Xavier Desurmont

Scene analysis for reducing motion JPEG 2000 video surveillance delivery bandwidth and complexity

In this paper, we propose a new object-based video coding/transmission system using the emerging motion JPEG 2000 standard for the efficient storage and delivery of video surveillance over low bandwidth channels. Some recent papers deal with JPEG 2000 coding/transmission based on the region of interest (ROI) feature and the multi-layer capability provided by this coding system. Those approaches allow delivering more quality for mobile objects (or ROI) than for the background when bandwidth is too narrow for a sufficient video quality. The method proposed here provides the same features while significantly improving the average bitrate/quality ratio of delivered video when cameras are static. We transmit only ROIs of each frame as well as an automatic estimation of the background at a lower frame rate in two separate motion JPEG 2000 streams. The frames are then reconstructed at the client side without the need of other external data. Our method provides both better video quality and reduced client CPU usage with negligible storage overhead. Video surveillance streams stored on the server are fully compliant with existing motion JPEG 2000 decoders.

Progressive Learning for Interactive Surveillance Scenes Retrieval

This paper tackles the challenge of interactively retrieving visual scenes within surveillance sequences acquired with fixed camera. Contrarily to todays solutions, we assume that no a-priori knowledge is available so that the system must progressively learn the target scenes thanks to interactive labelling of a few frames by the user. The proposed method is based on very low-cost features extraction and integrates relevance feedback, multiple-instance SVM classification and active learning. Each of these 3 steps runs iteratively over the session, and takes advantage of the progressively increasing training set. Repeatable experiments on both simulated and real data demonstrate the efficiency of the approach and show how it allows reaching high retrieval performances.

Content-based retrieval of video surveillance scenes

A novel method for content-based retrieval of surveillance video data is presented. The study starts from the realistic assumption that the automatic feature extraction is kept simple, i.e. only segmentation and low-cost filtering operations have been applied. The solution is based on a new and generic dissimilarity measure for discriminating video surveillance scenes. This weighted compound measure can be interactively adapted during a session in order to capture the users subjectivity. Upon this, a key-frame selection and a content-based retrieval system have been developed and tested on several actual surveillance sequences. Experiments have shown how the proposed method is efficient and robust to segmentation errors.

Browsing Sport Content through an Interactive H.264 Streaming Session

This paper builds on an interactive streaming architecture that supports both user feedback interpretation, and temporal juxtaposition of multiple video bitstreams in a single streaming session. As an original contribution, these functionalities can be exploited to offer improved viewing experience, when accessing football content through individual and potentially bandwidth constrained connections. Starting from a conventional broadcasted content, our system automatically splits the native content into non-overlapping and semantically consistent segments. Each segment is then divided into shots, based on conventional view boundary detection. Shots are finally splitted in small clips. These clips support our browsing capabilities during the whole playback in a temporally consistent way. Multiple versions are automatically created to render each clip. Versioning depends on the view type of the initial shot, and typically corresponds to the generation of zoomed in and spatially or temporally subsampled video streams. Clips are encoded independently so that the server can decide on the fly the version to send as a function of the semantic relevance of the segments (in a user-transparent basis, as inferred from video analysis or metadata) and the interactive user requests. Replaying certain game actions is also offered upon request. The streaming is automatically switched to the requested event. Later, the playback is resumed without any offset. The capabilities of our system rely on the H.264/AVC standard. We use soccer videos to validate our framework in subjective experiments showing the feasibility and relevance of our system.

Performance evaluation of real-time video content analysis systems in the CANDELA project

The CANDELA project aims at realizing a system for real-time image processing in traffic and surveillance applications. The system performs segmentation, labels the extracted blobs and tracks their movements in the scene. Performance evaluation of such a system is a major challenge since no standard methods exist and the criteria for evaluation are highly subjective. This paper proposes a performance evaluation approach for video content analysis (VCA) systems and identifies the involved research areas. For these areas we give an overview of the state-of-the-art in performance evaluation and introduce a classification into different semantic levels. The proposed evaluation approach compares the results of the VCA algorithm with a ground-truth (GT) counterpart, which contains the desired results. Both the VCA results and the ground truth comprise description files that are formatted in MPEG-7. The evaluation is required to provide an objective performance measure and a mean to choose between competitive methods. In addition, it enables algorithm developers to measure the progress of their work at the different levels in the design process. From these requirements and the state-of-the-art overview we conclude that standardization is highly desirable for which many research topics still need to be addressed.

