Research On Container Edge Detection Algorithm Based On FPGA

As an important tool in commercial transportation,the safety of the container is very important.Ensuring the integrity of the container is an important means to ensure the safety of the goods in the container.However,due to the interference of some external factors,defects such as deformation and damage occurred in the container,which seriously threatened the safety of the contents in the container.The traditional container inspection method is manual inspection,but this inspection method is inefficient,poor in effect,and easily affected by the staff themselves.In order to solve the above problems,this paper uses the characteristics of high timeliness and low power consumption of FPGA to study the classic filtering and edge detection algorithms,and proposes an improved Canny edge detection algorithm based on the FPGA platform,so as to realize the detection of container images.The main work of this paper is as follows:(1)Aiming at the problem that the traditional median filter smoothes the edge information of the container,resulting in the loss of high-frequency information and blurred imaging,a median-Gaussian filter algorithm based on FPGA is proposed.The algorithm weights and averages the pixels through Gaussian filtering,reduces noise and suppresses smoothing effects,and restores the clarity of the image.In this paper,Peak Signal to Noise Ratio(PSNR)and Structural Similarity Index(SSIM)are used as objective evaluation indicators.The experimental results show that the medianGaussian filter PSNR index is 31.2012 d B,and the SSIM index is 0.9605.Compared with the traditional median filter,they are respectively improved by 9.25% and 6.44%.(2)Aiming at the fact that when the traditional Canny algorithm performs edge point localization and edge refinement of the image,the region division in the gradient direction is not detailed enough,resulting in the problem of edge loss in the output container image,a six-direction Canny algorithm based on FPGA is proposed.Based on the traditional Canny algorithm,two more regions are divided into a total of 6gradient directions to enhance the refinement effect of the edge.Experimental results show that the improved six-way Canny algorithm detects more container edge information,and the edge is more detailed and complete.(3)Xc7z010clg400-1 of XILINX ZYNQ series is used as the main control chip.The proposed median-Gaussian filter,six-direction non-maximum suppression and other modules are transplanted to FPGA and verified by simulation.The edge detection effect of the improved algorithm and other mainstream algorithms is compared,and the ratio of 4-8 connected domain C/A and C/B,and the edge continuity ASL index are used as objective evaluation indicators.The experimental results show that the C/A and C/B values of the improved algorithm are the smallest,which are 0.0218 and 0.2795,respectively,and the ASL index of edge continuity is the maximum value of 1.Comparing the running time of the improved algorithm on the FPGA and Matlab platforms,the running time of the FPGA hardware platform is 0.56 ms,which is91.76 ms faster than that of the Matlab software platform.Compared with the resource consumption on FPGA,the improved algorithm reduces the hardware resource consumption of LUT by 3.94%,2.26% of FF,and 5.84% of BRAM compared with the traditional Canny algorithm.