Recent Trends in Machine Learning for Background Modeling and Detecting Moving Objects

Background modeling is often used in the context of moving objects detection from static cameras. Numerous methods have been developed over the recent years and the most used are the statistical ones. This paper describes the current state-of-art in background modeling methods for moving object detection. We also propose a method for background modeling based on texture features and self organizing map through Artificial Neural Networks. This approach can handle video scenes containing moving background, illumination variation, and also include into the background model shadow cast by moving objects.