With the rapid development of science and technology, modern warfare has developed to the stage of comprehensive competition in the four-dimensional space of land, sea, air and sky. In local wars under modern high-tech conditions, night observation capability has become an important component of the army's combat effectiveness. The single-band night vision technology has developed to the present, and has a fairly complete theory and relatively mature technology.

However, in many occasions, it is difficult to complete the detection and recognition tasks under various target backgrounds with only a single sensor. For example, a green tank in front of a concrete building will be difficult to distinguish in a low-light night vision system, but easy to distinguish in an infrared imaging system. If the night vision system is used to detect and image the green tank in front of the concrete building in two or more frequency bands, and further image fusion processing is performed on the formed image, the above scenes can be distinguished at night. Using the complementarity of image information from different sensors can greatly improve the detection capability of photodetection systems.

Information fusion (Information Fusion) is an information processing technology that processes multi-resource information to obtain improved new information.

In the field of multi-sensor information fusion, image fusion is one of the aspects, and it is also the most widely used and published aspect. It is a multi-sensor information fusion system only for visual information (image information). Image fusion can be described as follows: the images obtained by different sensors for the same target or scene, or multiple images (such as TVIRCTSAR multispectral images) obtained by the same sensor at different times or in different ways, are represented in different data representations. Synthesized at the level, the fusion information can reflect the information of multiple original images in a centralized manner, so as to achieve a more accurate and comprehensive analysis and judgment of the target and scene.

The above advantages of image fusion technology make it fully understood before the application of modern aviation customs, automatic control, weather forecast and other fields, especially in the field of military command is playing an increasingly important role. In terms of military applications, image fusion technology is widely used in reconnaissance, detection, identification, tracking and guidance systems; in terms of resource applications, monitoring of the utilization of the two places, statistics of vegetation coverage, environmental investigation and monitoring, flood disaster detection Prediction and evaluation, etc.; in computer vision, image fusion is considered to be a breakthrough direction to overcome some current technical difficulties; on various operating platforms of aerospace and aviation, the composite fusion of a large number of spectral remote sensing images obtained by various remote sensors, It provides an effective self-processing means for the efficient extraction of information.

At present, the main purposes of applying image fusion technology to digital image processing are as follows:

1) Increase the content of useful information in the image, improve the clarity of the image, and enhance some features that cannot be seen and clearly seen in a single sensor image.

2) Enhancing the content of light information, obtaining complementary image information for improving detection, classification, understanding, and recognition performance.

3) The image sequence fusion to detect the changes of the scene and the target.

4) Use images from other sensors to replace or make up for missing or faulty information in a certain sensor image. Obviously, image fusion technology is different from image enhancement in the general sense, it is a new technology in the field of computer vision and image understanding. Considering the purpose and advantages of image fusion, the image fusion system generally has the following requirements:

1) The image fusion algorithm used should make the fused image

Contains important information from the source image.

2) The image fusion algorithm should not introduce any misinformation that misleads human visual perception or image processing.

3) The image fusion system should have good stability, robustness and fault tolerance.