A Detection Method using Eigenvalue Distribution of Wishart Matrices for Video Surveillance Systems

• Main Authors: Wei-ren Jhu, 朱偉仁
• Other Authors: Tsang-Yi Wang
• Format: Others
• Language: zh-TW
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/2d8tb7

碩士 === 國立中山大學 === 通訊工程研究所 === 103 === In recent years, video surveillance systems have been widely used in our lives; it makes our environment more comfortable and secure. But the video surveillance systems are usually a source to evidence for occurrence of events. However, such video surveillance systems can’t provide real time alerts automatically when event occurs. The purpose in this thesis is combining statistical signal processing with digital image information to monitor the environment whether foreign matters invade or not. If the system detects the foreign matters invade; it can alert and save the important images in time. We have considered the image matrix in the literature. It can be divided into several blocks and make processing and distribution of each pixel for inspiration by Gaussian distribution respectively. We adopt two programs by using eigenvalue distributions of Wishart matrices to implement this system. In the end, we will simulate the programs we proposed and compare photos to video images by using Matlab.