Motion Detection in General Backgrounds

摘要: Once the basic case of motion detection problem has been studied and solved issues referred to use different imaging sensors, adaptation environments, speed, shape changes targets, some uncontrolled dynamic factors (e.g. gradual/sudden illumination changes), we have focused on when real scenes are considered. Therefore, in this chapter, goal is design a perfect segmentation technique based environments without any constraint about environment targets be identified. With that aim, such as, for instance, presence vacillating background elements or distinction between objects terms motionless situations, solved. Thus, brief review previous work carried out order introduce our approach. Then, deeper analysis problems as well proposed solutions explained. Finally, experimental results, from qualitative quantitative points view, presented discussed. As it will demonstrated, compared with classical techniques, algorithm faster, more robust, sensor-independent.