Partition Fusion Approach of Image Fusion Using Modified Pulse Coupled Neural Network for Differently Focused Images

DOI: 10.5176/978-981-08-7656-2ATAI2010-42

Authors: Dheeraj Agrawal and Jyoti Singhai

Abstract:

The paper introduces an approach for fusion ofdifferently focused images for improved human perceptionand machine recognition. The fusion algorithm uses apartition fusion approach where source images arePartioned into blocks and based on their clarity theseblocks are selected for composite image. The quality ofcomposite image thus depends upon the clarity measureand decision making system to select a block from givenmulti focus images. In the proposed algorithm the clarityof the image blocks is decided by EOL. The selection ofblocks is performed by the Modified PCNN based on theclarity of the image blocks. To improve the performance ofPCNN some modification are used in feeding and linkingfiled in PCNN. The performance of the proposed fusionalgorithm has been tested for RMSE, Mutual Informationand standard deviation. Considerable improvements areobserved in the results as compared to existing method ofimage fusion.