Abstract

SAR images are generally available as multi-temporal images. To recognize the changes between these images precise calibration and perfect spatial registration are required.Due to high asymmetry in the image histograms and minor unavoidable calibration errors, conventional image pre-processing techniques cannot be adopted in case of SAR images. A chain of pre-processing techniques with cross normalization is suggested. The pre-processing chains involving different histogram equalization techniques like Histogram Equalization, Adaptive Histogram Equalization, Local Histogram Equalization and Contrast Limited AHE are used in the image pre-processing chain and results are compared to identify best suited image pre-processing chain cross-normalization process.