Research on Container Edge Detection Algorithm Based on FPGA

In view of the complexity of container image noise and many edge details, the traditional Canny algorithm is improved in preprocessing module and non-maximal suppression module. By writing verilog in vivado software, the edge detection of canny algorithm is realized, and the effect of the algorithm is verified by MATLAB software. The improved edge detection algorithm is compared with the traditional canny algorithm. The experimental results show that the improved canny algorithm can restore more edge details, take less resources and process faster than the traditional algorithm for container image.

I. Introduction

Containers in the process of transportation due to various factors may cause a certain deformation of the entire structure failure, which will have unimaginable consequences, seriously affecting the safety and reliability of transport. The existing maintenance method is that maintenance workers check the deformation of the box one by one, but because of the large number of containers, and some shapeshifters can not observe the deformation effectively, it is very important to use automatic detection technology to detect the deformation of the box. Edge detection has a large amount of data and is faster for real-time image processing. With the development of FPGA chip performance such as data transmission speed, integration and data parallel processing, FPGA-based edge detection system has received wide attention[1]~[3]. Compared with traditional image processing system based on PC or DSP, FPGA-based image processing system has the advantages of high flexibility, low cost, fast parallel operation speed, abundant logical resources, assembly line technology and high execution efficiency[4].