A New Method Based On Evidence Theory And Its Application In Image Fusion

Image fusion has been an important and useful technique for image analysis and computer vision. With the use of multi-sensor, the fused image contains a more complete and accurate description of the scene than any of the individual source images. As a result of this processing, the fused image is more useful for human and machine perception or further image processing tasks such as object recognition and feature extraction. In recent years, it gets more and more attentions.This paper present a new method of image fusion based on evidence theory and fuzzy measure. Evidence theory is a very popular data fusion approach, but it has some difficulties in image fusion applications. One is that it's difficult to obtain Basic Probability Assignment of evidence theory, and the acquisition of Basic Probability Assignment is important to data fusion, it can affect the fuse performance directly. The other question is that when evidence increase, the numbers we have to deal with increase exponentially. The method of this paper proves the two questions in some degree. Fuzzy-C-Means clustering arithmetic is used to cluster the images, and Heuristic Least Mean Square is used to measure the fuzzy measure a ascertain mass function of evidence theory. Heuristic Least Mean Square decreases the complexity of parameter identification, and figure out the question of how to decide Basic Probability Assignment availably.The new method of image fusion is used in the multi-component flow recognition. We get clear fused result of multi-sensor pictures.