实时图像质心检测算法设计及其FPGA实现  

Design of real-time image centroid detection algorithm and its FPGA implementation

作　　者：张驾祥 谭会生[1] ZHANG Jiaxiang;TAN Huisheng(College of Railway Transportation,Hunan University of Technology,Zhuzhou,Hunan 412000,China)

机构地区：[1]湖南工业大学轨道交通学院,湖南株洲412000

出　　处：《湖南城市学院学报（自然科学版）》2024年第3期63-67,共5页Journal of Hunan City University:Natural Science

基　　金：湖南省学位与研究生教学改革研究项目(2022JGYB183)。

摘　　要：为了克服通用处理器实现图像质心检测算法时处理速度慢,难以应用于高性能实时计算机视觉和图像处理系统的缺点,采用并行处理、流水线等设计优化技术,并通过现场可编程门阵列(field-programmable gate array,FPGA)设计了一种实时图像质心检测算法。首先,分析了实时图像质心检测算法,并在MATLAB上对其进行仿真验证;然后,使用处理器系统(processing system,PS)和可编程逻辑(programmable logic,PL)相结合的方式,在ZYNQ SOC FPGA芯片上设计、实现了实时图像质心检测算法;最后,进行了实时图像质心检测的硬件验证。试验结果表明,当采集图像分辨率为1024×768、帧率为60时,该设计能够正确实时地计算并标定图像的质心。In order to overcome the disadvantages of low processing speed and difficulty in applying to high-performance real-time computer vision and image processing system,a real-time image centroid detection algorithm is designed and implemented by FPGA using parallel processing and pipeline design optimization techniques.Firstly,the real-time image centroid detection algorithm is analyzed and verified by MATLAB simulation.Then,a real-time image centroid detection algorithm is designed and implemented on ZYNQ SOC FPGA chip by combining the processor system(PS)and programmable logic(PL).Finally,the hardware verification of real-time image centroid detection is carried out.The experimental results show that when the resolution of the acquired image is 1024×768 and the frame rate is 60,the design can correctly calculate and calibrate the centroid of the image in real time.

关 键 词：图像质心检测 图像腐蚀膨胀 图像边缘检测 HU不变矩 PS+PL结构 FPGA 

分 类 号：TN402[电子电信—微电子学与固体电子学]