Mohamed Hammami

Data Mining, Algorithms

Associate Professor, PhD
13.09

Publications

  • [Show abstract] [Hide abstract]
    ABSTRACT: The efficient application of current methods of shadow detection in video is hindered by the difficulty in defining their parameters or models and/or their application domain dependence. This paper presents a new shadow detection and removal method that aims to overcome these inefficiencies. It proposes a semi-supervised learning rule using a new variant of co-training technique for shadow detection and removal in uncontrolled scenes. The new variant both reduces the run-time through a periodical execution of a co-training process according to a novel temporal framework, and generates a more generic prediction model for an accurate classification. The efficiency of the proposed method is shown experimentally on a testbed of videos that were recorded by a static camera and that included several constraints, e.g., dynamic changes in the natural scene and various visual shadow features. The conducted experimental study produced quantitative and qualitative results that highlighted the robustness of our shadow detection method and its accuracy in removing cast shadows. In addition, the practical usefulness of the proposed method was evaluated by integrating it in a Highway Control and Management System software called RoadGuard.
    No preview · Article · Aug 2015 · Multimedia Tools and Applications
  • Source
    Hazar Mliki · Emna Fendri · Mohamed Hammami
    [Show abstract] [Hide abstract]
    ABSTRACT: Face recognition has become an accessible issue for experts as well as ordinary people as it is a focal non-interfering biometric modality. In this paper, we introduced a new approach to perform face recognition under varying facial expressions. The proposed approach consists of two main steps: facial expression recognition and face recognition. They are two complementary steps to improve face recognition across facial expression variation. In the first step, we selected the most expressive regions responsible for facial expression appearance using the Mutual Information technique. Such a process helps not only improve the facial expression classification accuracy but also reduce the features vector size. In the second step, we used the Principal Component Analysis (PCA) to build EigenFaces for each facial expression class. Then, a face recognition is performed by projecting the face onto the corresponding facial expression Eigenfaces. The PCA technique significantly reduces the dimensionality of the original space since the face recognition is carried out in the reduced Eigenfaces space. An experimental study was conducted to evaluate the performance of the proposed approach in terms of face recognition accuracy and spatial-temporal complexity.
    Full-text · Article · Jan 2015 · Journal of Signal Processing Systems
  • Mayssa Frikha · Emna Fendri · Mohamed Hammami
    [Show abstract] [Hide abstract]
    ABSTRACT: Appearance based person re-identification attracts the attention of researchers and presents an active research area for intelligent video surveillance systems. In this paper, we propose a new approach for person re-identification in multi-camera systems. This approach consists in computing a new person signature by extracting a texture descriptor, not from the entire body, but only from stripes selected automatically. In addition, in this work, unlike existing solutions using gray leveled body, we propose to compute the texture descriptor from HSV colored body. Our approach has been compared to state-of-the-art methods using the highly challenging VIPeR dataset. We prove from this comparative study, that the proposed approach improves both, time and quality performances of person re-identification.
    No preview · Chapter · Sep 2014
  • Mohamed Hammami · Salma Ben Jemaa · Hanene Ben-Abdallah
    [Show abstract] [Hide abstract]
    ABSTRACT: Palmprint recognition, as a reliable personal identification method, has received an increasing attention and became an area of intense research during recent years. In this paper, we propose a generic biometric system that can be adopted with or without contact depending on the capturing system to ensure public security based on palmprint identification. This system is based on a new global approach that focuses only on areas of the image having the most discriminating features for recognition. The presented new approach was evaluated experimentally on two large databases, namely,“CASIA-Palmprint” and “PolyU-Palmprint”; the results of this evaluation show promising results and demonstrate the effectiveness of the proposed approach.
    No preview · Article · Feb 2014 · Multimedia Tools and Applications
  • Salma Ben Jemaa · Mohamed Hammami · Hanene Ben-Abdallah
    [Show abstract] [Hide abstract]
