Geometric Parameters Measurement Of Typical Hot Forging Of Agricultural Machinery Based On Computer Vision

The working environment of agricultural machinery, especially the field machinery, is mostly bad, so the mechanical performance of its working parts must be better, such as disc blade, plough, cutting knife, etc. The forming of these parts often need to forging. In the forging process, the wearing, the compressor deformation, the mold deformation, the partial load and the subjectivity of operators and other reasons lead to the inconsistence between the size of finished forged product and the size of the designed mold, so it is necessary to measure and control their size during the forging process.The plough body is the important foundation of the cultivating machine, size measurement is the key technology in the manufacturing process. At present, the size measurement of the plough body is mostly based on the artificial measurement with mechanical measuring instruments, the measurement error is large, waste time and energy and the working condition is poor. Computer stereo vision has the advantages of three dimensional, precision, real time, convenience, high efficiency, non-contact, automation, rich information, easy to analyze and so on. In this paper, in order to solve the difficult problem of size measurement of forging, research status at home and abroad is analyzed, and the machine vision measurement method is putted forward. Main contents are as follows:(1) The method of object extraction and edge detection of hot forging image is studied. Aiming at the intense light interference and high-temperature radiation of forgings, image features in various color space is studied, and the target extraction scheme is proposed based on the color space conversion and dynamic threshold segmentation; For the influence of oxide skin, ablation, burn and carbonization of hot forging, according to the shortcomings of the traditional edge detection operators of a fixed structure element is analysed, the influence of the shape, the size, the direction, the weight of morphological structure elements on edge extraction of hot forging is studied, the multiple, multi-scale morphological edge detection operator is constructed. The object extraction and edge detection of hot forging are realized.(2) The measurement method of the plough body contour based on binocular stereo vision is studied. Considering that the plough body contour has great effect on plow surface design, working performance and quality inspection at the scene of the forging, the plough body contour measuring based on binocular vision is introduced. Computer binocular stereo vision calibration method and 3D detection algorithm is studied, and secondary correction of 3D reconstruction method is proposed, the corresponding program is developed. With the realization of the plough body contour edge extraction, edge reconstruction, edge fitting and mathematical description, the difficult problem of the measurement of plough body contour, especially of hot forging plow body contour is solved.(3) The measurement of plough template curve based on binocular stereo vision and structured light is studied. Because the plow template curve is of great important for plow surface design, manufacturing, and quality inspection, the template curve measurement method of the binocular vision and structure light is presented. The structured light stripe center extraction method and the 3D reconstruction algorithm is studied, and the corresponding program is developed. With the realization of the light center extraction, fitting and mathematical description of plough body template curves, the problems of the measurement of the plough body template curve is solved, especially for hot forging plow body. It provides a new way for the mapping, description, design, manufacture and inspection of the plough. And this idea will be extended to inspect other characteristic curve and performance curve, including digging curve, soil crushing curve and contour etc.(4) The method of measuring the plow body surface based on structured light three-dimensional vision is studied. Based on structured light 3D measurement model and calibration theory, the "three steps" of the line structured light calibration method based on active vision and planar target is proposed; the measurement system with one dimensional translation based on the line structured light vision is constructed.3D reconstruction measurement algorithm is studied reconstruction program is compiled. The system has the advantages of simple structure, high precision, low cost, anti-interference, realizes the plough body surface 3D reconstruction and fitting, solves the difficult problem of the measurement of plow body surface especially of hot forging. It provides a new way for the mapping, description, design, manufacture and inspection of the plough body surface.(5) Based on computer vision, the method of the geometric tolerance measurement of the hot forging of agricultural machinery is studied. Geometric tolerance including parallelism, roundness, verticality, coaxial, and so on. It is the important content of the work piece quality, and it is difficult to measure for hot forging. In this paper, with examples of agricultural forgings, a right angle connecting plate, an annular shield, the method and algorithm of roundness, verticality measurement is studied, the corresponding program is developed.(6) Finally with hot plough shin’s geometry size and hot circular plate’s roundness tolerance as test objects, at the scene of the forging based on computer vision, the feasibility, the efficiency and the precision, working reliability and stability of the measurement system are tested.