工业过程噪声干扰下图像关心区域分割算法研究

Research on Image Care Area Segmentation Algorithm Under Industrial Process Noise

针对单纯的基于像素灰度值的图像分割方法难以精确分割工业过程有噪声干扰的情况,提出了基于自适应网格搜索的Gabor小波纹理粗糙度方法,用于对基于像素灰度值的FCM聚类结果进行去模糊化,从而分割出关心图像。将所提算法行了大量实验研究,实验结果表明,该算法能够很好地将所关心的图像从工业噪声中分割出来。

Image segmentation methods simply based on pixel gray values are difficult to accurately segment industrial process images disturbed by industrial noise. To solve this problem,a Gabor wavelet texture roughness method based on selfadaptive web search is proposed to remove fuzziness of FCM clustering results based on pixel gray values to separate required images. The algorithm has been included in large simulation studies,and simulation results demonstrate that the algorithm can do very well in separating required images from industrial noise.