The Study Of Some Key Techniques On Automatic Measurement And Process System Of Free-form Curved Surface

With the continuous increasing requirement, the parts and products that have the free-form curved surface have got more and more extensive application in modern society. At the same time, with the rapid development of the aviation, space flight, automobile and mould industry, the measurement and process of the free-form curved surface objects have got more and more important. The traditional measurement methods such as stereoscopic law, Moll's stripe law, grating cubic chart law and so on need expensive equipment and well-trained special technical staff, and they have long measurement cycle and low measurement efficiency. These shortcomings make the traditional contact measurement method hinder its wide application. The non- contact measurement system has paid more and more attention and got extensive application, which has advantages of high speed, no abrasion, and good continuity.The paper aims at realizing automatic online measurement of free-form surface with the system of processing for the background. The paper study how to constitute and implement an automatic measurement of thickness and form of planer or curved surface system combining non-contact photoelectric measuring technique and the corresponding robot system. The thickness measurement of the transparent curved surface object is investigated with the linear structured light method. This paper carries on the research of the theories and the experiment, main research contents are:(1) The non-contact measurement method based on the principle of linear structured light is studied according to the present conditions and development status of non-contact measurement. Laser scanning measurement using linear structured light is taken as emphases of research works. A mathematic models set up and a new algorithm of extracting the light strip center are developed.(2) The paper investigates the approach to camera measurement system calibration. An approach to hand-eye linear calibration based on project matrix M is developed, it improve the precision of the measurement system greatly.(3) The paper puts forward a system calibration scheme, and calibrates the measurement system by many experiments. The reason that caused measurement error is analyzed deeply. The paper makes use of the neural network to compensate linear structured light measurement system, so improve the measurement precision.