Design and Implementation of Hardware-SVM-based Shot Detection System

Structural analysis of video is an essential step to automatic video content analysis. In a general video structure, shot is the most basic unit, and it must be determined before any multimedia content-based retrieval applications. However, despite the rich research efforts in software shot detection system, there is a lack of hardware shot detection system. In this thesis, we designed a shot boundary detection system using hardware-based SupportVector-Machine(SVM). Our system can detect both cut shots and gradual shots. We optimized the feature extraction process based on global color histogram with a pipelined architecture to increase the overall speed. We solved the class imbalance problem in shot detection by applying random pseudo-sampling at the SVM training stage. The digital hardware SVM is designed using a fully-parallel pipelined architecture and is highly configurable on vector dimensions. Our data wordlength only needed five bits in fixed number, while the detection accuracy is competitive comparing with floating point software design. Our SVM classifier can run a maximum speed up to 251.62MHZ on Xilinx Virtex IV XC4VSX35 FPGA, and the video processing on our shot detection system can achieve 128 fps on PC matlab platform, it makes our system met the real-time constraints and detect shots in a continous video stream such as live-TV news or sports game.