Safety Surveillance Algorithm Research Around A Vehicle Based On Human Tacking And Image Processing

Nowadays driving safety becomes a more and more serious issue. A widelyemployed method for driver safety assistance is to remind a driver with the informationabout lane cues and obstacles on a road, running cars beside and various transportationsigns. Therefore, it is necessary to develop some methods or systems. Safety smartsurveillance is more and more urgent needed in various social occasions especially indriving assistant system.Motion target detection and tracking is one of the most important subjects incomputer vision, which combines advanced technologies in image processing, patternrecognition, automatic control, artificial intelligence, computer and other relative fields.Our target is to detect obstacles in the surrounding of the vehicle to remind the driverof safe driving and increasing tracking stability and robustness. We mainly aim tostudy algorithm of detection and tracking a moving human body in video imagesequences captured by a fisheye camera. Detection of moving human in stationaryscene is discussed, and then target tracking based on correlation tracking technique aredetailed studied. Considering existing problems, effective improved algorithms arepresented, by which tracking stability and robustness are increased. We studiedmethods for human body detection, segmentation and tracking in smart videosurveillance monitoring around a vehicle deeply.