基于多态蚁群算法的回转窑图像分割方法

Segmentation of kiln image based on polymorphic ant colony algorithm

为了解决回转窑图像背景噪声干扰较大且背景灰度变化较小,采用经典的算法不能有效分割出窑内黑把子区域的情况,提出了一种基于多态蚁群算法和小波变换相结合的图像分割方法。利用小波将图像变换,对低频系数进行分块,采取新的阈值选取方法并对每块进行分割,从而改善了黑把子区域的分割效果。实验结果表明,该算法提高了分割精度,缩短了程序运行的时间。

To solve the problem that some classic algorithms are not too effective in the segmentation of kiln image when noise exists and background image gray changes smaller, a new image segmentation method is proposed, which combines the poly morphic ant colony algorithm with wavelet transform. The lowest frequency coefficients divided into several blocks are extracted using wavelet transform, with different image areas, a new threshold method is applied. As a result, the segmentation result of kiln image is improved. Experimental results prove that the proposed method can enhance segmentation accuracy and reduce running time.