Processing Method Of Multi-temporal Remote Sensing Images And The Use In Damage Assessment

This thesis is based on the Basic Research Program of National Defense: Multi-Source Dynamic fusion, System Estimate and Damage Assessment Research. The main content is the research of multi-period images change detection, and an elementary damage assessment research with actual multi-period images of battlefield strike. The main purpose is developing an effective, fast multi-period remote sensing images change detection method, and finally applies it in the damage assessment system. The computer technology is used to deal with the multi-period images, find out the geometry characteristics of the objectives in the images, and detect changes in the area of interest, achieves the purpose of damage assessment. The remote sensing images change detection technology is one of the most widely and basic used remote sensing technology, and applying this technology in the damage assessment system is a new, quickly developing area.Existed image processing methods and change detection algorithms is summarized in this thesis, and a lot of experiments have been done, the main work and innovations are: 1. A new effectual image processing method based on edge-enhance was surveyed. Firstly, the proper convolution operator was chosen to do the crisp enhancement, and then an improved Sobel Operator was used to enhance the edge.2. An improved texture processing method was surveyed. A series of mathematics models were set up to do the operation of the images.3. Based on the characteristic of existing images and the application background of damage assessment, after the research and the summarization of existing change detection technology, the direct- discrepancy method was used to detect the change of multi-period remote sensing images by setting the relevant threshold of change detection.4. Based on a certain judgment criteria, both quantificational and qualitative elementary damage assessment analyses were made.Practical damage assessment system in military is very complex and all-sided. It has very high demand on pertinence, Real-time, validity and so on. Based on the experiment results, the image processing method surveyed has the practical sense in this thesis.