Method for Automatically Ignoring Cast Self Shadows to Increase the Effectiveness of Video Analytics Based Surveillance Systems

摘要

One of the challenges in video surveillance system is the detection of unwanted intruders in secure zones. These zones may be in large facilities such as airports, train stations, port facilities or warehouses. Because many of these locations are outdoors, the presence of shadows is detected on images. In order to effectively eliminate these shadows which will therefore result in a usable figure, these shadows must be eliminated. This is accomplished by forming a contour box around the image and then reducing the shadows based on preset parameters in the software.