A Region Feature Based Automatic Matching for Optical and SAR Images
Cai Zhi-gang · Journal of China University of Mining and Technology · 2007
Aiming at resolving the problem of lower-precision and labor-consuming in matching for synthetic aperture radar(SAR) and optical remote sensing images,a matching method based on region features was proposed.Speckle noises in SAR images were suppressed by a modified Lee filter which takes the advantage of median filter into account.Morphological dilations were used to link image edges extracted by Canny operator.Image segmentation based on region growing was used to extract closed regions in SAR and optical image as invariable features.Precise feature matching was guided by a cost function,which considering region areas,perimeters,positions of center points,etc.After cross matching,the center points of the matched region were used as matching control points.The experimentation of registering Envisat ASAR and Landsat TM images shows the whole workflow could be accomplished automatically with a precision of lower than 1 pixel.