Nadia Baaziz

Security and privacy protection for automated video surveillance

In this paper, we present an automated video surveillance system designed to 1) ensure efficient selective storage of data, 2) provide means for enhancing privacy protection, and 3) secure visual data against malicious attacks. The proposed solution is a 3-module system processing captured video data before storage. Salient motion detection is used to retain relevant sequences and identify regions of interest with potential privacy-sensitive details. Then, an invertible and secure privacy preserving process is performed using a DCT-based scrambling technique on selected regions. To secure visual data and allow for data authentication, a self-embedding watermarking technique is applied on each image sequence. It offers the capability of proving authenticity as well as locating manipulated regions. Furthermore, this technique is also able to recover and reconstruct a good approximation of original lost content. In addition to a low computational complexity, simulation results show the effectiveness of the whole system in achieving its goals in terms of security and privacy enhancement of automated video surveillance data.

Texture Modeling Using Contourlets and Finite Mixtures of Generalized Gaussian Distributions and Applications

In this paper, we develop a new framework for contourlet-based statistical modeling using finite Mixtures of Generalized Gaussian distributions ( MoGG). On the one hand, given the rich directional information provided by the contourlet transform (CT), we propose to use a redundant version of the CT, which describes texture structures more accurately. On the other hand, we use MoGG modeling of contourlet coefficients distribution, which allows for precise capturing of a wide range of histogram shapes and provides better description and discrimination of texture than single probability density functions (pdfs). Moreover, we propose three applications for the proposed approach, namely: (1) texture retrieval, (2) fabric texture defect detection, and 3) infrared (IR) face recognition. We compare two implementations of the CT: standard CT ( SCT) and redundant CT ( RCT). We show that the proposed approach yields better results in the applications studied compared to recent state-of-the-art methods.

Contourlet Domain Feature Extraction for Image Content Authentication

The achievement of multimedia content authentication by means of digital watermarking, while never easy, is further complicated by the continuing investigation of different ways of generating authentication signatures which survive specific acceptable manipulations. In this paper, we present a novel approach which takes advantage of a multiscale framework and directionality to extract the significant features of an image from its redundant contourlet transform. The introduced redundancy brings simplicity and accuracy for feature calculation. We also describe the applied postprocessing steps with the aim of stabilizing those features under acceptable image manipulations, namely lossy JPEG compression and additive noise corruption. Many experiments and comparative studies are performed to show the effectiveness of our technique in generating content signatures based on invariant image features, as well as to demonstrate its superiority when compared with a redundant wavelet approach

Laplacian pyramid versus wavelet decomposition for image sequence coding

Implementation of a set of pyramid transforms in a hierarchical approach for a real television sequence which consists of a 3-D data information set is presented. The investigation is based on several visual and objective comparative criteria, including the relevance to the motion information. Several pyramid transforms for image sequence coding are investigated and compared. The nonorthogonal pyramid transforms are separated from the orthogonal transforms. The nonorthogonal pyramid transforms are easily implemented, permit exact reconstruction, and accept the introduction of an error-compensation method if the different levels are coded but the nonorthogonality of the decomposition causes an increase in the number of transform samples and a high global entropy value (a disadvantage for data compression). The wavelet pyramid transform is more computationally complex but gives the best results for global entropy values. The quincunx pyramid seems to be the most interesting, since it features many advantages of the other including the visual representation of motion information through a pyramid sequence.<<ETX>>

Contourlet-based texture retrieval using a mixture of generalized gaussian distributions

We address the texture retrieval problem using contourletbased statistical representation. We propose a new contourlet distribution modelling using finite mixtures of generalized Gaussian distributions (MoGG). The MoGG allows to capture a wide range of contourlet histogram shapes, which provides better description and discrimination of texture than using single probability density functions (pdfs). We propose a model similarity measure based on Kullback-Leibler divergence (KLD) approximation using Monte-Carlo sampling methods. We show that our approach using a redundant contourlet transform yields better texture discrimination and retrieval results than using other methods of statistical-based wavelet/contourlet modelling.

Multiconstraint Wiener-based motion compensation using wavelet pyramids

We have investigated an original motion estimation method that exploits several frequency subbands of wavelet pyramids using a multigrid-multiconstraint strategy. A recursive and iterative solution based on the Wiener approach allows to take into account the reconstruction quality criterion that is crucial to image coding. Experiments show its performances in terms of compression ratio, reconstruction quality, and reduction of implementation complexity compared to a monoresolution case.

