for grayscale images, such as the 256-grayscale
standard image “Lena" with 256 × 256 pixels, we examined the
performance of the MPM estimate based on the performance measure
using the mean square error. We clarified that the MPM estimate via
the Gaussian probabilistic model of the likelihood is effective for
reducing noises, such as the blocking artifacts and the mosquito noise,
if we set parameters appropriately. On the other hand, we found that
the MPM estimate via the deterministic model of the likelihood is not
effective for noise reduction due to the low acceptance ratio of the
Metropolis algorithm.
Abstract: Like any sentient organism, a smart environment
relies first and foremost on sensory data captured from the real
world. The sensory data come from sensor nodes of different
modalities deployed on different locations forming a Wireless Sensor
Network (WSN). Embedding smart sensors in humans has been a
research challenge due to the limitations imposed by these sensors
from computational capabilities to limited power. In this paper, we
first propose a practical WSN application that will enable blind
people to see what their neighboring partners can see. The challenge
is that the actual mapping between the input images to brain pattern
is too complex and not well understood. We also study the
connectivity problem in 3D/2D wireless sensor networks and propose
distributed efficient algorithms to accomplish the required
connectivity of the system. We provide a new connectivity algorithm
CDCA to connect disconnected parts of a network using cooperative
diversity. Through simulations, we analyze the connectivity gains
and energy savings provided by this novel form of cooperative
diversity in WSNs.
Abstract: This paper presents a technique for diagnosis of the abdominal aorta aneurysm in magnetic resonance imaging (MRI) images. First, our technique is designed to segment the aorta image in MRI images. This is a required step to determine the volume of aorta image which is the important step for diagnosis of the abdominal aorta aneurysm. Our proposed technique can detect the volume of aorta in MRI images using a new external energy for snakes model. The new external energy for snakes model is calculated from Law-s texture. The new external energy can increase the capture range of snakes model efficiently more than the old external energy of snakes models. Second, our technique is designed to diagnose the abdominal aorta aneurysm by Bayesian classifier which is classification models based on statistical theory. The feature for data classification of abdominal aorta aneurysm was derived from the contour of aorta images which was a result from segmenting of our snakes model, i.e., area, perimeter and compactness. We also compare the proposed technique with the traditional snakes model. In our experiment results, 30 images are trained, 20 images are tested and compared with expert opinion. The experimental results show that our technique is able to provide more accurate results than 95%.
Abstract: Random and natural textures classification is still
one of the biggest challenges in the field of image processing and
pattern recognition. In this paper, texture feature extraction using
Slant Hadamard Transform was studied and compared to other
signal processing-based texture classification schemes. A
parametric SHT was also introduced and employed for natural
textures feature extraction. We showed that a subtly modified
parametric SHT can outperform ordinary Walsh-Hadamard
transform and discrete cosine transform. Experiments were carried
out on a subset of Vistex random natural texture images using a
kNN classifier.
Abstract: This paper proposes a novel multi-format stream grid
architecture for real-time image monitoring system. The system, based
on a three-tier architecture, includes stream receiving unit, stream
processor unit, and presentation unit. It is a distributed computing and
a loose coupling architecture. The benefit is the amount of required
servers can be adjusted depending on the loading of the image
monitoring system. The stream receive unit supports multi capture
source devices and multi-format stream compress encoder. Stream
processor unit includes three modules; they are stream clipping
module, image processing module and image management module.
Presentation unit can display image data on several different platforms.
We verified the proposed grid architecture with an actual test of image
monitoring. We used a fast image matching method with the
adjustable parameters for different monitoring situations. Background
subtraction method is also implemented in the system. Experimental
results showed that the proposed architecture is robust, adaptive, and
powerful in the image monitoring system.
Abstract: In this paper we present a novel approach for density estimation. The proposed approach is based on using the logistic regression model to get initial density estimation for the given empirical density. The empirical data does not exactly follow the logistic regression model, so, there will be a deviation between the empirical density and the density estimated using logistic regression model. This deviation may be positive and/or negative. In this paper we use a linear combination of Gaussian (LCG) with positive and negative components as a model for this deviation. Also, we will use the expectation maximization (EM) algorithm to estimate the parameters of LCG. Experiments on real images demonstrate the accuracy of our approach.
Abstract: Duplicated region detection is a technical method to
expose copy-paste forgeries on digital images. Copy-paste is one
of the common types of forgeries to clone portion of an image
in order to conceal or duplicate special object. In this type of
forgery detection, extracting robust block feature and also high
time complexity of matching step are two main open problems.
This paper concentrates on computational time and proposes a local
block matching algorithm based on block clustering to enhance time
complexity. Time complexity of the proposed algorithm is formulated
and effects of two parameter, block size and number of cluster, on
efficiency of this algorithm are considered. The experimental results
and mathematical analysis demonstrate this algorithm is more costeffective
than lexicographically algorithms in time complexity issue
when the image is complex.
Abstract: Gabor-based face representation has achieved enormous success in face recognition. This paper addresses a novel algorithm for face recognition using neural networks trained by Gabor features. The system is commenced on convolving a face image with a series of Gabor filter coefficients at different scales and orientations. Two novel contributions of this paper are: scaling of rms contrast and introduction of fuzzily skewed filter. The neural network employed for face recognition is based on the multilayer perceptron (MLP) architecture with backpropagation algorithm and incorporates the convolution filter response of Gabor jet. The effectiveness of the algorithm has been justified over a face database with images captured at different illumination conditions.
