Abstract: This article introduces a new application of Genetic Algorithm in the field of remote sensing (Signal's data classifier). Based on Simple Genetic Algorithm, an improved performance could be attained through the data mining process with necessary object-related information from a seties of EM radar signal images. After the preprocessing stage, the target's lacating task could be modeled into a curve imitation process. Beyond the heredity from SAG, expert knowledge and environment conditions are put into consideration in the practical problem resolution. We suggest a possible way to locate the underground object in the relics detecting field. Compare with ordinary calssifier method GA is better on speed and accuracy of searching on target. In further work we shall pattern 3-D model of the relics using complex GA.

Key words: genetic algorithm, remote sensing targets classifier, digital image process, genetic algorithm, remote sensing targets classifier