Henri Nicolas

Collusion, MPEG4 compression and frame dropping resistant video watermarking

We present in this article a new video watermarking which resists collusion, MPEG4 compression and frame dropping attacks. This scheme is based on video mosaicing. For that, we are going to start by describing the mosaicing technique in order to illustrate the contribution of this technique in video watermarking. In fact, mosaicing allows to select an interesting area where the mark should be embedded. The idea is to insert the same mark into the same pixels which represent the same physical point. This is exactly the information which can be provided by a mosaic image at least for the located points in the scene background. Next, we present extensive experimental simulations which prove the watermark imperceptibility and robustness against several video attacks.

An Approach to Trajectory Estimation of Moving Objects in the H.264 Compressed Domain

This paper presents a simple and fast method for unsupervised trajectory estimation of multiple moving objects within a video scene. It is entirely based on the motion vectors that are present in compressed H.264/AVC or SVC video streams. We extract these motion vectors, perform robust frame-wise global motion estimation and use these estimates to form outlier masks. Motion segmentation on the spatio-temporally filtered outlier masks is performed to detect moving regions in the scene, which are analyzed over time in order to identify similar objects in adjacent frames. The construction of so-called Object History Images (OHIs) is proposed to stabilize the trajectories, which are finally interpolated with X-splines. The system enables real-time analysis with standard hardware.

Collusion-Resistant Video Watermarking Based on Video Mosaicing

With the rapid development of multimedia applications, and the fact that a large quantity of multimedia data is distributed via the Internet network, the protection of video becomes a major problem. It is therefore necessary to protect them. This can be done using video watermarking methods. To be really efficient, the mark has to be resistant to malicious attacks. Methods developed for still images cannot be efficiently used for video. Effectively, if a mark is embedded in only one image of the video, a temporal filtering along the motion displacement can easily removed it without significant degradation of the video. It is therefore necessary to define specific techniques for video. In this context, we present in this paper a video watermarking method based on the use of video mosaicing. We show that the proposed method is resistant to temporal filtering attack

A multi-resolution particle filter tracking in a multi-camera environment

This paper presents a novel tracking method with the multi-resolution technique and a Kolmogrov-Smirnov test for model update to track a non-rigid target in an uncalibrated multi-camera environment. It is based on particle filter method using color appearance model. Compared to the related work, our method improves the tracking performance by proposing: i) a multi-resolution technique to rapidly locate the estimate of the target state and refine it gradually, ii) the Kolmogrov-Smirnov test to evaluate the reliability of the estimate so as to take the decision on further updating/ reinitialization of the estimate, as well as iii) an interaction of cameras approach to reinitialize the estimate by information detected in other cameras in case of tracking failures. After being tested in a multi-camera environment for one person tracking, our system is shown to give a better tracking result in comparison with mono-camera tracking, especially when occlusions occur.

Compressed domain aided analysis of traffic surveillance videos

We present a novel system to perform efficient, compressed domain aided video analysis in the context of traffic surveillance applications. After camera installation, the system initializes by performing two short and fully automatic learning stages to gather information about the background and the principal moving directions in the scene. This knowledge is later used to assist the detection and tracking of vehicles. We combine processing in the pixel domain on decoded I-frames with motion based information from the H.264/SVC compressed domain in order to obtain a hybrid solution that delivers robust results at low computational complexity. Pan-Tilt-Zoom cameras are supported by the system, since global motion estimation is performed using the motion vectors that are present in the compressed stream.

Improved gaussian mixture model for the task of object tracking

This paper presents various motion detection methods: temporal averaging (TA), Bayes decision rules (BDR), Gaussian mixture model (GMM), and improved Gaussian mixture model (iGMM). This last model is improved by adapting the number of selected Gaussian, detecting and removing shadows, handling stopped object by locally modifying the updating process. Then we compare these methods on specific cases, such as lighting changes and stopped objects. We further present four tracking methods. Finally, we test the two motion detection methods offering the best results on an object tracking task, in a traffic monitoring context, to evaluate these methods on outdoor sequences.

Adaptive Video Watermarking using Mosaic Images

The watermarking is a technique that gives protection to multimedia data such as images and video. Actually, their protection from malicious manipulations as unauthorized copying and illegal distribution becomes a necessity. A simple re-used of still images watermarking techniques for video watermarking are not sufficiently efficient. In fact, if the mark is embedded in every frame of the video without taken into account the strong temporal correlation between them, a simple temporal filtering can destroy this mark. In this paper, we present an adaptive video watermarking using mosaic images which resists this attack. This method is also tested against MPEG compression.

Object-based video coding using a dynamic coding approach

In this paper we propose a video-object based coding scheme using dynamic coding. The principle of dynamic coding is to set on competition different coders on each video object. Thus, we are proposing a video-object based dynamic coding scheme using four completely different coders. The novelty of our work firstly comprise a global rate-distortion optimization enabling an optimal selection of a coder and its parameters for each object, and secondly the definition of a distortion metric. Our work is thus confirms that dynamic coding is efficient. It shows that a video object based coding approach is competitive. It improves object based video coders such as MPEG4 and it gives interesting comparison results between different state-of-the-art coders.

Human Behaviour Analysis and Event Recognition at a Point of Sale

This paper presents a new application that aims at improving communication and interactions between digital media and customers at a point of sale. Our system analyzes in real-time human behaviour while shopping. In particular, the system detects customer’s interest in products and interaction such as people grabbing products. This system is based on a behaviour model. A video analysis module detects motion, tracks moving object, and describes local motion. Then specific behaviours are recognized and sentences are generated. Finally, our approach is tested on real video sequences.

Object-based motion classification

With the development of object-based video representations such as MPEG-4, object based video manipulations applications become more and more important, such as 2D video scene composition. Semi-automatic algorithms designed to perform that task use meaningful informations about the behavior of the video object in the original sequence. This article proposes a method to split the sequence in temporal segments in which the object-camera orientation is either constant or changing. Then, a key view can be extracted for each segment. This video object sequence decomposition relies on the meaning of the affine motion model parameters obtained with a robust estimation algorithm.