Search

The skipping frame algorithm in TMN8 rate control only depending on buffer state regardless of image characteristic, may skip important frames with large motion and, as a result, video quality seriously reduced. An adaptive skipping frame algorithm is proposed for low-bit rate real-time video encoding. The occurrence of frame skipping is jointly dependent on the temporal and spatial contents of the video, and achieves a balanced spatial and temporal quality to enhance the overall perceptual quality. Experimental results show that the proposed algorithm can achieve the better subjective and objective video quality than TMN8s algorithm, without introducing any computation complexity.

Traffic conflict studies have been accepted because the data and the effectiveness of a treatment of safety problems can be collected and evaluated in a short period of time. Driving behavior is the reflection of the vehicles operating characteristic, and it has close relation with traffic safety. In the paper, the process of vehicle coming and going at intersection, and driving behavior when conflict occurring were analyzed by field observation and video record observation. Based on the above analysis, driving behavior was classified into five types. Because each driving behavior has influence of different degree on traffic safety, driving behavior needs to be weighted. The paper used expert consultation method, and obtained multiple expert scoring for each kind of driving behavior. Then, the more reasonable weights were given by statistical method determining weights. Comprehensive value of driving behavior during statistical time is gotten according to the survey data. Finally, the relation model between driving behavior and conflict was established.