The remote sensing images captured at different time may be quite different in color

brightness because of various atmospheric conditions

seasonal changes in vegetation characteristics and other reasons.These differences not only cause great impact on subsequent image classification and change detection

but also bring difficulties in remote sensing image mosaicking.This paper presents a color calibration method based on remote sensing automatically classification and lαβ color space transformation.Multi-images mosaic with ETM captured at different time are experimented by applying this method.The results of experiments show that color calibration method can calibrate difference of color and brightness in images caused by the different time more effectively than conventional methods

such as overlapping zone calibration and histogram matching.So it can improve the accuracy of follow-up image classification and change detection

as well as the remote sensing images mosaicking effect.