Abstract
The information of the pixel inside consists of synthesis signals which depend on the characteristics of reflectance factors of landcover materials. This study aims to analyze the model of synthesis signals of multispectral data, and to develop an estimation method of pixel inside information.
If it is possible to estimate or extract the information of pixel inside with a simple method, the contents of landcover classified data will change from qualitative or logical information to quantitative information. In the field of water resources, it is already possible to extract the details of the snow-covered area by using LANDSAT data. But obtaining unclouded LANDSAT data is not so easy in the winter season of Japan. On the other hand, it is possible to obtain unclouded NOAA AVHRR data. Because the NOAA satellite covers the same area twice a day. But, spetial resolution of AVHRR data is very low, compared with LANDSAT MSS and TM.
This study aims also to develop the methodology and procedures on how to extract details of pixel inside information from low spatial resolution sensor data such as NOAA AVHRR data.
The estimation method of pixel inside is very simple, then it is easy to use in image operation, and it's results have some accuracy to use practical analysis.
