This article attempts to get the information on rocky desertification under different background conditions in the karst area using spectral features and mathematical method,so as to improve the accuracy and timeliness of desertification intelligent classification.The Yachi demonstration area is taken as a study area and different platforms and different bands of remote sensing data are used.Based on the 3S technologies,the spatial distribution and characteristics of the karst area are summarized through processing,analysis,and classification.A quantitative research is carried out to understand the spectral characteristics in different conditions and diversities.Results from the study provide some references for other areas to develop the interpretation of remote sensing intelligent desertification.