三维连接系数矩阵的脉冲耦合神经网络彩色图像分割

Color Image Segmentation Based on Pulse Coupled Neural Network with 3-D Weighting Matrix

提出三维连接系数矩阵的脉冲耦合神经网络(3D-PCNN)模型,将二维连接系数矩阵扩展成三维,值取空间欧氏距离的倒数,提出指数上升的动态阈值。利用神经元脉冲同步发放特性和自动波特性,直接分割彩色图像。结果表明,3D-PCNN算法与其他分割算法相比,运行时间减少了25%以上;该算法能够将不同区域信息以多层次彩色显示,改变RGB分量输入顺序时,同样可以分辨出更多的图像细节信息,分割精度高。

A pulse coupled neural network with 3-D weighting matrix is presented,two-dimensional weighting matrix is expanded to three-dimensional weighting matrix in PCNN model,values are adopted as reciprocal of neural to neural Euclidean distance,iterative threshold is modified as index increased dynamic threshold. The autowave spreading characters which generated by similar neurons firing synchronously is used for color image segmentation directly. Experiments are carried out based on standard color image,comparing with other 3D segmentation method. Experimental results show that more particulars of color image are preserved by color image segmentation based on the three-dimensional weighting matrix,running time is reduced by more than 25%,different pixels with same color are segmented out and displayed with multi-gradation without improving computational consuming. When the input sequence of RGB component is changed,also segmented more details,have high segmentation accuracy.