基于模糊推理的边缘检测算法

Edge detection algorithm based on fuzzy inference

针对传统模糊推理边缘检测算法存在抗噪性能差、边缘为非单像素边缘等缺点,提出一种基于模糊推理的边缘检测新方法。首先根据全向小波变换获得4个方向的小波变换幅值,并将该幅值作为模糊推理系统输入;然后通过比较解模糊之后的值和自适应阈值得到二值边缘图像,再细化边缘得到最终边缘图像。实验结果表明:与传统微分算法和模糊推理算法相比,该算法对图像中噪声和伪边缘的抑制以及边缘提取的完整性都具有很好的效果。

For the shortcomings of traditional fuzzy inference edge detection algorithm, such as anti-noise performance is poor and the edge is not a single pixel edge, a new edge detection method based on fuzzy inference is proposed. Firstly, the wavelet transform amplitude in four directions is obtained based on the omni-directional wavelet transform and then the amplitude is input as the fuzzy inference system. Then, by comparing the value after defuzzification with the adaptive threshold, the binary edge image is obtained and then the edge is thinned to get a final edge image. The experimental results show that the algorithm has a good effect on the suppression of noise and pseudo edge in the image and the integrity of the edge extraction compared with the traditional differential algorithm and the fuzzy inference algorithm.