Detecting motion from noisy scenes using Genetic Programming

Abstract

A machine learning approach is presented in this study to automatically construct motion detection programs. These programs are generated by genetic programming (GP), an evolutionary algorithm. They detect motion of interest from noisy data when there is no prior knowledge of the noise. Programs can also be trained with noisy data to handle noise of higher levels. Furthermore, these auto-generated programs can handle unseen variations in the scene such as different weather conditions and even camera movements.