    ABSTRACT: Hand detection is the first step of any hand biometric recognition process, which determines the outcome of the following treatments. In this study, the authors propose a robust method for hand detection without contact and without constraints on the capture environment. This method is based on a data-mining process for skin-colour modelling. The presented data-mining process offers several advantages like the choice of the most relevant colour axes and the automatic choice of the decision rules. To improve the achieved results of skin detection and to determine the hand region in the image, a succession of postprocessings was proposed. The authors hand detection method was evaluated experimentally on a real database, namely, `Sfax-Miracl hand database'; the outcomes of this evaluation show promising results and demonstrate the effectiveness of the proposed method.
    No preview · Article · Nov 2013 · IET Image Processing
  • Mohamed Hammami · Nadra Ben Romdhane · Hanene Ben-Abdallah
    [Show abstract] [Hide abstract]
    ABSTRACT: Lane detection and tracking are very crucial treatments in lane departure warning systems as they help the vehicle-mounted system to keep its lane. In this context, the authors' work aims to develop vision-based lane detection and tracking method to detect and track lane limits in highways and main roads. The authors' contribution focuses on the detection step. By exploiting the fact that, in an image, the road can be formed by linear and curvilinear portions, the authors propose two types of appropriate treatments to detect the lane limits. The authors' method offers high precision rates independently of the painted lane marking's characteristics, of the time of acquisition and in different weather conditions. Besides the challenges it overcomes, the authors' method has the advantage of operating with a timing complexity that is reasonable for real-time applications. As shown experimentally, compared to three leading methods from the literature, the authors' method has a higher efficiency.
    No preview · Article · Jul 2013
  • Source
    Hazar Mliki · Nesrine Fourati · Souhatt Smaoui · Mohamed Hammami
    [Show abstract] [Hide abstract]
    ABSTRACT: Over the last two decades, the advances in computer vision and pattern recognition power have opened the door to new opportunity of automatic facial expression recognition system. In this work, we have introduced a new feature-based approach for facial expressions recognition. The proposed approach provides full automatic solution to identify human expressions as well as overcoming facial expressions variation and intensity problems. Facial features component were automatically detected and segmented. Then, we have detected facial feature points which go with facial expression deformations. Afterwards, distances between these points were computed and used through Data mining technique to generate a set of relevant prediction rules able to classify facial expressions. We took into account the intensity of JOY expression. Thus, we have defined SMILE expression as the lowest intensity of JOY. Seven facial expression classes were defined: JOY, SMILE, SURPRISE, DISGUST, ANGER, SADNESS, and FEAR We have appraised experimental study to evaluate the performance of the proposed solution.
    Full-text · Conference Paper · May 2013
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, we introduce a new facial-expression analysis system designed to automatically recognize facial expressions, able to manage facial-expression intensity variation as well as reducing the doubt and confusion between facial-expression classes. Our proposed approach introduces a new method to segment efficiently facial feature contours using Vector Field Convolution (VFC) technique. Relying on the detected contours, we extract facial feature points which go with facial-expression deformations. Then we have modeled a set of distances among the detected points to define prediction rules through data mining technique. An experimental study was conducted to evaluate the performance of our proposed solution under varying factors.
    Full-text · Conference Paper · Jan 2013
  • S. Ben Jemaa · M. Hammami · H. Ben-Abdallah
    [Show abstract] [Hide abstract]
    ABSTRACT: Hand detection is the first step of any hand biometric recognition process, which determines the outcome of the following treatments. In this paper, we propose a robust method for hand detection without contact and without constraints on the capture environment for hand biometric applications. This method is based on a color based approach adopting a non-parametric modeling. Our contribution consists mainly in the choice of the most relevant color axes and the choice of the decision rules automatically using a process of the data-mining as a new philosophy of data processing. To improve the achieved results of skin detection and to determine the hand region in the image, a succession of post-processing was proposed. Our hand detection method was evaluated experimentally on a real database; the outcomes of this evaluation show promising results and demonstrate the effectiveness of the proposed method.
    No preview · Conference Paper · Jan 2013
  • Source
    Hazar Mliki · Mohamed Hammami · Hanêne Ben-Abdallah
    [Show abstract] [Hide abstract]
