We present an efficient approach that merges the virtual objects into video sequences taken by a freely moving camera in a realistic manner. The composition is visually and geomet...
Guofeng Zhang, Xueying Qin, Xiaobo An, Wei Chen, H...
This paper presents a content-based approach for temporal segmentation of videos. Tracked objects are characterized by their 2D trajectories which are used in a meaningful way to ...
Alexandre Hervieu, Patrick Bouthemy, Jean-Pierre L...
Segmentation, tracking, and description extraction are important operations in smart camera surveillance systems. In this paper, a robust segmentation-and-descriptor based trackin...
The appearance of non-rigid objects detected and tracked in video streams is highly variable and therefore makes the identification of similar objects very complex. Furthermore, i...
We present here a prototype video analysis and retrieval system, called NeTra-V, that is being developed to build an object-based video representation for functionalities such as s...