针对简单线性迭代聚类(SLIC)方法对图像边缘细节处理效果不佳的问题,提出一种参数可控、改进的简化脉冲耦合神经网络模型(PC-MSPCNN)与SLIC结合的彩色图像分割方法。该方法首先改进MSPCNN模型的加权矩阵和连接系数,并增设辅助参数,以提高分割准确度。随后将彩色图像输入至PC-MSPCNN模型中,依据改进模型中输出Y值的分布划分物体的边缘,使分割结果更好地贴合物体的边缘,利用所提出的相似性准则合并散布的碎片,减小后续处理的复杂度;其次,在SLIC度量相似距离的基础上引入PC-MSPCNN中RGB三个通道的内部活动项U值,完成对图像剩余部分的加权融合聚类,改进聚类效果。实验结果表明,本文方法能更精确地贴合图像中物体的边界,大幅减少碎片,有效提高图像的边缘贴合度。

The simple linear iterative clustering (SLIC) method does not perform well in edge detail processing in image segmentation. Thus, a modified color image segmentation algorithm is proposed combining a “parameter-controlled modified simplified” pulse coupled neural network (PC-MSPCNN) and SLIC. The proposed algorithm works in two steps. First, the weighted matrix and connection coefficient of the MSPCNN model are improved, and the auxiliary parameters are added to improve the accuracy of the segmentation. Then, the color image is input into the PC-MSPCNN model, and the edge of the object is divided according to the distribution of the output Y value in the improved model so that the segmentation results appropriately fit the edge of the object, and the proposed similarity criterion is used to merge the scattered fragments to reduce the complexity of subsequent processing. Second, based on the measurement similarity of the SLIC, the internal activity term U values of the three RGB channels in the PC-MSPCNN are introduced to achieve weighted fusion clustering for the remaining parts of the image to improve clustering. Experimental results show that the proposed algorithm can accurately fit the edge of an object, considerably reduce the number of pieces, and effectively improve adherence of the image edge.