    ABSTRACT: This paper presents a new system to achieve face detection and tracking in video sequences. We have performed a combination between detection and tracking modules to overcome the different challenging problems that can occur while detecting or tracking faces. Our proposed system is composed of two modules: Face detection module and face tracking module. In the face detection module, we have used skin color and motion information to extract regions of interest and cut off false positive face. This filtering step has enhanced the next face tracking processing step, as it helps to avoid tracking false positive faces. Regarding tracking module, we have used face detection results to keep the face tracker updated. In order to carry on tracking face we have used particle filter technique which was adapted to track multiple faces. Moreover, each tracked face was described by a defined state: tracked, occluded, entered, left or stopped. The performance of our detect-track system was evaluated using several experiments. This evaluation proved the robustness of our face detection-track system as it supports automatic tracking with no need to manual initialization or re-initialization and reaches best performance to deal with different challenging problems.
    Full-text · Conference Paper · Sep 2012
  • Source
    Mliki Hazar · Hammami Mohamed · B.-A. Hanene
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper we deal with the problem of lowering down the difficulty of face detection in video. Most of the recently developed systems swap detection accuracy for higher speeds, or vice versa. We have proposed a robust approach which makes use of spatial and temporal information in video to reduce time execution and improve precision rate. Our experiments show that our proposed approach proves efficiency without sacrificing real-time performance which makes it well-suited for live video applications.
    Full-text · Conference Paper · Jul 2012
  • [Show abstract] [Hide abstract]
    ABSTRACT: A new branch of biometrics, hand recognition, has attracted increasing amount of attention in recent years. In this paper, we propose an approach of hand detection based skin color pixel for biometric applications using multi layer perceptron (MLP) neural network. This later has the ability to classify skin pixels belonging to people with different skin tones and captured in different lighting conditions and complex background environments. To improve the achieved results, a succession of post-processing was proposed. The choice of the database is an important step in testing a biometric process. For this, we build a database named "Sfax-Miracl hand database". This database contains a total of 1080 images having the advantage of being captured from freely posed hands in contact free settings. Various conducted experiments on this database show promising results and demonstrate the effectiveness of the proposed approach.
    No preview · Conference Paper · Jun 2012
  • [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, we present an algorithm for tracking objects in road traffic sequences which is based on coherent strategy. This strategy relies on two times processing. Firstly, a Short-Term Processing (STP) based on spatial analysis and multilevel region descriptors matching allows identification of objects interactions and particular objects states. Secondly, a Long- Term Processing (LTP) is applied to cope with track management issues. In fact LTP feedbacks objects and their corresponding regions in each frame to update tracked object attributes. In case of merging objects, attributes are obtained using Template matching. An experimental study by quantitative and qualitative evaluations shows that the proposed approach can deal with multiple rigid objects whose sizes vary over time. The obtained results prove that our method can provide an effective and stable road objects tracks.
    No preview · Conference Paper · Jun 2012
  • [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, we tackle the shadow problem in depth for a better foreground segmentation. We propose a novel variant of co-training technique for shadow detection and removal in uncontrolled scenes. This variant works according to a powerful temporal behavior. Setting co-training parameters is based on an extensive experimental study. The proposed co-training variant runs periodically to obtain more generic classifier, thus improving speed and classification accuracy. An experimental study by quantitative, qualitative and comparative evaluations shows that the proposed method can detect shadow robustly and remove the ‘cast' part accurately from videos recorded by a static camera and under several constraints.
    No preview · Conference Paper · Jun 2012
  • Nadra Ben Romdhane · Mohamed Hammami · Hanene Ben-Abdallah
    [Show abstract] [Hide abstract]
    ABSTRACT: Lane detection consists of detecting the lane limits where the vehicle carrying the camera is moving. The aim of this study is to propose a lane detection method through digital image processing. Morphological filtering, Hough transform and linear parabolic fitting are applied to realize this task. The results of our proposed method are compared with three proposed researches. The method presented here was tested on video sequences filmed by the authors on Tunisian roads, on a video sequence provided by Daimler AG as well as on the PETS2001 dataset provided by the Essex University. KeywordsDriver assistance system–Lane detection–Hough transform–Morphological filtering
