Research On Line Laser Scanning Technology Based On Machine Vision

Line laser scanning technology is one of the important means to obtain object information.It is an urgent technology in the field of measurement and detection to accurately grasp the shape,distance,height and other high-precision information of the measurement object in a short period of time,and not be affected by the material,color and gloss of the object.In this thesis,a line laser scanning technology based on machine vision is proposed,which combines the triangulation principle and image processing technology to achieve high-precision and stable measurement of objects of different materials.The specific research contents include:(1)To analyze the principle of laser ranging,we build a line laser vision measurement system with the principle of direct laser triangulation,to confirm the specific selection of the hardware part of the system,and then complete the construction of the actual system.Task of completing the calibration of the line laser vision measurement system.Describe the conversion relationship between the pixel coordinate system in the camera imaging model and the world coordinate system.The Zhang zhengyou calibration method was selected to calibrate the camera,calibrate the camera to obtain the camera’s parameters,and calibrate the light plane to obtain the light plane in the camera coordinate system.(2)Aiming at the problem that different material objects have different sensitivity to light,resulting in uneven image brightness,the proposal of an improved two-dimensional gamma luminance correction algorithm.Perform Robert sharpening and bilateral filtering on the image,convolve the filtered image and the sharpened image to obtain the illumination component,extract the illumination component and adjust the two according to its distribution characteristics.The parameters of the dimensional gamma function can finally achieve uneven brightness correction of images with different reflectances due to different materials.Compared with the effects obtained by other brightness correction algorithms,as well as histogram,standard deviation,information entropy,and average gradient data indicators,the effectiveness of the proposed algorithm is proved and better than other algorithms.(3)The proposal of an improved laser fringe center sub-pixel extraction algorithm,which realizes the high-precision extraction of object surface feature information.According to the characteristic that the gray value of the line laser irradiation area is higher than that of other regions in the same column,this thesis proposes a segmented optimal threshold segmentation algorithm.The grayscale image is divided longitudinally into n equal parts,and the optimal threshold T of each segment is calculated,which improves the segmentation accuracy of the line laser scanning position.Finally,according to the width of the line laser scanning position,the appropriate Gaussian variance is set for Steger algorithm extraction.Compared with the effect obtained by its centerline extraction algorithm,the centerline extracted by this algorithm is smoother and achieves better results.(4)Objects with different materials have been experimented,which verified the feasibility of the experimental device in this thesis.Combined with the parameter matrix obtained by calibration,the extracted pixel coordinates of the line laser centerline are converted into three-dimensional point cloud data.The experimental results show that the point cloud data of the three materials of mold 1 are basically consistent,which verifies the stability of the experimental device in this thesis.The experiment shows that the width error of the aluminum alloy groove is within0.07 mm and the height error is within 0.05 mm,which verifies that the experimental device has high measurement accuracy.Compared with other algorithms,the algorithm extracted in this thesis has higher accuracy and faster speed.