Attacks on collusion-secure fingerprinting for multicast video protocols

Achievement of digital rights management standards in digital multimedia distribution applications, while never easy, is further compounded by the persevering investigation of different ways of attacking. In this paper, we analyze a well-known fingerprinting scheme for video distribution. We show the vulnerability of this scheme to specific attacks, namely, the copy attack and the combined collusion with frame-dropping attack. Several experiments were performed to show the effects of theses attacks. Some solutions are suggested in order to prevent these attacks and to enhance the ability of the fingerprinting scheme to reach its most significant aims, namely, the copyright protection and detection of pirates. These solutions are based on valid watermarks incorporating digital signatures, and selective watermarking. Moreover, the application of this revised fingerprinting scheme on a well known multicast protocol yields a significant decrease in the bandwidth and gets around the problem of excessive length of c-secure identity strings.

Content-based image copy detection using dual signatures

We are interested in content-based copy detection of images as a means for protecting intellectual property. The proposed methodology makes use of the discrete cosine transform (DCT) of an averaged image to extract two complementary features, namely ordinal measures and sign information, yielding a dual signature, i.e., a compact feature vector ensuring efficient storage in the image database. Moreover, a specific similarity measurement scheme is designed to handle dual signature comparison during the image retrieval process. Simulation results show the proposed method to outperform two known copy detection methods in terms of retrieval accuracy. Many common image manipulations can be handled such as noise addition, image resizing, Gamma and contrast adjustment, slight shifting, image flipping and 180° rotation. Achieved retrieval rates are very high and confirm the superiority of the proposed scheme.

Automatic Fabric Defect Detection Using Learning-Based Local Textural Distributions in the Contourlet Domain

We propose a learning-based approach for automatic detection of fabric defects. Our approach is based on a statistical representation of fabric patterns using the redundant contourlet transform (RCT). The distribution of the RCT coefficients are modeled using a finite mixture of generalized Gaussians (MoGG), which constitute statistical signatures distinguishing between defective and defect-free fabrics. In addition to being compact and fast to compute, these signatures enable accurate localization of defects. Our defect detection system is based on three main steps. In the first step, a preprocessing is applied for detecting basic pattern size for image decomposition and signature calculation. In the second step, labeled fabric samples are used to train a Bayes classifier (BC) to discriminate between defect-free and defective fabrics. Finally, defects are detected during image inspection by testing local patches using the learned BC. Our approach can deal with multiple types of textile fabrics, from simple to more complex ones. Experiments on the TILDA database have demonstrated that our method yields better results compared with recent state-of-the-art methods.Note to Practitioners—Fabric defect detection is central to automated visual inspection and quality control in textile manufacturing. This paper deals with this problem through a learning-based approach. By opposite to several existing approaches for fabric defect detection, which are effective in only some types of fabrics and/or defects, our method can deal with almost all types of patterned fabric and defects. To enable both detection and localization of defects, a fabric image is first divided into local blocks, which are representative of the repetitive pattern structure of the fabric. Then, statistical signatures are calculated by modeling the distribution of coefficients of an RCT using the finite MoGG. The discrimination between defect-free and defective fabrics is then achieved through supervised classification of RCT-MoGG signatures based on expert-labeled examples of defective fabric images. Experiments have shown that our method yields very good performance in terms of defect detection and localization. In addition to its accuracy, inspection of images can be performed in a fully automatic fashion, whereas only labeled examples are initially required. Finally, our method can be easily adapted to a real-time scenario since defect detection on inspected images is performed at the block level, which can be easily parallelized through hardware implementation.

Impact of scan conversion methods on the performance of scalable video coding

The ability to flexibly access coded video data at different resolutions or bit rates is referred to as scalability. We are concerned here with the class of methods referred to as pyramidal embedded coding in which specific subsets of the binary data can be used to decode lower- resolution versions of the video sequence. Two key techniques in such a pyramidal coder are the scan-conversion operations of down-conversion and up-conversion. Down-conversion is required to produce the smaller, lower-resolution versions of the image sequence. Up- conversion is used to perform conditional coding, whereby the coded lower-resolution image is interpolated to the same resolution as the next higher image and used to assist in the encoding of that level. The coding efficiency depends on the accuracy of this up-conversion process. In this paper techniques for down-conversion and up-conversion are addressed in the context of a two-level pyramidal representation. We first present the pyramidal technique for spatial scalability and review the methods used in MPEG-2. We then discuss some enhanced methods for down- and up-conversion, and evaluate their performance in the context of the two-level scalable system.