The two volume set LNCS 4843 and LNCS 4844 constitutes the refereed proceedings of the 8th Asian Conference on Computer Vision, ACCV 2007, held in Tokyo, Japan, in November 2007.

The 46 revised full papers, 3 planary and invited talks, and 130 revised poster papers of the two volumes were carefully reviewed and seleceted from 551 submissions. The papers of this volume are organized in topical sections on shape and texture, fitting, calbration, detection, image and video processing, applications, face and gesture, tracking, camera networks, face/gesture/action detection and recognition, learning, motion and tracking, retrival and search, human pose estimation, matching, face/gesture/action detection and recognition, low level vision and phtometory, motion and tracking, human detection, and segmentation.

The two volume set LNCS 4843 and LNCS 4844 constitutes the refereed proceedings of the 8th Asian Conference on Computer Vision, ACCV 2007, held in Tokyo, Japan, in November 2007.

The 46 revised full papers, 3 planary and invited talks, and 130 revised poster papers of the two volumes were carefully reviewed and seleceted from 551 submissions. The papers of this volume are organized in topical sections on shape and texture, fitting, calbration, detection, image and video processing, applications, face and gesture, tracking, camera networks, face/gesture/action detection and recognition, learning, motion and tracking, retrival and search, human pose estimation, matching, face/gesture/action detection and recognition, low level vision and phtometory, motion and tracking, human detection, and segmentation.

Vision is a very important sense to humans and animals. It is mimicked for robots with cameras. However, conventional cameras have a small field of view, which implies that important difficulties might be encountered (such as motion ambiguity, occlusion and lack of information), particularly in robotics. This book is therefore dedicated to omnidirectional vision systems, especially catadioptric cameras, which permit these difficulties to be overcome. A catadioptric camera is a special type of omnidirectional system: it is composed of a camera and a convex mirror, permitting a great enlargement of the field of view and thus acquiring much more information from the environment. The objective of this book is to study the role of omnidirectional vision for robotic applications such as low-level feature extraction, target tracking, motion estimation and 3D reconstruction, while addressing the fundamental issues pertaining to omnidirectional vision. This book is accompanied with "OmniToolbox v2.1," a Matlab toolbox dedicated to omnidirectional vision (freely available on the website of the authors), so that the readers of this book can start working with omnidirectional images directly.

The goal of object recognition is to label objects from images and to estimate the poses of the labeled objects. The field of object recognition has seen tremendous progress with successful applications in some specific domains such as face recognition. However, the current state-of-the-art methods show unsatisfactory results for more general object domains in complex natural environments with visual ambiguities. In this dissertation, we aim to enhance the object identification and categorization with the guide of visual context and graphical model. In this work, we propose a general framework for the cooperative object identification and object categorization. Examplars used in identification provide useful information of similarity in categorization. Conversely, novel objects are rejected in identification but the proposed object categorization can label the novel objects and segment them out for database update in identification. This work can be helpful to the engineers in artificial intelligence and machine vision.