Aouatif Amine

Driver's fatigue and drowsiness detection to reduce traffic accidents on road

This paper proposes a robust and nonintrusive system for monitoring drivers fatigue and drowsiness in real time. The proposed scheme begins by extracting the face from the video frame using the Support Vector Machine (SVM) face detector. Then a new approach for eye and mouth state analysis -based on Circular Hough Transform (CHT)- is applied on eyes and mouth extracted regions. Our drowsiness analysis method aims to detect micro-sleep periods by identifying the iris using a novel method to characterize drivers eye state. Fatigue analysis method based on yawning detection is also very important to prevent the driver before drowsiness. In order to identify yawning, we detect wide open mouth using the same proposed method of eye state analysis. The system was tested with different sequences recorded in various conditions and with different subjects. Some experimental results about the performance of the system are presented.

Noise Reduction in Medical Images - comparison of noise removal algorithms -

The Medical community uses several image acquisition techniques for diagnosing and suggesting the corresponding therapies. Therefore the obtained images from clinical examinations should be treated to assist doctors in results interpretation. In this paper, we focus on denoising task in order to determine the benefits and drawbacks for each algorithm. For this, we used as database, images acquired from the most common techniques namely Magnetic Resonance (MR), Computed Tomography (CT), Ultrasounds, Scintigraphy and X-Ray. The effectiveness of discussed algorithms is compared on the basis of: Signal to Noise Ratio (SNR), Peak to Signal noise (PSNR), Root Mean square Error (RMSE) and the Mean Structure Similarity Index (MSSIM). Experimental results demonstrate that the NL-Means algorithm clearly outperforms the others denoising approaches for all noises levels.

The Performance of LBP and NSVC Combination Applied to Face Classification

The growing demand in the field of security led to the development of interesting approaches in face classification. These works are interested since their beginning in extracting the invariant features of the face to build a single model easily identifiable by classification algorithms. Our goal in this article is to develop more efficient practical methods for face detection. We present a new fast and accurate approach based on local binary patterns LBP for the extraction of the features that is combined with the new classifier Neighboring Support Vector Classifier NSVC for classification. The experimental results on different natural images show that the proposed method can get very good results at a very short detection time. The best precision obtained by LBP-NSVC exceeds 99%.

Study of the relative magnitude in the wavelet domain for texture characterization

Wavelet-based transforms have emerged as efficient directional multiscale schemes able to provide advanced analysis for the textural content of an image. Making use of their statistical dependencies, wavelet coefficients have been recognized as good basis for texture analysis. In this paper, we propose a new feature vector called relative magnitude (RM) which incorporates local statistical dependencies within the neighborhood of magnitude coefficients. Its discriminative power is evaluated on multiclass grayscale texture classification. The generalized Gaussian distribution and the Laplace Model are used to study the statistical behavior of the proposed feature vector. Experiments were conducted on textures from the VisTex, Brodatz, Outex_TC10, UMD, UIUC, and KTH_TIPS databases. Quantitative results demonstrate the efficiency of the RM feature vector for texture discrimination in the wavelet domain.

Towards a Novel Reidentification Method Using Metaheuristics

Tracking multiple moving objects in a video sequence can be formulated as a profile matching problem. Reidentifying a profile within a crowd is done by a matching process between the tracked person and the different moving individuals within the same frame. In that context, the feature matching task can be approximated to a search for the profile that maximizes a considered similarity measure. In this work, we introduce a novel Modified Cuckoo Search (MCS) based reidentification algorithm. A complex descriptor representing each moving person is built from different low level visual features such as the color and the texture components. We make use of a database that involves all previously detected descriptors, forming therefore a discrete search space where the sought solution is a descriptor and its quality is represented by its similarity to the query profile. The approach is evaluated within a multiple object tracking scenario, and a validation process using the normalized cross correlation method to accept or reject the obtained reidentification results is included. The experimental results show promising performances in terms of computation cost as well as reidentification rate.

Texture classification using relative phase and Gaussian mixture models in the complex wavelet domain

The importance of phase features for texture analysis has been earlier established for many image processing applications. However, the modeling of the phase data faces some difficulties as its information gathers data with rotating values and thus highly sensitive to distortions. Motivated by its ability to capture different shapes of histograms, in this communication, we propose the Gaussian Mixture Model (GMM) to characterize the behavior of relative phase. The Maximum-likelihood Estimator (MLE) is used to estimate the GMM parameters. To investigate the relevance of the GMM model for relative phase data, a feature vector incorporating the estimated parameters is proposed for a multiclass classification task. Experiments are conducted on textures from VisTex and Brodatz databases. Results demonstrate that the GMM model fit well relative phase data. In addition, higher rates of accuracy, precision and recall, 95.35%, 95.50% and 95.40%, respectively, were achieved for Brodatz textures using the proposed feature vector. This suggests the potential usefulness of the probabilistic proprieties for texture analysis.

Mobile object tracking using the modified cuckoo search

Nature is a great source of inspiration for explaining hard and complex problems using simple means in computer science. This chapter attempts to use artificial intelligence algorithms unlike the traditional classifiers for image classification in order to improve the robustness of object tracking and reidentification systems which have an extensive search space to increase their efficiency. The goal of this chapter is to introduce and investigate some of the many possible ways to use bioinspired algorithms for solving object tracking problems. This study projects the cuckoo search algorithm on two different applications for mobile object tracking, which can be considered an optimization problem. The main problem in that context is the execution time. The obtained experimental results prove the efficiency of such an algorithm to solve the problem in very reasonable execution time, providing a satisfying accuracy rate.