Image is composed of the digital numbers including information on natural phenomena, their condition and the kind of objects. Digital numbers are changed in geometric correction preprocess.
This change of digital numbers give an effect on results of ...
Image is composed of the digital numbers including information on natural phenomena, their condition and the kind of objects. Digital numbers are changed in geometric correction preprocess.
This change of digital numbers give an effect on results of land-cover classification.
This study can analyze the effect of resampling in classifying land-cover using the image reconstructed by geometric correction. Study area is selected Chun-Cheon basin which have very variable land-cover pattern in North-Han river region and is used of RESTEC data resampled in preprocessing.
Land-covar is classified by six classes of LEVEL I by maximum likelyhood classification method.
Land-cover classification was preformed with the image resampled by two methods - bilinear interpolation and nearest. Bilinear interpolation method was more accuracy in five classes excepted bare-land according the result of comparing each class with topographic map.
I also could understand the difference the seasonal patterns in the of spring and autumn image.
Resultly, the resampling method must be choosed according to the class, and accuracy of classification can be elevated by using four-season's image.