Abstract. This work is concerned with real-time feature tracking for long video sequences. In order to achieve efficient and robust tracking, we propose two interrelated enhancements to the well-known Shi-Tomasi-Kanade tracker. Our first contribution is the integration of a linear illumination

This paper presents a method for computing strain images of a deformable object in a video sequence. The method includes two steps: in the first step, the motion data between a pair of video frames is generated using a robust optical flow al-gorithm. In the second step, a strain image is computed

Introduction: New technology is creating large stores of digital video. Real time processing of these huge data sets is extremely challenging; retrospective or forensic analysis creates even greater problems when one must rapidly examine hours or days of video from thousands of cameras. We develop

be accurately described using "heavy-tailed" distributions (e.g., Pareto); (2) the autocorrelation of the VBR video sequence decays hyperbolically (equivalent to long-range dependence) and can be modeled using self-similar processes. We combine our findings in a new (non-Markovian) source model

Abstract. In this paper we present an approach for automatically detecting and tracking humans in very long video sequences. The detection is based on background subtraction using a multi-mode Codeword method. We enhance this method both in terms of representation and in terms of automatically

Xenosaga: Episode 1 are learning machines. They get themselves learned and learned well, so that they get played long and hard by a great many people. This is how they and their designers survive and perpetuate themselves. If a game cannot be learned and even mastered at a certain level, it won

Abstract. In this paper, we propose a 2D feature tracking method that is stable to long video sequences. To improve the stability of long tracking, we use trajectory information about 2D features. We predict the expected feature states and compute a rough estimate of the feature location

-time features and integrate such representations with SVM classification schemes for recognition. For the purpose of evaluation we introduce a new video database containing 2391 sequences of six human actions performed by 25 people in four different scenarios. The presented results of action recognition justify

Imagine a video taken on a sunny beach, can a computer automatically tell what is happening in the scene? Can it identify different human activities in the video, such as water surfing, people walking and lying on the beach? To automatically classify or localize different actions in video sequences

Human action in video sequences can be seen as silhouettes of a moving torso and protruding limbs undergoing articulated motion. We regard human actions as three-dimensional shapes induced by the silhouettes in the space-time volume. We adopt a recent approach [14] for analyzing 2D shapes