| Aiming at the application of monitoring land use in the region of returning cultivated land to forest by remote sensing, this thesis studied the methods and techniques of picking up the land use change area automatically, and mainly studied monitoring woodland dynamic, which was based on change detection of land use. Moreover, this thesis analyzed land use dynamic and the transform rules of land types at the different time, and analyzed mostly woodland dynamic. In the end, according to analyzing the source of woodland from farmland, the impacts of carrying out returning cultivated land to forest are monitored.Two satellite images, one from TM and the other one from SPOT, were used. The technique of change detection was post-classification comparison. The resultant classification of land use was obtained using the methods of supervised classification and expert classification. The change area of studied region was obtained to monitor the land use dynamic by overlaying the resultant classification of two periods.The results indicate that:â‘ The precision of RS classification of land use is improved by the three methods, fusing TM image and SPOT PAN image, creating new bands NDVI BI GVI WI PC2 using TM bands, and combining the supervised classification and expert classification.â‘¡ The structure of land use of studied region changes in evidence from 1997 to 2003, and the rate of change dynamic of land use is very high. Cultivated land area decreases greatly, and woodland and construction land increase greatly, and water area and traffic land increase gently. The rate of change dynamic of cultivated land is very high, which is -5.4%. The rate of change dynamic of woodland is 0.9%, but the change gross is great. It is the residential land that is the highest one of all change rates, which is 5.4%. All these indicate that the project, returning cultivated land to forest, affects greatly the land use of studied region.â‘¢ Most of new forest land is from cultivated land. Among the amount of returning cultivated land to forest land in 2003, the probability of re-forest land from cultivated land is 81.24%, and the probability of re-forest land is 91.7 percent of cultivated land transferred. It is indicated that the source of returning cultivated land to forest is reasonable,and that the land use of re-forest region changes in favor of eco-environment.For the problems of the precision finite and the weather influence, it is proposed to classify land use using high resolution, high spectrum, radar and infrared remote sensing. It is also proposed to monitor completely returning cultivated land to forest, according to space distribution. |