摘要
基于Vivado HLS工具封装了包括FAST算法和AXI接口的FPGA硬件IP,并将该IP集成到工业摄像机的嵌入式系统中,工作摄像机中FAST算法的最终处理结果传输给视觉检测设备主机。利用相同图像的FAST角点检测算法对FPGA算法电路和OpenCV算法程序进行对比实验,完成相同功能的情况下前者的处理速度比后者快约60倍,该功能在工业摄像机中使数据带宽最多减小9%,显示出本文的工作可提高工业视觉检测设备的检测效率和性能。相关技术可应用于其他图像处理算法,也可推广到二维传感器阵列数据的处理。
Based on the Vivado HLS tool,the FPGA hardware IP is encapsulated including the FAST algorithm and AXI interface,and integrates the IP into the embedded system of the industrial camera.The final processing result of the FAST algorithm in the working camera is transmitted to the visual inspection equipment host.Using the FAST corner detection algorithm of the same image to compare the FPGA algorithm circuit and the OpenCV algorithm program,the processing speed of the former is about 60 times faster than the latter under the same function.This function realizes the data bandwidth in the industrial camera the reduction is up to 9%,showing that the work in this article can improve the detection efficiency and performance of industrial visual inspection equipment.The related technology can be applied to other image processing algorithms,and can also be extended to the processing of two-dimensional sensor array data.
作者
王柔
冯冲
张俊星
孙福明
Wang Rou;Feng Chong;Zhang Junxing;Sun Fuming(College of Information and Communication Engineering,Dalian Nationalities University,Dalian 116600,China)
出处
《单片机与嵌入式系统应用》
2021年第2期55-58,共4页
Microcontrollers & Embedded Systems
基金
国家自然基金面上项目(61976042)
辽宁省自然科学基金计划项目(2019 ZD 0173)
大连民族大学学科团队项目(0919110163)
横向项目(2019040037)
辽宁省科技厅博士启动基金(20180550560)。
