A Study On Land Use Survey And Dynamic Change Monitoring By Using Remote Sensing

Remote sensing is used for land resource survying, monitoring and land of resources programming etc. In order to improve the interpretation precision of remote sensing images, this paper selects Dingxiang county of Shanxi province, processes the TM/ETM images of this area obtained by America's Landsat-5and Landsat-7 by using ERDAS, EVNI and ArcGIS.Before the image processing, geometric correction has done for the RS images and scanned relief map, ect. Then, mask is used to cut the images into different images for the Dingxiang area. In order to enhance spectrum characters and difference, image interpreter is done such as convolution, principal components, decorrelation stretch, RGB to HIS, HIS to RGB, ect.The study area land-use classification system is thoroughly sumed up in this paper. Two methods are adopted to extract land information. One is visual interpretation, for combineing interpretation indoor and investigate in field to different land-use in images. Another is computer automatic classification, which is produced by integrating unsupervised classification and supervised classification procedures before image interpretation. The precision of visual interpretation is higher. However, it needs more manual work and good specialty skill of interpreters. It is not suit for large areas and complicated land use areas. Computer automatic classification based on Maximum likelihood needs to choose training areas as many as possible and the precision is lower, but it is more quickly. Post classification process is done by clump and eliminate in ERDAS. And then, the classification result changed into vector carry on by ArcGIS for geography information analysis and statistic.Three methods are carried out in monitoring land use, band differencing, multi-temporal composite, and post classification comparison. Band differencing is feasible in water and irrigable field, but it couldn't monitor the changes of whole land use and its precision is the lowest. For the mufti-temporal composite method, it is difficult to choose the exact regions for change detection. Based on the previous accurate classification result, the post classification comparison method gets the best result compared with the other two methods, but also with the hand-on mistakes.In conclusion, to cut images into several pieces based on geomorphological structure, using the computer automatic classification and post classification comparison methods are most effective in land use information extracting and dynamic change monitoring, especially in the alluvial plain regions.