The Extraction And Change Detection of Plants Based on High-resolution Imagery at Haikou

摘要

Hainan Island (Province)is the second largest island in China. With the rapid urban growth and development of economy, some problems on ecology and environment emerge, among which the degradation of woodland and tropical rain forests is an important one, because it is partially caused by human activities and conversely it leads to other ecological problems. To make most of people realize the ponderance of the problem, we need to find the problem scientifically and describe it in a direct and pellucid way. Remote sensing is a powerful tool in the field of environment monitoring and change detection. To exert its larger potential, we have to strengthen the understanding of high-resolution imagery in order to extract more information. Haikou, the capital of Hainan Province, with its surrounding area is chosen as the study area in this research. Firstly, high-resolution SPOT and ALOS images are pre-processed and conducted image fusion to exploit the complimentary nature of various types of resolution in the same image type. Then, we select three methods to perform classification in order to extract the information about plant. Finally, we evaluate the accuracy of classification results, compare them and analyze the factors leading to the change.