    No preview · Chapter · Aug 2011
  • Nadra Ben Romdhane · Mohamed Hammami · Hanene Ben-Abdallah
    [Show abstract] [Hide abstract]
    ABSTRACT: Obstacle detection is an important component in driver assistance as it helps systems to locate obstacles and then to prevent collisions. The aim of this study is to develop an obstacle detection module through digital images processing. We present a hybrid stereo vision-based method that combines stereo matching and homographic transformation methods. We use a sparse matching method in order to get a rapid geometric representation of the road scene that allows us to extract the upper and lower parts of obstacles. According to the position of the lower part, our method uses either the dense stereo matching or the homographic transformation methods to extract the candidate obstacles regions. A verification test is performed to verify whether the retained region is an obstacle or not. In order to avoid collisions, we compute the distance to the preceding obstacle to maintain the vehicle carrying the camera at a safety distance. The method presented here was tested on DIPLODOC 1 road stereo sequence captured on a highway. The obtained results prove the efficiency of our proposed method.
    No preview · Article · Jun 2011
  • [Show abstract] [Hide abstract]
    ABSTRACT: Fast and accurate moving object segmentation in dynamic scenes is the first step in many computer vision applications. In this paper, we propose a new background modeling method for moving object segmentation based on dynamic matrix and spatio-temporal analyses of scenes. Our method copes with some challenges related to this field. A new algorithm is proposed to detect and remove cast shadow. A comparative study by quantitative evaluations shows that the proposed approach can detect foreground robustly and accurately from videos recorded by a static camera and which include several constraints. A Highway Control and Management System called RoadGuard is proposed to show the robustness of our method. In fact, our system has the ability to control highway by detecting strange events that can happen like vehicles suddenly stopped in roads, parked vehicles in emergency zones or even illegal conduct such as going out from the road. Moreover, RoadGuard is capable of managing highways by saving information about the date and time of overloaded roads.
    No preview · Article · Apr 2011 · Multimedia Tools and Applications
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The detection of moving objects is a basic task for computer vision system. The performances of these systems are not sufficient for many applications. One of the main reasons is that the moving objet detection task has many difficulties in dealing with various constraints like the variations of the environment. A great number of methods were already proposed. We classify contributions reported in the literature in four approaches with a categorization based on inter-frame processing they adopt methods based on Inter-Frame Difference (IFD), those based on Background Modeling (BM), methods based on the Optical Flow (OF), and hybrid methods. In this paper, we present our proposed methods to detect moving objects. The first is a hybrid method that combines the inter-images difference based on entropy image and optical flow computed by a local method with a hierarchical coarse-to-fine optical flow estimation. The second is an adaptive background modeling based on dynamic matrix and spatio-temporal analyses of scenes. A comparative study by quantitative evaluations shows that the proposed BM method can detects foreground robustly and accurately from videos recorded by a static camera and which include several constraints such as sudden and gradual illumination changes, shaking camera, background component changes, ghost, and foreground speed.
    Full-text · Article · Apr 2011 · Journal of Next Generation Information Technology
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, we propose a new approach, called RoadGuard, for Highway Control and Management System. RoadGuard is based on counting and tracking moving vehicles robustly. Our system copes with some challenges related to such application processing steps like shadow, ghost and occlusion. A new algorithm is proposed to detect and remove cast shadow. The occlusion and ghost problems are resolved by the adopted tracking technique. A comparative study by quantitative evaluations shows that the proposed approach can detect vehicles robustly and accurately from highway videos recorded by a static camera which include several constraints. In fact, our system has the ability to control highway by detecting strange events that can happen like sudden stopped vehicles in roads, parked vehicles in emergency zones or even illegal conduct such going out from the road. Moreover, RoadGuard is capable to manage highways by saving information about date and time of overloaded roads
    Full-text · Conference Paper · Jan 2011
  • Salma Ben Jemaa · Mohamed Hammami

    No preview · Conference Paper · Jan 2011

15 Following View all

64 Followers View all