1. Image correction
In the process of remote sensing imaging, system distortion and random distortion are inevitable. Generally, the data purchased by satellite ground stations have undergone geometric coarse correction. This module mainly realizes geometric fine correction of purchased remote sensing image data, that is, geometric correction is carried out by manually selecting control points. It uses a mathematical model to approximately describe the geometric distortion process of remote sensing images, and obtains this geometric distortion model by using some corresponding points (that is, control point data pairs) between distorted remote sensing images and standard images (maps). The calibration object is the original remote sensing image data without precise geographical coordinates, and the reference object is the calibration image data of the same area. Users interactively select the ground control point (GCP) for correction, and finally output the corrected image data.
2. Image mosaic
In the practical analysis and application of the system, when the research area is at the junction of multiple images or the research area is large enough to be covered by multiple images, it is necessary to register the images covering the research area, and then splice these images to facilitate unified processing, interpretation, analysis and research. There are great defects in manual stitching, but using computer to stitch remote sensing images will greatly improve the quality of stitched images and create a good foundation for later image processing applications.
This function completes the splicing of image data of two adjacent scenes. It is required that the input image data of each scene has overlapping areas in geographical position and is the image data after geometric fine correction. The module can automatically judge the overlapping area according to its geographical location, and adjust the color and feather the boundary of the adjusted image according to its statistical characteristics, so as to obtain seamless and color-matched large-format mosaic data.
3. Image cropping
A lot of mosaic data are needed in interpretation and analysis, but there are more practical needs in data output. Therefore, the image clipping function based on standard map sheet and arbitrary boundary is developed. The cutting functions of1:10000 and1:100000 standard plates have been developed, which are used in key monitoring areas with medium resolution monitoring system (1:100000 precision) and high resolution monitoring system (/kloc)
At present, the river management in China implements the parallel system of river basin management and administrative management, and the data of resources and environment are counted by river basin and administrative division respectively. The system has developed the output image data clipping function with arbitrary vector boundaries. Users can clip the remote sensing image data of Wanquan River basin according to the boundaries of each basin or administrative boundaries by calling the "four sources and one stem" boundary data or the boundaries of administrative divisions within the basin.
4. Information enhancement
Information enhancement is one of the most basic methods of remote sensing image processing, which aims at: ① adopting a series of technologies to improve the visual effect of the image and improve the clarity of the image; ② Transform the image into a form more suitable for human or machine interpretation and analysis. This function is not based on the principle of image fidelity, but tries to selectively highlight some interesting information that is convenient for human or machine analysis and suppress some useless information to improve the use value of images, that is, image enhancement processing only enhances the resolution of some information.
Information enhancement is a relative concept. The enhancement effect is not only related to the quality of the algorithm itself, but also directly related to the data characteristics of the image. At the same time, because the evaluation of image quality often depends on the subjective of the observer, there is no universal quantitative standard, so the enhancement technology is mostly problem-oriented, and the specific enhancement methods are selectively used by users.
The system has developed functions such as data stretching, saturation stretching and decorrelation stretching, and users can set the stretching range and data output range as required.
Step 5 register quickly
Complete the fast registration of the data before and after the two periods, quickly and automatically select matching points for the data before and after the two periods, and then directly correct the later data to achieve the purpose of fast registration.
6. Image fusion
In the construction of remote sensing monitoring system, remote sensing image data obtained by many different sensors such as MODIS, ETM, ASTER and SPOT are used. These data constitute multi-source data in the same area in space, time and spectrum. The image data of a single sensor usually cannot extract enough information to complete some applications. Multi-sensor data fusion can give full play to the characteristics of various sensor images and obtain more information. This is also the purpose of developing image fusion function.
The system develops an image fusion function module based on HSV fusion method, which integrates three steps: low-resolution data resampling, HSV forward transformation and HSV inverse transformation, making the fusion operation simple and convenient. By fusing image data with different resolutions, the high-resolution color image obtained not only has high spatial resolution, but also has the same hue and saturation as the image, which is beneficial to visual interpretation and automatic information extraction.