A Fast Algorithm for SAR Image Segmentation Based on Key Pixels
Ronghua Shang,Yijing Yuan,Licheng Jiao,Biao Hou,Amir Masoud Ghalamzan Esfahani,Rustam Stolkin
DOI: https://doi.org/10.1109/jstars.2017.2743338
IF: 4.715
2017-12-01
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Abstract:Recent high-performance clustering methods process all pixels when segmenting an image, which results in a very large time complexity of these algorithms. Additionally, the performance of such algorithms can be severely affected by noise when dealing with highly polluted images. To address these problems, we propose a new unsupervised algorithm for segmenting synthetic aperture radar images based on a fuzzy clustering approach, called fast fuzzy C-means clustering based on key pixels. Our algorithm first selects a subset of special key pixels based on the rule of local extrema, and then performs image segmentation on only these key pixels using fuzzy clustering combined with nonlocal information. Next, the remaining non-key pixels can be rapidly segmented by combining the clustering results of the key pixels with a similarity metric rule which is robust to speckle noise. This approach greatly accelerates overall image segmentation because the time-consuming clustering operation is only performed on a small subset of pixels. We show the effectiveness of our proposed algorithm by a series of experiments including segmenting twelve simulated and four real synthetic aperture radar images. Moreover, to validate our results, we compare the segmentation results obtained by our algorithm with those obtained by seven other state-of-the-art segmentation algorithms from the literature. The experimental results suggest that our algorithm outperforms other state-of-the-art segmentation algorithms in both computational speed and speckle noise suppression.
imaging science & photographic technology,remote sensing,engineering, electrical & electronic,geography, physical