A New Method for Image Fusion Based on Pulse Coupled Neural Networks with Adaptive Parameters

• 作 者: JiangXiangYng;XuXinZheng;NieRu;WangGuanYin;
• 机 构: 中国矿业大学
• 出 处: 中国计算机学会多值逻辑与模糊逻辑专业委员会第十四届学术会议
• 发表日期: 2014年期

　　Image fusion,which plays a dominant role in many fields such as computer vision,manufacturing and medical images,attracted the attentions of numerous researchers.For Pulse coupled neural network(PCNN)is derived from the synchronous neuronal burst phenomena in the cat visual cortex,it is very suitable for image processing.However,the fusion method based on conventional PCNN model is tedious and time consuming because the parameters are adjusted manually.In this paper,we proposed an image fusion method using the modified PCNN based on different evolution algorithm,which can adjust parameter of PCNN on the basis of source images automatically.In order to explain efficiency and validity of our proposed method,we take multi-focus images and medical images for example to explain further the advantages in comparison to other image fusion method.Experiments illustrate that the proposed method is better and more efficient than other methods,and the images fused by our method includes more information.Moreover,our method not only adapts multi-source images very well but also enhances the quality of the fused image.