Abstract
This paper proposed a shot boundary detection method based on color and texture features of frames. The shot boundary detection is employed in order to build a highly efficient method for video description. The research methods include feature extraction using color and texture feature, and shot boundary detection offering an adaptive measuring method. In this paper, the color and texture features are Traditional Color Histogram (TCH) and Histogram of Gradient Directions (HGD), respectively. The adaptive measuring method was proposed for the threshold of distance as the basis of identifying shot changes between frames. In order to validate the shot boundary detection method proposed herein, the video database provided by Carleton University website was used for the experiment. The experimental and comparison results of shot boundary detection were validated, and the experimental results were analyzed. This study carried out a series of comparisons and analyses for the above video database, and proved that the shot boundary detection method proposed could effectively detect the shot boundary.
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call