Study On Industrial Robot Application Based On Visual Guidance

Intelligent manufacturing has become the development direction of manufacturing industry and its core is industrial robot.The production cost can be reduced and the output and quality of products can be improved by the application of industrial robot in industry.The perception and adaptability of the robot to the external environment are improved by the visual guidance system.It has great significance on intelligent manufacturing.Based on the background of sorting system,the application of visual guidance technology in industrial robot is researched.The following three aspects are researched in this paper:(1)The object recognition and location method based on contour matching is researched.First,the source image is segmented by the edge-based method.The contour features of object image and template image are extracted.Then the moment invariants and the contrast metric between object and template contours are calculated.The incorrect matches are removed by metric threshold.Finally,the center pixel coordinate is calculated by the standard moments of image.The rotation angle is calculated by included angle between two vectors which are composed of center coordinate and barycentric coordinate of object image and template image.(2)The object recognition and location method based on feature points matching is researched.First,the source image is segmented by the edge-based method.The feature points of object image and template image are detected.Then the Hamming distance between feature points is calculated to determine whether the feature points match.The incorrect matching results are removed by the threshold of the best feature points.Finally,the center pixel coordinate is calculated by two triangles which are composed of the center coordinate of the template and any two feature points.The rotation angle is calculated by included angle between two vectors which are composed of any two feature points of object image and template image.(3)A sorting system based on visual guidance is designed.The simple workpieces and chess pieces are sorted in the system.First,the source image is obtained from industrial digital camera and objects are recognized and located by visual processing program with C++.Then the object information is sent to the industrial robot by Ethernet communication,the data analysis and industrial robot control procedures are programed by MELFA-BASIC.Finally,the objects are automatically sorted by RV-13 F 6-DOF industrial robot and pneumatic nozzle.The object recognition and location experiments are carried out to verify the effectiveness of the object recognition method.The absolute errors between calculated coordinates and actual coordinates are employed to analyze the accuracy of the recognition and location results.The actual world coordinates are obtained by teaching device.The experimental results show that the errors between calculated and actual coordinates are within 0.6mm and objects can be grasped by industrial robot.The performance of the system is tested.All objects in the image can be arranged in the same direction and automatically sorted using the calculated rotation angle.