Research And Realization Of Portable Obstacle Avoidance System Based On Embedded Platform

With the rapid development of computer vision and embedded system technology,obstacle identification technology has been widely used in military,scientific research and other fields.At present based on recognition of visual image is used for scene is relatively fixed field,such as license plate recognition,face recognition,position detection,etc.,but complex scenarios used in the experiment and research and development stage,more difficult to applied in the actual life and promotion.1.The research status of obstacle avoidance equipment at home and abroad was studied,and the advantages and disadvantages of existing obstacle avoidance equipment were analyzed and compared..2.Sorted out the key technologies of binocular machine vision and image processing.The mathematical model and theory involved in binocular machine vision are derived in detail,and the principle of image feature extraction and classifier(SVM)is emphatically analyzed.3.from the Angle of hardware,software and system,this paper expounds the structure and function of the system,based on S5PV210 embedded hardware platform,the hardware part mainly includes: the A8 core board module,communication module,GPS module,3 g SMS transceiver module and speech module.Adopt Linux as the front-end operating system,and drive and program development under Linux.Finally,through the experiment to verify the operation effect of the entire obstacle avoidance system.From the test results,the accuracy of the system basically meets the expected requirements,the work is relatively stable,the main obstacles can be identified more accurately,and it has good real-time performance.