Research On 3D Measurement Method With Anti-reflection Light Interference Under GPU Acceleration

Coded structure light 3D measurement technology is widely used in industrial production and life applications in various fields.In this paper,based on the extensive understanding of the coded structure light 3D measurement method's domestic and foreign development,this paper uses the gray code structured of the 3D measurement method.For the case where there is interference with the reflected light on the surface of the object,as well as the requirement of measuring speed and measurement result integrity,a 3D measurement scheme with anti-reflection light interference under GPU acceleration is proposed.The hardware and software structure of the system is designed,the key technology is researched and analyzed.Experiments on the scene with reflected light interference show that the system has good robustness which can reduce the number of projection patterns,improve processing speed and validate that algorithm is feasible.My work is mainly composed of the following aspects:1.A projection compensation algorithm for reducing reflection of light interference in a 3D measurement system is proposed based on the fact that there is a wrong 3D measurement result due to global illumination interference in the actual 3D measurement of the structured light.By using the projection pattern of high spatial frequency,denoising the piece in the scene that is disturbed by reflected light.And then through the statistical analysis algorithm to denoise the point cloud data,feedback and get the mask pattern of the projection,and finally after projection compensation to get the scene of the 3D point cloud data.2.To improve the speed of 3D measurement,the use of NVIDIA Tesla C1060 graphics card to speed up the image processing speed.Based on the analysis of GPU hardware structure and software programming mode,four algorithms are used in the 3D measurement system of Gray code structure: the estimate global and direct component algorithm,the pixel classification algorithm,the Gray code conversion algorithm and the triangular method to calculate the three-dimensional information,and make the improvement and implementation based on the CUDA parallel architecture platform..3.Write image acquisition program,projection control program and synchronization control program,to achieve 3D measurement device.Completing the camera calibration.The global and the direct component light of the scene are analyzed.On this basis,the gray code coding pattern acquired by the camera is classified,and the exact coding value of each point of the scene is obtained.According to the decoding value and the corner of the coherence matrix calculation to complete the projector calibration,and then get all the internal parameters and external parameters of the system.4.Combining the 3D measurement algorithm with the reflected light interference and the GPU acceleration image algorithm,the 3D measurement method with anti-reflection light interference under GPU acceleration is realized.The results of the physical measurement of the object with the reflected light interference are obtained,and the result of 3D point cloud data is complete.Compared with the CPU reconstruction,the speed increase 430 times.Verify the applicability of the method.