Data Fusion and Information Comparison of Different Data Sources

The fusion of remote sensing data with different features enhances the interpretation ability of images and improves the application effect. Image fusion has become a very valuable technology in remote sensing application technology. In the fusion method, ISH color transformation has the advantages of flexibility and practicality, which has been paid more and more attention by researchers at home and abroad in recent years, so various forms of IHS transformation (Jia Yonghong,1998; Li Xuwen,1992; Sun et al., 1998), in which image fusion based on pixel weighting, image fusion based on wavelet theory and image fusion based on category (Gendern,1994; Agostinelli (1992) is a mature technology at present.

In this study, three data sources, TM, SPOT and aerial color infrared remote sensing photography, were collected, with resolutions of about 30m, 10m and 2m respectively. Due to the low flying height of aerial data, the image distortion from the center to the edge of the image is great. In addition, the resolution of SPOT and TM data is quite different from that of other two data sources, and the fusion effect is not good through experiments. However, SPOT and TM data are complementary. In plain areas, these two data sources are often used for fusion processing, and applied to the monitoring and investigation of urban land use, environment and ecology. SPOT sensor is different from TM, which often obtains data from the side, resulting in different projection distortion of the two data sources. This distortion is greater in mountainous areas, and it is sometimes difficult to record and correct these two data. The research on these two kinds of data fusion is also based on the following reasons:

1) It is impossible to extract both spectral information and texture information by using two data sources alone.

2) Although the resolution of 2)SPOT panchromatic data image is high, it is difficult to reflect the shadow information in the study area because of the large terrain fluctuation, and the shadow area accounts for a large proportion.

3) The fused information is more suitable for human eyes to recognize and analyze various remote sensing information.

Through the strict registration and correction of SPOT and TM images, the fusion image of spot and TM in the study area (plate13) is obtained by using the pixel-based ISH transformation method. By comparing the fused image with the SPOT panchromatic image (plate 14 and plate 13), it is found that most details of the original SPOT shadow area are clearly reflected, which provides important help for the next remote sensing interpretation.