Abstract: This paper is devoted to present and discuss a model that allows a local segmentation by using statistical information of a given image. It is based on Chan-Vese model, curve evolution, partial differential equations and binary level sets method. The proposed model uses the piecewise constant approximation of Chan-Vese model to compute Signed Pressure Force (SPF) function, this one attracts the curve to the true object(s)-s boundaries. The implemented model is used to extract weld defects from weld radiographic images in the aim to calculate the perimeter and surfaces of those weld defects; encouraged resultants are obtained on synthetic and real radiographic images.
Abstract: Modern culture, based on disinhibition of cultural trends and on heterodirection, is promoting openmindedness attitudes towards ethnic diversity, but on the other hand also new forms of social representations of the foreigner. Social representation is situated between the psychic field and the social one; it is the representation of oneself and of the other one, hanging between social categories and individual inner world. We will produce the results of a research on the representation of the foreigner, built on the type of prejudice prevailing among middle-low or middle-high educational qualification subjects, in which prejudicial attitudes seem to descend from precise mental images of the foreigner.
Abstract: This paper presents an authoring tool which makes a
user easily and intuitively design vibrotactile sensation. A mobile
hardware platform powered by ANDROID, a multi-purpose haptic
driver and a linear resonance actuator are used to implement the
system of the presented authoring tool. The tool allows users to easily
and simply create a vibrotactile sensation by drawing vibrotactile
images and to feel the sensation by rubbing drawn images on the touch
screen of a mobile device. The tool supports a graphical interface for
designing, editing and playing vibrotactile images as well as a
pre-defined file format for save and open.
Abstract: Counting people from a video stream in a noisy environment is a challenging task. This project aims at developing a counting system for transport vehicles, integrated in a video surveillance product. This article presents a method for the detection and tracking of multiple faces in a video by using a model of first and second order local moments. An iterative process is used to estimate the position and shape of multiple faces in images, and to track them. the trajectories are then processed to count people entering and leaving the vehicle.
Abstract: We present here the results for a comparative study of
some techniques, available in the literature, related to the relevance
feedback mechanism in the case of a short-term learning. Only one
method among those considered here is belonging to the data mining
field which is the K-nearest neighbors algorithm (KNN) while the
rest of the methods is related purely to the information retrieval field
and they fall under the purview of the following three major axes:
Shifting query, Feature Weighting and the optimization of the
parameters of similarity metric. As a contribution, and in addition to
the comparative purpose, we propose a new version of the KNN
algorithm referred to as an incremental KNN which is distinct from
the original version in the sense that besides the influence of the
seeds, the rate of the actual target image is influenced also by the
images already rated. The results presented here have been obtained
after experiments conducted on the Wang database for one iteration
and utilizing color moments on the RGB space. This compact
descriptor, Color Moments, is adequate for the efficiency purposes
needed in the case of interactive systems. The results obtained allow
us to claim that the proposed algorithm proves good results; it even
outperforms a wide range of techniques available in the literature.
Abstract: This paper presents a general trainable framework
for fast and robust upright human face and non-human object
detection and verification in static images. To enhance the
performance of the detection process, the technique we develop is
based on the combination of fast neural network (FNN) and
classical neural network (CNN). In FNN, a useful correlation is
exploited to sustain high level of detection accuracy between input
image and the weight of the hidden neurons. This is to enable the
use of Fourier transform that significantly speed up the time
detection. The combination of CNN is responsible to verify the
face region. A bootstrap algorithm is used to collect non human
object, which adds the false detection to the training process of the
human and non-human object. Experimental results on test images
with both simple and complex background demonstrate that the
proposed method has obtained high detection rate and low false
positive rate in detecting both human face and non-human object.
Abstract: Image mosaicing techniques are usually employed to offer researchers a wider field of view of microscopic image of biological samples. a mosaic is commonly achieved using automated microscopes and often with one “color" channel, whether it refers to natural or fluorescent analysis. In this work we present a method to achieve three subsequent mosaics of the same part of a stem cell culture analyzed in phase contrast and in fluorescence, with a common non-automated inverted microscope. The mosaics obtained are then merged together to mark, in the original contrast phase images, nuclei and cytoplasm of the cells referring to a mosaic of the culture, rather than to single images. The experiments carried out prove the effectiveness of our approach with cultures of cells stained with calcein (green/cytoplasm and nuclei) and hoechst (blue/nuclei) probes.
Abstract: Ethnicity identification of face images is of interest in
many areas of application, but existing methods are few and limited.
This paper presents a fusion scheme that uses block-based uniform
local binary patterns and Haar wavelet transform to combine local
and global features. In particular, the LL subband coefficients of the
whole face are fused with the histograms of uniform local binary
patterns from block partitions of the face. We applied the principal
component analysis on the fused features and managed to reduce the
dimensionality of the feature space from 536 down to around 15
without sacrificing too much accuracy. We have conducted a number
of preliminary experiments using a collection of 746 subject face
images. The test results show good accuracy and demonstrate the
potential of fusing global and local features. The fusion approach is
robust, making it easy to further improve the identification at both
feature and score levels.