Study Classification Of DMKD In Remote Sensing Data Based On Rough Sets

In recent years, with the fast development of Earth Observation, database and network technology, and the prevalence and improvement of the observation station constructed, all types of spatial data, including resource, environment and disaster, is increasing in exponent way. It's the key field to research how to discover the knowledge people needed from the data is a very absorbing and challenging field. Because of the Remote Sensing data a very important part of the spatial data, how to obtain the connotative knowledge from the Remote Sensing data is the bottleneck of spatial data mining, and should be resolve firstly. As the Remote Sensing data had the characteristic of complicacy and uncertainty, the general mathematical statistic method can not deal with the uncertainty effectively. Rough sets theory is a new mathematical approach to uncertain and vague data analysis. It is, no doubt, one of the most challenging areas of modern computer applications nowadays and a new very important and rapidly growing area of research and applications.This paper treated the Remote Sensing data like a knowledge system, and discussed how to use the Rough sets theory to classify the Remote Sensing data automatically. The paper also designed and carried out a model of automatic classification based on Rough sets theory. The main research work and suggestions is:â‘ Investigated the uncertainty mechanism in Remote Sensing data and the mechanism Rough sets deal with the uncertainty data. And the theoretical framework of uncertainty in Remote Sensing information processing via Rough sets theory is analyzed in detail.â‘¡Discretize the Remote Sensing data before classification can improve the efficiency and precision of the classification. This paper discussed and carried out the discretization of Remote Sensing data.â‘¢Knowledge reduction is very important in the process of obtain rules based on Rough sets theory. Because of the high waste of time and memory space the general algorithm has, this paper proposed a new algorithm based on discernibility matrix and analogical matrix, and carried out the algorithm.â‘£We can get the rules after knowledge reduction. Then we use these rules to design the tools to classify Remote Sensing data, and evaluate the precise of the result to prove the validity of the. tools.(5) Last, this paper summed up the model of automatic classification based on Rough sets theory.The question how to handle the uncertainty during the process of dealing with Remote Sensing data affected the precise of the result of classification; well mechanism can improve the precise of classification.