On the performance evaluation of tracking systems using multiple pan-tilt-zoom cameras

Object tracking from multiple Pan Tilt Zoom (PTZ) cameras is an important task. This paper deals with the evaluation of the result of such a system. This performance evaluation is conducted by first considering the characterization of the PTZ parameters and then by the trajectories themselves. The camera parameters with be evaluated with the homography errors; the trajectories will be evaluated according to the location and miss-identification errors.

On the use of Video Content Analysis in ITS : A review from academic to commercial applications

Stand-alone cameras or CCTV networks are nowadays commonly present in public areas such as city centers, stores and more recently in transportation infrastructures. In the meantime, automatic processing of video data is a field of activity stirring up the utmost attention in the pattern recognition community; state-of-the-art advances in this area enable the reliable extraction of features and the investigation of numerous applications dedicated to ITS. A first obvious field of application of Video Content Analysis (VCA) consists in improving safety and security in transports context. Embedded VCA in vehicles can track pedestrians to avoid collisions, improving safety. Used in railway station, VCA are able to detect left luggage allowing to enhance security. Video streams available from such installations may also represent a useful source of information for statistical transportation applications, e.g. monitoring of road traffic conditions or providing accurate counting statistics in railway/subway stations. This paper proposes an overview of the VCA applications in terms of safety, security and efficiency for ITS, with a specific focus on the usability of such VCA systems (emerging research topics, state-of-the-art studies, already commercialized applications, etc).

Adaptive Metadata Management System for Distributed Video Content Analysis

Scientific advances in the development of video processing algorithms now allow various distributed and collaborative vision-based applications. However, the lack of recognised standard in this area drives system developers to build specific systems, preventing from e.g. content analysis components upgrade or system reuse in different environments. As a result, the need for a generic, context-independent and adaptive system for storing and managing video analysis results comes out as conspicuous. In order to address this issue, we propose a data schema-independent data warehouse backed by a multiagent system. This system relies on the semantic web knowledge representation format, namely the RDF, to guarantee maximum adaptability and flexibility regarding schema transformation and knowledge retrieval. The storage system itself, namely data warehouse, comes from the state-of-the-art technologies of knowledge management, providing efficient analysis and reporting capabilities within the monitoring system.

Communications challenges in the Celtic-BOSS project

The BOSS project [1] aims at developing an innovative and bandwidth efficient communication system to transmit large data rate communications between public transport vehicles and the wayside to answer to the increasing need from Public Transport operators for new and/ or enhanced on-board functionality and services, such as passenger security and exploitation such as remote diagnostic or predictive maintenance. As a matter of fact, security issues, traditionally covered in stations by means of video-surveillance are clearly lacking on-board trains, due to the absence of efficient transmission means from the train to a supervising control centre. Similarly, diagnostic or maintenance issues are generally handled when the train arrives in stations or during maintenance stops, which prevents proactive actions to be carried out. The aim of the project is to circumvent these limitations and offer a system level solution. This article focuses on the communication system challenges.

A generic flexible and robust approach for intelligent real-time video-surveillance systems

In this article we present a generic, flexible and robust approach for an intelligent real-time video-surveillance system. A previous version of the system was presented in [1]. The goal of these advanced tools is to provide help to operators by detecting events of interest in visual scenes and highlighting alarms and compute statistics. The proposed system is a multi-camera platform able to handle different standards of video inputs (composite, IP, IEEE1394 ) and which can basically compress (MPEG4), store and display them. This platform also integrates advanced video analysis tools, such as motion detection, segmentation, tracking and interpretation. The design of the architecture is optimised to playback, display, and process video flows in an efficient way for video-surveillance application. The implementation is distributed on a scalable computer cluster based on Linux and IP network. It relies on POSIX threads for multitasking scheduling. Data flows are transmitted between the different modules using multicast technology and under control of a TCP-based command network (e.g. for bandwidth occupation control). We report here some results and we show the potential use of such a flexible system in third generation video surveillance system. We illustrate the interest of the system in a real case study, which is the indoor surveillance.