复杂图像的Kaniadakis熵阈值分割方法

Threshold segmentation method of complex image based on Kaniadakis entropy

针对工业无损检测及红外等图像,基于信息论中的Kaniadakis熵理论,提出一种用于复杂图像分割的阈值化方法。该方法能综合图像像素灰度分布的复杂信息实现图像的有效分割,且能通过调节熵参数k适应不同的分割任务。实验结果表明,提出方法在复杂图像的分割中获得的结果优于相比较的其他方法,且具有良好的实时性能,满足实践任务需求。

A threshold segmentation method based on Kaniadakis entropy theory was proposed for complex images. The complicated information of image pixel gray distribution can be used to achieve effective segmentation of the image, and the different segmentations are adapted by adjusting entropy parameter/c . The experimental results show that the proposed method is superior to other methods and has good real-time performance, which can meet practical needs.