论文部分内容阅读
针对利用TM影像进行土地利用传统分类精度不高的问题,该文提出了一种综合应用影像纹理与光谱特征对TM影像进行土地利用模糊分类的方法。采用主成分分析法对研究流域TM影像的光谱及纹理特征信息进行压缩与融合,并对融合后的TM影像数据进行3个组别的多尺度分割,在影像分割对象单元的基础上应用面向对象的模糊逻辑隶属度函数法实现影像的软语义分类。相对传统分类方法而言,该方法在充分利用影像光谱信息的基础上综合了影像的纹理信息,且分类理论思想更加符合人们对于客观事物的认知规律,分类精度有了显著的提高,为TM影像分类方法的改进提供一定的参考。
Aiming at the problem of using traditional TM images for the traditional classification of land use, this paper presents a method for fuzzy classification of TM images based on the combination of image texture and spectral features. The principal component analysis (PCA) was used to compress and fuse the spectral and texture features of the TM images in the study watershed. The fused TM images were divided into three groups based on the multi-scale segmentation. Based on the image segmentation unit, Fuzzy Logic Membership Function Method to Realize Soft Semantic Classification of Images. Compared with the traditional classification methods, this method synthesizes the texture information of the image based on the full use of the spectral information of the image, and the classification theory is more in line with the people’s cognizance of objective things, and the classification accuracy has been significantly improved. Image classification methods to improve the reference to provide some.