The measurement method based on the block adaptive double thresholds of motion target detection

Be proposed in this paper, based on the panoramic stereo vision mobile robot for automatic detection and tracking of moving targets and positioning, research on moving target detection algorithm, is put forward based on the block adaptive double threshold of moving object detection algorithm. First of all, the extracted from video sequence frame difference. For separation of target and background, next second range value of the utilization of the result of the difference. Then, morphological filtering. Finally, using the connectivity analysis to fill the hole. Experiments show that this method increases the system reliability, in the autonomous positioning test shows great advantages.

I. Introduction

Catadioptric panoramic camera is composed of reflector and ordinary camera, space of the light into the camera after a mirror reflection imaging, generate ring panoramic images. Because of catadioptric panoramic vision system level viewing Angle is 360 0, can realize the close wide field of three-dimensional visual perception, is widely attention in recent years, popular research hotspot technology. Moving target tracking and location based on panoramic vision is a new target tracking technology.