A multi-line laser scanning system for 3D topography measurement is proposed. This method not only has the advantages of high precision of laser scanning technology, but also has high reconstruction efficiency. In this paper, speckle reconstruction technique, multi-line laser technique and Biocular reconstruction technique are used to construct a 3D reconstruction system, and test equipment is built, and the problems existing in the system establishment process are actually studied. In order to solve the problem of mismatching in binocular multi-line laser matching, a method to sort out the correspondence of multiple laser lines in binocular images based on speckle matching results is proposed. In order to optimize the multi-line laser matching effect, a speckle matching network based on deep learning is proposed, which integrates the grayscale images of the left and right cameras as supplementary information, and takes the speckle image and grayscale image as the input of the network model to obtain more accurate and edge-complete matching results. Finally, the matching results of the multi-line laser and the camera calibration parameters were used to reconstruct the object point cloud. Experimental results show that the proposed speckle matching method can make binocular multiline laser point cloud reconstruction more robust and stable than the traditional method, and through the accuracy analysis of the system, it is proved that the average measurement accuracy of the proposed method can reach 0.05 mm.
Read full abstract