Research On Visual Inspection Method Of Airline Baggage Based On Laser Coding

At present,the Self Check-in with Baggage Systems can basically complete the baggage suitability testing,but because of its use of LiDAR propulsion scanning scheme can only get the sparse point cloud of baggage,can not accurately and reliably restore three-dimensional information on the baggage surface,which makes the baggage suitability detection there is a possible increase in miscalculation,and the dense point cloud will completely restore the three-dimensional information on the surface of the baggage,which will significantly improve the accuracy of baggage suitability detection,so it is important to study the 3D sensor and the baggage suitability algorithm based on the dense point cloud.The main work and contributions of the paper include:Firstly,the dynamic texture enhancement method of active laser encoding projection is studied.According to the significant texture feature requirements of stereo matching,the pixel encoding rules of projection image are designed.The active projection of weak texture airline baggage is carried out by using the encoding image to solve the difficult problem that the 3D reconstruction of the texture characteristics are not obvious or pure color airline baggage.Secondly,the 3D reconstruction method of the non-polar line constraint of binocular convergent mode is studied.According to the principle of the polar geometry,the optimal corresponding point search along the non-parallel pole line is derived,and the true three-dimensional coordinates of the object point are inversely solved by the coordinate mapping formula,thereby realizing the three-dimensional reconstruction of the measured object.Thirdly,the algorithm of seed region growth and gravitational segmentation is studied to determine the number of baggage.The effectiveness of the algorithm is verified by practical experiments,which can significantly improve the accuracy of baggage transportability detection.Finally,using the image texture enhancement method of laser projection coding and the3 D reconstruction method of non-polar constraints in binocular convergent mode,an industrial-grade embedded system with integrated laser projection control,image synchronous acquisition and processing,stereo matching and three-dimensional reconfiguration is designed and implemented,and the experiment is carried out in the self-check in with baggage system,which achieves satisfactory results.