Research On Mechanical Measure And Recognition System Based On Machine Vision

Along with the advance of science and technology, mechanical parts type of recognition and measurement of size also gradually towards informationization, the direction of high speed and high precision, the general detection means no longer meet the requirement of modern manufacturing, machinery manufacturing industry in the realization of product, the problem of automatic and intelligent production gradually looking for a new technology for manufacturing of mechanical parts for automatic detection, automatic assembly. Machine vision technology is the current implementation of mechanical parts type recognition and measurement technology research. Machine vision technique has high accuracy, high efficiency, not the advantage of contact with the mechanical parts, mechanical parts based on machine vision recognition and measurement technology gradually shows great potential in terms of precision measurement, is one of the focus on the current manufacturing technology.Based on the recognition and measurement of mechanical parts as the research object, through to conduct the thorough research to the machine vision technology research status at home and abroad, to find a suitable class automotive spare parts recognition and measurement of machine vision technology and method, by combining Matlab tools and Lab VIEW development tool for identification and measurement system research and its experimental verification. Mainly studied the following aspects of content:(1) for further study of machine vision recognition of mechanical parts and measurement technology at home and abroad, analyzes the application of machinerecognition technology at home and abroad and the development trend in the future. In this paper, the basic principle of image acquisition and recognition of mechanical parts is expounded, and the application status of image acquisition card and the important role of mechanical parts based on machine vision recognition and measurement are introduced. The basic method of image gray level transformation is studied, and the processing technique of mechanical parts and the edge processing method are analyzed.(2) for mechanical parts of basic characteristics, combined with machine vision recognition and measurement technology of machinery parts recognition and measurement of the overall scheme, the detailed design of the system hardware structure, determine the optical lighting system solutions and image acquisition system and a pneumatic control device. The system software development platform were introduced, the detailed design of the system software to achieve the overall scheme, implementation of the system is designed based on the machine vision machinery parts recognition and measurement functions, including parts of image acquisition, part image preprocessing, parts recognition and measuring and result output control.(3) for mechanical parts based on image recognition and machine vision, first carries on the analysis of mechanical parts image recognition process, determine the identification scheme; discusses the part image recognition algorithm, using the part recognition algorithm based on gray classification of mechanical parts and the delineation of matching area is studied; combined with gray algorithm is the complex problem of mechanical parts inside and outside diameter to achieve two times at the same time, area classification delineated matching method is proposed based on improving efficiency of regional matching; according to the actual operation in the process of using existing parts position uncertainty problem, need to image any angle of rotation, to ensure that the rotating parts can image with thetemplate matching method by rotating the template image in the recognition process have different direction of template matching, the most After describing the method of image recognition and parts making image template algorithm.(4) the machine vision measurement system calibration method and measuring test, respectively, of camera, pixel equivalent calibration, according to corresponding relationship between calibration to determine the image plane coordinate system and the world coordinate system, combined with image plane coordinate system and computer image coordinates corresponding relationship between pixel coordinates of parts in computer image coordinate system and in the world coordinate system in physical coordinates of the corresponding relation, which parts of the pixel coordinates obtained the corresponding physical coordinates, finally after conversion that parts of the actual physical dimensions. Describes the calibration method, the comparison of experimental results before and after the calibration graph is given by the calibrated system of parts; the size measurement, and the measurement results are analyzed.