In this paper, an unsupervised segmentation methodology is proposed for remotely sensed images by using Fractional Differential (FD) based texture analysis model and Iterative Self-Organizing Data Analysis Technique Algorithm (ISODATA). Essentially, image segmentation is used to assign unique class labels to different regions of an image. In this work, it is transformed into texture segmentation by signifying each class label as a unique texture class. The FD based texture analysis model is suggested for texture feature extraction from images and ISODATA is used for segmentation. The proposed methodology was first implemented on artificial target images and then on remote sensing images from Google Earth. The results of the proposed methodology are compared with those of the other texture analysis methods such as LBP (Local Binary Pattern) and NBP (Neighbors based Binary Pattern) by visual inspection as well as using classification measures derived from confusion matrix. It is justified that the proposed methodology outperforms LBP and NBP methods.
TopIntroduction
Image segmentation is the process of subdividing an image into multiple regions. In other words, a label could be assigned to each pixel in the image such that pixels with the same label share a particular characteristic. Especially, in this work, each image region is assigned with a unique label representing a texture class. It is always true that low level image analysis that is based on classified image regions is more expressive than the one based on individual pixels. Hence, image segmentation is found by many researchers to be the most significant task for precise image interpretation (Sezgin, 2004).
Most of the conventional methods being adopted for image segmentation (Roy, 2014; Lu, 2007) depend on spectral features of an image, and they may lead to wrong segmentation, particularly in case of remotely sensed images. Therefore, image spatial features are preferred over spectral features for segmentation. It is noticed that texture, one of the spatial features, is contained in most of the images of natural scenes, ragged or worn surfaces of many objects and particularly in remote sensing images. Hence, the texture features are expected to provide a significant contribution for image segmentation (Haralick, 1973; Chen, 1995; Ojala, 2001; Li, 2003; Liu, 2006; Ranjan, 2014). It is also found that texture feature based segmentation is found to be more appropriate in many of the computer vision applications such as remote sensing (Chakraborty, 2012; Tsai, 2005; Roy, 2014; Cheng, 2013), defect detection in industrial applications (Xie, 2008; Iyer, 2014), medical imaging (Hajar Danesh, 2014) and content based image retrieval (Yue, 2011; Zhang, 2000).