城市绿地遥感信息自动提取研究--以广州市为例

提出了一种新的基于像元信息分解和神经网络分类相结合的城市绿地遥感信息自动提取方法。选择广州市作为研究区,先用像元信息分解法把绿地从TM影像中分离出来,再以分离出来的绿地作为分类掩膜,采用BP神经网络法进行分类。并开展野外遥感调查以提高和验证分类精度。结果表明:该方法保证了分类时绿地的纯洁度,有效地排除和避免了提取绿地信息时其它多余信息的干扰和影响,提高了分类的精度。

An Automatic Extraction Model of Urban Greenbelt Remote Sensing Image--A Study on Guangzhou City

A new automatic classification model of remote sensing image using pixel information decomposition combined with neural network classification is proposed in this paper. The author chooses Guangzhou City as a study area. At first, the greenbelt are separated from the TM image of Guangzhou by pixel information decomposition. Then the author takes it for the classified covering. Afterward, based on the classified covering, the author continues to subdivide by BP neural network classification. Ultimately, it carries out field investigation to improve and validate the classification precision. The results show that this method ensures the greenbelt purity and eliminates the disturbance and influence of unwanted objects effectively so it improves the classification precision.