Gradual shot boundary detection using SVM optimization technique and HOG feature

Abstract

Rapid growth in the field of data capturing and storage along with the increase in the availability of multimedia content in the web, has resulted in many large personal and public digital video databases. The Shot Boundary Detection (SBD) is the very first step in the research arear of video application. The objective of the SBD is to address the issues in Video Content Analysis (VCA), namely, Video Segmentation. From the analysis it is identified it is difficult for the human being to analyze the contents manually. Therefore, many methods is needed to automate the process of videos. Gradual Transition detection is based on the Support Vector Machine (SVM). To improve the result of this shot boundary detection, SVM with Histogram of Gradient (HOG) is used. HOG feature is calculated for further shot detection. Finally gradual shots are detected based on the above methods. Here the performance is evaluated by the parameters called precision and recall value.