Study On LUCC By Remotely Sensed Data

As the advance of research upon global change, the study of LUCC has been one of foci. The technology of Remote Sensing and GIS offers an effective tool for the rapid, precise and dynamic acquisition of information on LUCC.The main purpose of the paper was to find effective change detection techniques. Wuxi, Jiangsu province was taken as the test area. By using data mining and knowledge recovery, some improvements have been made upon the traditional Image Differencing Method (IDM). First, subtraction between the chosen bands was conducted; second, the potential changed pixels in each subtracted image were extracted by using a threshold to generate a binary image; third, the changed of the imagery was masked with the binary image and classified; at last, land usage transformation matrix was computed. Post-classification Method (PCM) also has been applied to the same imagery to derive change of land use /cover and transformation matrix to cross-validate the advanced IDM. The result showed that the improved IDM has characteristics of maneuverability, veracity and convenience. Landscape pattern has been explored and discussed. Summarily, the main content was as follows:1. The extraction of information of LUCC and the verification of advanced IDM. Wuxi was taken as an instance, and two methods were used in this paper to generate LUCC information upon land use in Wuxi. After independent extraction of the changed of LUCC respectively by two methods, the comparison between them has been made. As a result, the trend of change derived from two methods were the same, the only difference was that the results from the PCM were less than that of the post-classification method. For more exact results from advanced IDM, two crucial steps-bands selection and threshold value finding should be optimized further.2. Landscape pattern and its spatio-temporal change of Wuxi. This paper discussed about the landscape pattern of Wuxi in 1991 and 2001, and also about the spatio-temporal change of that. The results showed that the fragmentation of landscape was to some extent deeper and the shape of it was becoming more complex. The dominance and evenness of the landscape had no change because of the shortage of dominant class and the even distribution of each patches.