An Improved PCNN Image Segmentation Algorithm
PCNN is a sort of neural network that simulates the synchronous oscillation of the pulse clusters produced by the animals visual cortex nerve cell. Because there are a lot of parameters in the PCNN model and many experiments are required to confirm the optimal values, they limit the application of PCNN in the image segmentation. The improved algorithm in the text makes use of the characters of the PCNN to achieve the automatic confirmation of the parameter value during the iteration calculation. The improved algorithm also strengthens the abilities of the border extraction of the PCNN by increasing the threshold of border points.
Max Entropy Max Variance between clusters (OTSU) PCNN Image Segmentation
XIA Hui MU Xihui MA Zhenshu WANG Hao LAN Jian
Ordnance Engineering College, Shijiazhuang , 050005
国际会议
北京
英文
2007-08-05（万方平台首次上网日期，不代表论文的发表时间）