基于TM影像的福州市区绿地提取研究

Extraction of Green Space in Fuzhou Based on TM Image

城市绿地是影响市区环境的重要因素,对建设城市生态文明起着重要作用。本文在TM影像的基础上,通过对各个波段的分析,求出最优波段组合并基于这种组合进行影像分类,通过与原始影像分类和NDVI分类结果的对比得出:1)最优波段组合影像分类效果优于原始影像分类;2)最优波段组合影像分类结果与NDVI分类结果十分相近,与实际情况相符;3)NDVI分类效果虽然较好,但与基于最优波段组合影像分类相比,其阈值确定困难,需要多次提取感兴趣区,随机性较大。

Green space is an important part of the urban environment, and plays a significant part in the ecological of urban. Based on the TM images, through the analysis of each band, to find the optimal band combination ,Then, Classification of it. Compare the clas- sification results with the classification of original image, the classification of NDVI, draw a conclusion as ：①Best band combination is better than the original image classification. ②The optimal band combination image classification results were very similar with the re- sults of classification of NDVI, accords with the actual situation. ③the classification of NDVI is better, but compared with the optimal band combination based on image classification, the threshold determination difficult and requires multiple extractions region of interest randomness.