Research On Key Technology Of Workpiece Inspection Manipulator Control System

With the rapid development of industrial intelligent manufacturing,the automation system combining machine vision and manipulator has been applied more and more in the production and manufacturing industry,and has become a hot research topic at present.The traditional manipulator imitates the action function of the human arm,to replace the simple and repeated labor,according to the immobilized program to grasp the handling and other operations,but can not judge the merits and demerits of the production of workpiece,with strong limitations.With the development of vision field,detection technology has greatly improved the accuracy and efficiency of automatic control system,so it has a better research prospect for the control system of manipulator.In this context,this thesis studies the modern industrial field of manipulator control system,analysis of the research status at home and abroad,seize the direction of its development,analysis of the current manipulator control system in which there are shortcomings,the key technology in the related field of a comprehensive study.In this thesis,combined with image enhancement,stereo vision and other machine vision technology and machine learning field to study their application in the manipulator,developed a manipulator control system based on Raspberry Pi overall scheme.The manipulator control system in this thesis carries out filtering,sharpening,morphological denoising and other operations on the collected workpiece images to improve the image quality and reduce the error influence caused by the interference of external conditions such as illumination.In the experiment,the improved iterative maximum inter-class variance method is used to segment the image,and the target is segmented from the complex background area.The centroid calculation method of chain code is applied to locate the target,which greatly improves the efficiency of the manipulator to track the target location.Then,this thesis introduces the image defect feature description method,studies the machine learning method to identify and classify the defect image,analyzes the advantages and disadvantages of each algorithm model,and selects the appropriate algorithm to detect the defect.Then through the experiment verified that the manipulator can accurately identify and detect the target workpiece and determine the quality of the workpiece,verified the design of the workpiece detection manipulator control system based on machine vision feasibility,and the experiment may produce error causes are analyzed.Finally,the research results of this thesis are summarized and some problems that need to be solved in this system are analyzed.