The light-weighting realization of video structured description on android-based terminals

In this paper,we introduced the application of mobile device on video data acquisition and intelligent content-wise analysis. Specially, aiming at the solution of the deepening video application on intelligent transportation and public security, the video structured description technique on android platform was utilized to execute the scene description task and implement the recognition task such as face recognition and vehicle recognition. The schema of video structured description on mobile computing architecture was figured out and the synthesis searching framework was introduced by utilized the semantic web search engine. As our experiments proved, the OpenCV for Android based video analysis and event detection is eligible for the present ARM-based hardware configuration. Furthermore, the offline training and online recognition mechanism can utilize enough the advantage of the mobile computing environment in the WLAN and can guarantee the real-time of the video structured description on Android platform. The extensive application in video portals shows the inspiring prospect of the proposed video analysis technology. The present deficiency of the light-weighting video structured description also was discussed.