A Study on Classifying Forest Type and Preparing Digital Forest Type Map Using High Resolution Satellite Imagery of IKONOS.
Woo-Kyun Lee, Jae-Seo Chong\ Greg S. Bigiog, Peng Gong and Hyun-Kook Cho

 

ABSTRACT : This study proved if the high resolution satellite imagery of IKONOS is suitable for preparing digital forest type map. Three methods, the pixel based classification with maximum likelihood(PML), the segment based classification with majority principle(SMP). and the segment based classification with maximum likelihood(SML), were applied to classify and deliminate forest type of IKONOS imagery taken in May 2000 in a forested area in the central Korea. The PML showed the poorest accuracy with overall accuracy of 0 . 57 and kappa value of 0.50. Through the SMP. the overall accuracy could be improved to 0.67 and the kappa value to 0.62. The SML showed the best performance, resulting in overall accuracy of 0. 70 and kappa value of 0. 64. And the both segment based classifications could attenuate the salt and pepper effect, while the pixel-based classification was not free from the salt and pepper effect. The segment based classification was proved to be more suitable for classifying and deliminating the forest type of the IKONOS imagery. And we could validate that a detailed forest type delimitated in the form of a palygon can be mapped on the basis of the segment based classification of high resolution IKONOS imagery.

 

Key words: digital forest type map, IKONOS imagery, pixel based classification. segment basea classification