论文部分内容阅读
针对斑点噪声严重的高分辨率SAR数据开展智能化影像分割方法研究的不足,该文基于面向对象的思想和极化SAR目标分解理论,提出基于目标分解的面向对象SAR影像分割算法,实现了Pauli基分解分割、Krogager基分解分割、Freeman基分解分割这3种典型的SAR目标分解下的影像分割方法。实验结果表明:所提出的不同极化基下的分割算法取得的结果较为理想、精度高,且整体上具有一致性;对于不同类型的特定目标,Pauli基分解分割算法对道路等表面散射体分割效果较好,Freeman基分解分割算法则更有利于植被等体散射体的分割。
To solve the problem of intelligent image segmentation based on high speckle SAR data, this paper proposes object-oriented SAR image segmentation algorithm based on object-oriented theory and polarimetric SAR target decomposition theory, and realizes Pauli Basic decomposition, Krogager basis decomposition and Freeman basis decomposition are the three typical SAR image decomposition methods. The experimental results show that the proposed segmentation algorithm with different polarizations has better results, higher accuracy and overall consistency. For different types of specific targets, Pauli-based decomposition algorithm can be used to segment surface scatterers The effect is better, Freeman-based decomposition segmentation algorithm is more conducive to the segmentation of vegetation body scatterers.