PROJECT TITLE :
Gabor Feature Based Unsupervised Change Detection of Multitemporal SAR Images Based on Two-Level Clustering
During this letter, we have a tendency to propose a straightforward yet effective unsupervised modification detection approach for multitemporal synthetic aperture radar pictures from the angle of clustering. This approach jointly exploits the robust Gabor wavelet representation and also the advanced cascade clustering. First, a log-ratio image is generated from the multitemporal pictures. Then, to integrate contextual information in the feature extraction method, Gabor wavelets are utilized to yield the representation of the log-ratio image at multiple scales and orientations, whose maximum magnitude over all orientations in each scale is concatenated to form the Gabor feature vector. Next, a cascade clustering algorithm is intended in this discriminative feature area by successively combining the first-level fuzzy c-suggests that clustering with the second-level nearest neighbor rule. Finally, the 2-level combination of the changed and unchanged results generates the final modification map. Experimental results are presented to demonstrate the effectiveness of the proposed approach.
Did you like this research project?
To get this research project Guidelines, Training and Code... Click Here