Research on the complete dimension measurement of truck compartment with binocular vision
In this paper,a binocular vision based measurement method is proposed to rapidly measure the full size of truck compartment.The binocular camera is calibrated according to the camera imaging model and calibration principle.The calibrated binocular camera is employed to collect the image of the truck car from its rear side.Then U2-Net the salient-object detection algorithm is introduced to extract the outline of the truck car image collected,and the edge of the truck car is extracted through the improved line segment detection algorithm based on image segmentation.The specific corner coordinates are calculated by using the line segments of the edge of the carriage,and the dimensions of the carriage are obtained from the corner spacing of the carriage after the three-dimensional coordinates are restored.Our results show the average error of carriage size detection is less than 4.17％,and the average error of volume detection is less than 4.1％,demonstrating its feasibility for actual measurement.
国家科技期刊平台
NSTL
万方数据
