基于小波域灰度拉伸的Ostu图像分割

New Ostu image segmentation based on intensity stretching on DWT fields

针对Ostu分割方法简单有效但不适用于没有明显双峰直方图图像分割的问题,提出了一种基于小波域灰度拉伸的Ostu图像分割方法,结合小波变换和最大类间方差分割的原理对图像进行灰度拉伸和边缘信息增强后再利用Ostu方法进行分割.实验结果表明,新的分割方法在没有明显增加计算复杂度的情况下增大了Ostu方法的普适度,保留了更丰富的细节成分,提高了分割精度.

Ostu image segmentation is most useful threshold segmentation technique,but it is powerless for those image,which hasn t distinct double peak in histogram.A new Ostu image segmentation method based on intensity stretching and edge enforcement on DWT fields is presented.First,stretching the intensity and enforcing the edge are done by means of the theory of wavelet transform and maximum variance between different classes,then image segmentation is done by Ostu method.The experience results show new method is universal for masses of image without much more complexity, and retain more affluent detail and develop segmentation precision.