基于独立成分分析和形态特征的干涸水系提取

Extraction of dry drainage system based on independent component analysis and morphological characteristics

基于遥感影像的水系提取在地学研究中应用广泛。目前的提取方法大多是依据水系中的水体光谱特征。但是在冲沟、处于枯水期的季节性河流以及长期干涸的河道中,基于水体光谱特征的方法并不适用。为此提出了基于独立成分分析(independent component analysis,ICA)和形态特征的干涸水系提取方法。利用ICA技术将水系与其他地物分解至不同的独立分量,并针对水系分量中少量的非水系信息和噪声,利用基于中值滤波的背景抑制、数学形态学滤波和基于形态特征的噪声消除,对ICA提取的水系进行进一步的图像增强和去噪,最终分割出水系。以内蒙古自治区乌拉特后旗获各琦地区为例,对提出的模型进行实际测试,并与传统的监督分类方法进行了对比。结果表明,本文提出的方法去噪效果更好,在干涸水系提取中的应用效果更为理想,且不需要训练数据,操作简便,具有很强的实用性。

The extraction of drainage system is necessary in many geoscience research fields.For instance,drainage system is an important indicator for structure and lithologic interpretation,sample sites in stream sediment geochemical exploration are designed according to drainage system,and drainage system needs to be recognized and masked in mineral alteration extraction.The drainage system in remote sensing image is generally extracted according to spectral features of water body.However,in the dry drainage systems,such as gullies and seasonal rivers in dry season and under prolonged dry condition,the method based on water body is not applicable.To tackle this problem,the authors propose a method based on independent component analysis(ICA).ICA is a signal decomposition technique that converts multispectral data to independent components which represent independent signal sources,thereby enhancing and separating the specific target in the image.The streambed system extracted by ICA may be still accompanied by noisy data.A series of methods are used to enhance image and remove noise,which include background suppression,morphological filtering and de-noising based on morphological features.The proposed method was tested with ASTER data from Huogeqi area of Urad Rear Banner in Inner Mongolia Autonomous Region,and the result was compared with that derived from supervised classification.The results indicate that the method proposed in this paper can be used to identify dry drainage system,and the recognition result is better than the traditional supervised classification method.The method put forward by the authors performs better in interference information reduction and de-noising,and training data are not needed in this method.In conclusion,the method proposed in this paper is ideal and practical in the extraction of dry drainage system.