实时图像自适应阈值边缘检测FPGA实现

Adaptive Threshold Edge Detection of Real-Time Image Based on FPGA

传统的Sobel边缘检测算子对噪声变化抑制能力较差,并且其使用固定阈值导致适应性差。该方法利用中值滤波思想来确定图像的每一像素点局部边缘阈值来实现自适应阈值Sobel边缘检测,并在FPGA平台实现。实验结果表明,该算法检测出的实时图像边缘更加精细,并且能很好地抑制背景噪声。

The traditional Sobel operator is sensitive to noise and lack of adaptability for its fixed threshold. This paper calcu- lated local threshold of each pixel by using median filtering to meet the demand of self-adaptive computation, then implemented it on FPGA. The results demonstrate that the proposed method has better performance, of which the image edge is clearer and the background noise is filtered as welt.