子视光谱相关属性的相干目标点探测方法

Identification of Coherent Scatters Based on Sub-look SAR Image Correlation Properties

鉴于目前的PS点识别方法都基于大数据集情况,适用于小数据集的PS点识别方法鲜有研究,相干点目标与分布式散射体相比,可视为由一个或者多个主散射体构成的反射体,对应的单视复数影像的相位值在频谱域显现一种显著的光谱特征,依据相干目标点具有强光谱相关性的特点,提出了基于子视光谱属性的相干目标点探测方法,通过案例验证该方法识别的相干目标点不仅具有强振幅值和干涉相干值,而且在时间序列中保持稳定的相位值。通过分析相关窗口大小与子视相干值的关系,发现子视相干值和窗口大小成反比,窗口越大,相干值越小;相干窗口的大小变化对人工建筑物的影响较小,对植被覆盖区域的影响较为明显。该方法不对数据量要求,即使单景SAR影像也能识别出相干目标点,为小数据集的时间序列分析提供了新的相干目标点识别方法。

Comparing with the distributed scatters,coherent scatter target is considered as a reflector constituted by one or more main scatters.Its phase value of single look complex image（SLC）includes only the distance phase between the scatter target and the sensor,and the phase of coherent scatter target shows a kind of significant spectral features in the spectral domain.The identification of coherent scatters target in single and multi-images has been addressed based on coherent scatter target with high spectral correlation among sub-look SAR images.The properties of the detected coherent scatter target can be regarded as the PS point,which shows that they not only have high amplitudes and high interferometric coherence,but also maintain stable phase in the time series.By analyzing the relationship between the correlation window size and sub-look coherent value,we found the window size is inversely proportional to sub-look coherent value.The change of coherent window size has little effect on artificial structures but obvious influence on the vegetation coverage area.The proposed method needs a limited number of images different from the identification of PS that relies on the availability of time series with multiple images.Even in single SAR image,coherent scatters can be efficiently identified.Therefore,this paper provides a new coherent target identification,at the same time it extends the applications of the InSAR.