FPGA优化的快速边缘检测方法研究

Research on Fast Edge Detection Method Based on FPGA

对于移动设备处理图像效率低、功耗高的情况,本文对比几种经典边缘检测算法后,设计了基于Sobel算子的高效边缘检测实现方法。系统主要工作是把经过MATLAB处理后的灰度图像通过串口通信发送给FPGA,经过Sobel算子计算,将计算完成后的图像数据通过VGA时序输出至VGA显示器显示,最终实现一种基于FPGA的快速且消耗资源较少的图像边缘检测方式。在设计算法实现模块时,以信号流程图来体现模块间的关联关系,同时用Modelsim仿真来验证所设计程序的时序逻辑是否正确。

For the situation of low efficiency and high power consumption in image processing by mobile devices,an efficient edge detection method based on Sobel operator is designed after comparing several classical edge detection algorithms.The main work of the system is to send the gray image processed by MATLAB to FPGA through serial communication.After the calculation of Sobel operator,the calculated image data is output to VGA display through VGA time sequence,and finally realize a fast and resource-consuming image edge detection mode based on FPGA.In the paper,when designing the algorithm implementation module,signal flow chart is used to reflect the correlation between modules,and Modelsim simulation is used to verify whether the timing logic of the designed program is correct.