Formulation and Use of Time-to-Collision in Content-Based Retrieval

Main Article Content

Ankush Mittal
Loong-Fah Cheong

Abstract

Time-To-Collision (TTC) is a common vision feature used to avoid obstacles, and is frequently used in cinematography to induce emotional effects. In this paper, we show how TTC can lead to several high-level video categories. The extracted TTC shots (low-level feature) are mapped to their corresponding high-level indices. The information conveyed by neighboring frames or shots (i.e., contextual information) is used to facilitate the mapping process. Our proposed TTC detection algorithm is based on computing TTC from the divergence of the image velocity field. A simple and novel method known as the pilot cue is used to further refine our algorithm. The experiments results are promising and show the effectiveness of the approach.

Article Details

How to Cite
Mittal, A., & Cheong, L.-F. (2007). Formulation and Use of Time-to-Collision in Content-Based Retrieval. INFOCOMP Journal of Computer Science, 6(1), 18–29. Retrieved from https://infocomp.dcc.ufla.br/index.php/infocomp/article/view/160
Section
Articles