Recent Advances in MPEG-7 Cameras Frederic Dufaux and Touradj Ebrahimi Institut de Traitement des Signaux Ecole Polytechnique Fédérale de Lausanne (EPFL) CH-1015 Lausanne, Switzerland Frederic.Dufaux@epfl.ch, Touradj.Ebrahimi@epfl.ch Emitall Surveillance S.A. CH-1820 Montreux, Switzerland Frederic.Dufaux@emitall.com, Touradj.Ebrahimi@emitall.com ABSTRACT We propose a smart camera which performs video analysis and generates an MPEG-7 compliant stream. By producing a content-based metadata description of the scene, the MPEG-7 camera extends the capabilities of conventional cameras. The metadata is then directly interpretable by a machine. This is especially helpful in a number of applications such as video surveillance, augmented reality and quality control. As a use case, we describe an algorithm to identify moving objects and produce the corresponding MPEG-7 description. The algorithm runs in real-time on a Matrox Iris P300C camera. Keywords: smart camera, MPEG-7, video analysis, content description 1. INTRODUCTION Recent advancements in acquisition, data processing and storage technologies have enabled the emergence of a new class of cameras, referred to as smart cameras, which perform video analysis to generate a content-based metadata description of the scene. These smart cameras are especially suitable for applications such as video surveillance, augmented reality or industrial manufacturing. In these applications, the camera output is meant to be received by a machine rather than by a human being. In this case, it is advantageous to extract the information relevant for the target application directly in the camera. It is then sufficient to only transmit this information, which is directly interpretable by a machine, unlike a video signal. An MPEG-7 smart camera has first been introduced in [1]. The MPEG-7 camera is a particular case of a smart camera which produces a stream compliant with the MPEG-7 Multimedia Content Description Standard [2][3]. The Eptacam MPEG-7 network camera [4] is based on the same concept. In this paper, we propose an MPEG-7 camera using a Matrox Iris P300C camera [5]. It features an Intel 400 MHz Celeron processor, enabling embedded image processing. As a use case, we consider a system which identifies moving objects, extracts their shape, and generates the corresponding MPEG-7 description. The algorithm runs in real-time on the camera. This paper is structured as follow. We first introduce the concept of smart camera in Sec. 2. An overview of MPEG-7 and the MPEG-7 camera is given in Sec. 3. An implementation example is described in Sec. 4, where as experimental results are proposed in Sec. 5. Finally, we draw some concluding remarks in Sec. 6.