Extracting crop information with remote sensing is important contents that the remote sensing technology was used in the field of agriculture, it is the important component of the resource remote sensing too. Remote sensing technology as front technology of the geo-information science, it is the most effective observing technology and the means of obtaining information at present and has objective, prompt characteristics. It can obtain the ground information on large-scale in succession in a short time and realize collecting fast and analyzing quantitative for agricultural information. So remote sensing technology provides new opportunity for agricultural modernize management, and it is the basic means with scientific decision for agriculture China has carried on agricultural monitor work with remote sensing since 1980s, and has made great progress now.However, crop information was often extracted based on NOAA/AVHRR data in the past year. For restriction of AVHRR data, its monitoring accuracy needs further improvement in crop monitor.Terra/MODIS is a new remote sensing sensor. It provides new opportunity for crop monitor on large-scale. This thesis sets out to research how to make full use of MODIS data for crop monitor, and develops new methods and models for extracting crop information according to MODIS spectral characteristic on large-scale.Main research results and initiatives in this thesis include as following:(1) Characteristic bands and time series selection based on MODIS data Based on the agricultural knowledge and biological characteristic of the crop, combined with MODIS data and its agricultural engineering application, the bands and phonological stage fitting for crop monitor were selected. Red, blue, NIR and ESWIR were selected as working bands and land surface water index (LSWI), enhanced vegetation index (EVI) were selected as working parameters. In the end, there were 9 phonological stages as the best time phase of remote sensing. So we...
|