A survey of multi-source image fusion

Multi-source image fusion has become an important and useful new technology in the image understanding and computer vision fields. The purpose of multi-source image fusion is to intelligently synthesize image data from multiple information sources, to generate more accurate and reliable descriptions and judgments than single-sensor data and to make fused images more consistent with human and machine visual features. Although there are many studies on multi-source image fusion, few papers summarize both theoretical and experimental aspects. This paper reviews, classifies and discusses the more advanced multi-source image fusion methods. We comprehensively introduce existing image fusion evaluation methods and compare them based on different standards. The representative algorithms are evaluated by using 12 famous target fusion metrics, and the advantages and disadvantages of each type are discussed in detail. Through research, the challenges encountered in this field and possible future research directions and development prospects are discussed.