基于改进SLIC算法的SAR图像海陆分割
|
摘要
海陆分割在提高SAR图像舰船目标检测精度方面具有十分重要的意义。针对传统算法不能很好地对SAR图像进行海陆分割,提出了基于改进SLIC超像素分割和分层区域合并准则(HSWO)的海陆分割算法。针对SAR图像统计特性,首先对SLIC超像素分割和HSWO算法模型分别进行改进,然后用SLIC超像素分割算法对图像进行超像素分割,并按照分层区域合并准则对超像素块进行聚类,最终实现海陆分割。实验表明,所提出的改进模型具有较高的处理精度和处理效率,相比于其他算法更适用于SAR图像的海陆分割,具备一定的工程应用价值。
Sea-land segmentation is of great significance for improving the accuracy of SAR target ship detection. Aiming at the fact that the traditional algorithm can not carry out high-quality sea-land segmentation for the SAR image this paper proposes a segmentation algorithm based on the improved SLIC superpixel partitioning and the Hierarchical Area Combination Criterion( HSWO). According to the statistical characteristics of SAR images the SLIC and HSWO algorithm models are improved respectively at first. Then the SLIC algorithm is used to segment the superpixels in the image and the superpixel blocks are clustered according to the merging rules of the hierarchical regions. Finally the segmentation between sea and land is implemented. Experiments show that the improved model proposed here has higher processing precision and efficiency. Compared with other algorithms it is more suitable for the sea-land segmentation of SAR images and has certain engineering application value.
Sea-land segmentation is of great significance for improving the accuracy of SAR target ship detection. Aiming at the fact that the traditional algorithm can not carry out high-quality sea-land segmentation for the SAR image this paper proposes a segmentation algorithm based on the improved SLIC superpixel partitioning and the Hierarchical Area Combination Criterion( HSWO). According to the statistical characteristics of SAR images the SLIC and HSWO algorithm models are improved respectively at first. Then the SLIC algorithm is used to segment the superpixels in the image and the superpixel blocks are clustered according to the merging rules of the hierarchical regions. Finally the segmentation between sea and land is implemented. Experiments show that the improved model proposed here has higher processing precision and efficiency. Compared with other algorithms it is more suitable for the sea-land segmentation of SAR images and has certain engineering application value.
引文
[1]王春瑶,陈俊周,李炜.超像素分割算法研究综述[J].计算机应用研究,2014,31(1):6-12.
[2] REN X FMALIK J. Learning a classification model for segmentation[C]//The 9th IEEE International Conference on Computer VisionWashington DC2003:10-17.
[3] FELZENSWALB P FHUTTENLOCHER D P. Efficient graph-based image segmentation[J]. International Journal of Computer Vision200459(2):167-181.
[4] SHI J BMALIK J. Normalized cuts and image segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence200022(8):888-905.
[5] MOORE APRINCE SWARRELL Jet al. Superpixel lattices[C]//IEEE Conference on Computer Vision and Pattern Recognition2008:1-8.
[6] VINCENT LSOILLE P. Watersheds in digital spaces:an efficient algorithm based on immersion simulations[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence199113(6):583-598.
[7] COMANICIU DMEER P. Mean Shift:a robust approach toward feature space analysis[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence200224(5):603-619.
[8] LEVINSHTEIN ASTERE AKUTULAKOS K Net al. Turbopixels:fast superpixels using geometric flows[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence200931(12):2290-2297.
[9] ACHANTA RSHAJI ASMITH Ket al. SLIC superpixels compared to state-of-the-art superpixel methods[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence201234(11):2274-2282.
[10]蔡淑宽,刘金清,施文灶,等.基于改进SLIC与区域邻接图的高分辨率遥感影像建筑物提取[J].计算机系统应用,2017,26(8):99-106.
[11] BEAULIEU J MGOLDBERG M. Hierarchy in picture segmentation:a stepwise optimization approach[J].IEEE Transactions on Pattern Analysis&Machine Intelligence198911(2):150-163.
[12]韩翠娟.高分辨率Pol SAR图像的超像素分割方法研究[D].哈尔滨:哈尔滨工业大学,2017.
[13] OTSU N. A threshold selection method from gray-level histogram[J]. IEEE Transactions on SystemsManand Cybernetics19799(1):62-66.
[14] LIU M YTUZEL ORAMALINGAM Set al. Entropy rate superpixel segmentation[C]//Computer Vision and Pattern RecognitionIEEE2011:2097-2104.
[2] REN X FMALIK J. Learning a classification model for segmentation[C]//The 9th IEEE International Conference on Computer VisionWashington DC2003:10-17.
[3] FELZENSWALB P FHUTTENLOCHER D P. Efficient graph-based image segmentation[J]. International Journal of Computer Vision200459(2):167-181.
[4] SHI J BMALIK J. Normalized cuts and image segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence200022(8):888-905.
[5] MOORE APRINCE SWARRELL Jet al. Superpixel lattices[C]//IEEE Conference on Computer Vision and Pattern Recognition2008:1-8.
[6] VINCENT LSOILLE P. Watersheds in digital spaces:an efficient algorithm based on immersion simulations[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence199113(6):583-598.
[7] COMANICIU DMEER P. Mean Shift:a robust approach toward feature space analysis[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence200224(5):603-619.
[8] LEVINSHTEIN ASTERE AKUTULAKOS K Net al. Turbopixels:fast superpixels using geometric flows[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence200931(12):2290-2297.
[9] ACHANTA RSHAJI ASMITH Ket al. SLIC superpixels compared to state-of-the-art superpixel methods[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence201234(11):2274-2282.
[10]蔡淑宽,刘金清,施文灶,等.基于改进SLIC与区域邻接图的高分辨率遥感影像建筑物提取[J].计算机系统应用,2017,26(8):99-106.
[11] BEAULIEU J MGOLDBERG M. Hierarchy in picture segmentation:a stepwise optimization approach[J].IEEE Transactions on Pattern Analysis&Machine Intelligence198911(2):150-163.
[12]韩翠娟.高分辨率Pol SAR图像的超像素分割方法研究[D].哈尔滨:哈尔滨工业大学,2017.
[13] OTSU N. A threshold selection method from gray-level histogram[J]. IEEE Transactions on SystemsManand Cybernetics19799(1):62-66.
[14] LIU M YTUZEL ORAMALINGAM Set al. Entropy rate superpixel segmentation[C]//Computer Vision and Pattern RecognitionIEEE2011:2097-2104.