Research On The Method Of Laser Target Tracking Based On Machine Vision

In this paper,aiming at the problem of coal mining machine in the process of mechanization of coal mining,the method of laser target tracking based on,machine vision is studied.Image detection and target tracking algorithms will be researched,relevant experiment platform design,complete algorithm validation and analysis of positioning error and the reason,for further research of coal winning machine laser positioning based on machine vision to provide theoretical basis and technical support.In terms of moving target detection and tracking,first of all,the common method to target detection and tracking are analyzed,compared to frame differential method,background difference method,optical flow method detecting method,the advantages and disadvantages of five kinds of target tracking method,suitable for frame differential method in this paper,the detection and target tracking method is determined,At the same time,in view of the traditional three frame difference method to obtain the target features appear lack of redundancy,and the boundary is not clear phenomenon,the traditional three frame difference method and the way the edge detection was improved,and the fusion of background modeling based on the average method,background difference method,makes the detection of target information more accurate;Through the four sets of different movement targets,the improved algorithm is simulated.The experimental results show that the improved algorithm can achieve a more complete detection of motion target characteristics.In the aspect of laser target detection and tracking system,the overall design scheme is developed,and the structure composition,technical index and realization method of specific functions are studied;for the captured image capture lens,the internal parameters are calibrated through the camera calibration;considering the precision of the hardware,the camera search system and the laser target distance,designing a new laser target structure,avoiding the fact that the hardware and measuring distance is causing the laser beam to extract the wrong features.Finally,according to improve the target detection and tracking method and hardware experiment platform was set up,including the man-machine interface design and the selection of key hardware,drawn up in accordance with the experiment scheme,laser target detection and tracking system function of the basic implementation.