一种融合PCNN和Watershed变换的图像分割方法

Image segmentation based on fusion of WST and PCNN

提出了一种融合Watershed变换(WST)和主成分神经网络(PCNN)的图像分割方法.该方法充分利用WST和PCNN在图像分割方面各自的优点,首先用PCNN对待分割的图像进行预处理,自动生成对象标记,然后在标记的指引下,进行WST.有了标记,WST就会以一种受控制的方式进行,这样过度分割的问题就会得到解决.实验表明,该方法能够快速、准确地实现图像分割.

This paper presents a novel approach for image segmentation based on the fusion of watershed transformation（WST） and the principal component neural network（PCNN）.With the help of the markers that have been located by PCNN automatically and previously;marker-controlled WST can avoid over-segmentation.Experiment shows that the proposed method can achieve good image segmentation result effectively and efficiently.