Study On The Environment Perception System Based On Vision Of Robot

Over 70 percent of human's sense of environment is gained by vision, which is also considered as the main method for robots. It is the precondition for them to be independent and the core of vision is the recognition of the objects. By setting autonomous robot as the object of study, setting the perception of environment by vision as the project of study, the target of this article is to establish a robot system which has the ability to detect the known dynamic environment all by itself. Our main study is the methods of color-based object recognition and shape-based object recognition.Through the discussion of the distill method of the characteristic of the color threshold of the aim and the problem of color matching and recognition, this article presents an algorithm for the color-based recognition of multi-objects at real time. It improves the quality of real time for the vision system and makes the system more robust to a certain extent.Through the study of getting figures of object, this article presents a fast-algorithm for the recognition of simple shape based on the improved Hough transform. It is proved by experiments that those algorithms can recognize objects with simple shape and single color in complex background, and it can meet the requirement of the quality of real time for robots.This article describes a perception system of autonomous robot and applies it on the automatic transportation system for the storeroom of aviation equipments, and applies the algorithm above and the algorithm of recognition by color code and the algorithm of localization with single camera on this system, it realizes the visual servoing for robot and reaches a good effect.