Research And Application Of Sorting Algorithm On Tripod Robot Based On Machine Vision

With the development of modern industry, the demand for the intelligent level of industrial robot is increasing. Industrial robot based on machine vision can sense its surroundings through the visual system and make corresponding decisions independently, which can greatly improve the flexibility of robots, and make them be widely used in the electronic, automobile, aerospace, food and pharmaceutical industries.One of a typical application of industrial robots based on machine vision is to identify different artifacts and to sort them out. In this paper, a visual system is designed for the tripod robot of B&R to recognize the shape, color, and location of nine different geometries. Finally, the sorting device of tripod robot will place the nine different geometries to required position with the optimal path that is optimized by the system.In the process of target recognition and location, a series of processing is done for the image, such as perspective transform, graying, smoothing, threshold segmentation, edge extraction and morphology processing. After the processing, the fairly complete contours of the geometries are obtained. Then, the color of the geometries are identified according to the RGB model of the color image, the shape and position of the geometries are identified according to Hough transform, corner detection and area detection.The optimization of sorting path is a Travelling Salesman Problem with constraints. In this paper, the traditional ant colony algorithm is improved by putting forward the concept of static and dynamic constraints. Then, the improved ant colony algorithm is applied to the optimization of sorting path with appropriate parameters that is determined according to theoretical guidance and simulation experiment. Finally, the effectiveness of the improved ant colony algorithm is verified though MATLAB.