Path Planning Research Of Binocular Stereo Vision Robot Based On Cortex A9 Platform

With the development of computers and automation, the research of robot is deepened more and more, and the robots are increasingly intelligent. Whether a robot is intelligent depends largely on the performance of its motor processes. There are two essential directions in the field of robot movement, including the vision system and the path planning. The binocular stereo vision proposed by mimicking the human eyes are widely used in extracting 3D information form images and high-precision and short distance measure system. By using the situational awareness information provided by the binocular vision system, path planning makes it possible for robots to avoid obstacles independently and reach the destination.In this paper, due to the fact that the embedded platform has specific resistance; high reliability; and being easy to be cut down, it is chosen as the development platform of robots, making it an ideal platform for Robot development. The binocular vision and path planning are combined together and put into a more and more powerful general-purpose embedded platforms. Relying on the strong binocular vision algorithms of OpenCV, the identify of obstacles and distance measurement can be achieved by the robots. The artificial potential field, it adopt the relatively simple principle and the amount of computation is small enough. What’s more, there is no need to use other algorithms which require a large storage space to achieve Path Planning. The existing problems of the artificial potential field has also been solved, thus it suit better for the intelligent robot movement obstacle avoidance system with embedded as the core.In this paper, instead of using the previous ARM9 and ARM 11 as a development platform, the author select the more powerful Cortex-A9 platform to complete the work of binocular ranging. By transplanting and building the necessary embedded environment as well as conducting experiments and simulations, it can effectively complete the environment depth information acquisition and path planning task.