基于FPGA的图像处理实验平台设计  

Design of Image Processing Experimental Platform Based on FPGA

作　　者：林科业 黄仕凰 LIN Keye;HUANG Shihuang(School of Physics and Electronic Engineering,Hanshan Normal University,Chaozhou Guangdong 521041,China)

机构地区：[1]韩山师范学院物理与电子工程学院,广东潮州521041

出　　处：《微处理机》2023年第6期23-26,共4页Microprocessors

基　　金：广东省教育厅创新强校项目“基于FPGA数字图像处理技术的瓦楞纸产品质量检测”(2019KQNCX081)。

摘　　要：为克服PC软件平台体积较大、应用不便等缺陷,设计一种基于FPGA的图像处理实验平台,尝试将图像处理技术移植到嵌入式开发平台上。采用模块化的形式,设计OV5640、SDRAM、VGA显示等外设的驱动模块。结合FPGA自身的结构特点,设计YCbCr、均值滤波、sobel边缘检测、图像二值化、图像膨胀和腐蚀等图像预处理算法。将预先设计好的各个驱动和图像预处理模块综合成起来,在FPGA硬件平台上对图像处理进行实验演示。该实验平台充分利用FPGA的数据并行处理能力,在图像处理中同时运行多个算法,大大提高处理效率。In order to overcome the shortcomings of PC software platform,such as large volume and inconvenient application,an image processing experimental platform based on FPGA is designed,in which the image processing technology is tried to be transplanted to the embedded development platform.The driver modules of OV5640,SDRAM,VGA display and other peripherals are designed in modular form.Combined with the structural characteristics of FPGA,image preprocessing algorithms such as YCbCr,mean filtering,Sobel edge detection,image binarization,image expansion and erosion are designed.The pre-designed drivers and image preprocessing modules are integrated,and the image processing is demon-strated experimentally on the FPGA hardware platform.The experimental platform makes full use of the data parallel processing ability of FPGA,and runs multiple algorithms simultaneously in image processing,which greatly improves the processing efficiency.

关 键 词：现场可编程逻辑门阵列 图像处理 OV5640传感器 SOBEL算法 灰度转化 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]