Study On The Detection Method Of Medium Steel Plate Flatness Based On Cross Laser

Flatness is one of the principal indexes reflecting the quality of rolling products. It has a direct influence on yield of steel products. In recent years, with the raise of development plan of Chinese industry 4.0, intelligent detection technology has been more and more widely used in all industrial detection fields. Especially, machine vision has been widely applied in shape and morphology measurement of the medium and heavy plate, with the advantages in its non-contact type, high precision, and quick date processing and high real-time. With the rapidly development of modern manufacturing, there is a higher quality demand for flatness. For example, the height precision of extensibility fiber should be millimeter-scaled. Detection precision of superficial flatness for medium and heavy plate embodies development level of production equipment and rolling technology, but also it reflects detection level and control capability in system, and it also reflects a nation’s precision rolling development status and the overall strength. Therefore, conducting research about flatness measurement technology has significant meaning.This paper puts forward a non-contact plate flatness detection method based on cross laser and CMOS image sensor, taking the moving medium steel plate as the research object and realizing the fiber measurement by using a cross laser and CMOS image sensor. This thesis puts emphasis on the fundamental of laser triangulation, the calibration methods, and the precision contrast of point laser and cross laser, the influence of laser source voltage for the spot detection accuracy and the analysis and experimental research of flatness measuring system. In this paper, the following work has been done:(1) The current status of medium and heavy plate flatness is depicted, with the trend of future development in this field analytically studied. Furthermore, a new detection method is proposed according to the current status and requirements of medium and heavy plate detection.(2) A new platform is built for non-contact flatness detections, considering about the low precision of traditional laser spot center coordinate, a cross laser source is selected to improve the extraction accuracy.(3) A new calibration board is designed to calibrate the camera in coordinate with the camera calibrating principle. The relationship between the image motion and the shooting height is simulated based on the date measured as different heights, and according to the related calibration parameters, the interconversion between space and plan coordinate systems of image is realized.(4) A speed measurement approach based on magnetic grids sensor is designed, and match the processed speed dates with the height dates to calculate the flatness.(5) The Matlab software platform is utilized to study the algorithm of individual phases in the flatness detection of medium and heavy plate, including image pre-progressing, image segmentation, image coordinates extraction of feature points, image date fitting and so on. Finally the precision of detection is analyzed with samples and the inspection system achieved flatness detection accuracy to 1mm.(6) The research work at present stage was summarized. And the further study in the future was prospected.