欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
基于聚类的分割算法能够有效地分析目标特征在特征域的分布结构，进而准确判断目标的所属类别，但难以利用图像的空间和边缘信息，而基于区域增长的分割算法能够在空间域利用多种图像信息计算目标之间的相似性，但缺乏对特征结构本身的深层挖掘，容易出现欠分割或过分割的结果．本文结合这两种算法各自的优势，针对合成孔径雷达（Synthetic aperture radar，SAR）图像的特点，提出了一种基于上下文分析的无监督分层迭代算法．该算法使用过分割区域作为操作单元，以提高分割速度，降低SAR图像相干斑噪声的影响．在合并过分割区域时，该算法采用了分层迭代的策略：首先，设计了一种改进的模糊C均值聚类算法，对过分割区域的外观特征进行聚类分析，获得其类别标记，该类别标记包含了特征的分布结构信息．然后，利用多种SAR图像特征对同类区域的空域上下文进行分析，使用区域迭代增长算法对全局范围内的相似区域进行合并，直到不存在满足合并条件的过分割区域对为止，再重新执行聚类算法．这两种子算法分层交替迭代，扬长避短，实现了一种有效的方法来组织和利用多种信息对SAR图像进行分割．对模拟和真实SAR图像的实验表明，本文提出的算法能够在区域一致性和细节保留之间做到很好的平衡，准确地分割出各类目标区域，对相干斑噪声具有很强的鲁棒性．
英文摘要：
Cluster based segmentation algorithms can effectively capture the structure of features so as to accurately determine the classes of objects, but they are difficult to make use of spatial information and edges in images. Region growing based segmentation algorithms can adopt different kinds of features to compute the similarity between objects, but they lack the analysis of features~ structure and often result in under-segmentation or over-segmentation. This paper combines the advantages of the two kinds of segmentation algorithms~ and proposes a context based unsupervised hierarchical iterative algorithm for synthetic aperture radar （SAR） image segmentation. This algorithm adopts over- segmented regions as operation elements to improve computation speed and reduce the influence of speckle noise. While merging the over-segmented regions, this algorithm chooses a hierarchical iterative strategy： a modified fuzzy C-means algorithm is first designed to analyze the appearance-based features of over-segmented regions, and then a region iterative growing scheme is used to merge the similar regions based on contextual analysis in space domain. After that, a new loop of these two iterative sub-algorithms begins, which is a hierarchical process and realizes a natural and effective way to use different kinds of information to segment SAR images. Experiments on synthetic and real SAR images indicate that the proposed algorithm can obtain excellent segmentation results and make a good balance between region consistency and preserving （SAR） image details.
同期刊论文项目
同项目期刊论文
Hierarchical spatial pyramid max pooling based on SIFT features and sparse coding for image classifi
Shape-Adaptive Reversible Integer Lapped Transform for Lossy-to-Lossless ROI Coding of Remote Sensin
A PSO-based automatic relevance determination and feature selection system for hyperspectral image c
Minimum-Entropy-Based Autofocus Algorithm for SAR Data Using Chebyshev Approximation and Method of S
Unsupervised Change Detection in SAR Image Based on Gauss-Log Ratio Image Fusion and Compressed Proj
Improve the performance of co-training by committee with refinement of class probability estimations
Multivariate Compressive Sensing for Image Reconstruction in the Wavelet Domain: Using Scale Mixture
An efficient matrix bi-factorization alternative optimization method for low-rank matrix recovery an
Compressive spectrum sensing in the cognitive radio networks by exploiting the sparsity of active ra
Change detection in synthetic aperture radar images based on evolutionary multiobjective optimizatio
期刊信息
位置：