摘要
将１９９６年获取的４个时相的Ｒａｄａｒｓａｔ图像用于广东肇庆地区的稻田分类试验，结果表明，多时相Ｒａｄａｒｓａｔ数据对水稻类型的识别精度较高，而且稻田的轮作规律容易推测出来。本文系统地介绍了这一试验研究的最新进展，探讨了神经网络分类方法在ＳＡＲ图像处理中的应用潜力和Ｒａｄａｒｓａｔ数据在中国南方水稻监测中的最佳时相选择和有效分辨率问题。
It gets a good result, that the multi-temporal Radarsat data are applied to paddy Field Classification in Zhaoqing area, Guangdong province, China. The discrimination of paddy field can reach a high accuracy, and the rotation process in paddy field can also be inferred easily. This paper demonstrate the latest development in this experimental research. Emphasis has been placed on the potential of neural network classifier's application to SAR image processing and the optimum Radarsat data selection for paddy monitoring in the southern China.
出处
《国土资源遥感》
CSCD
1997年第4期1-6,13,共7页
Remote Sensing for Land & Resources
