Abstract
Abstract. In this paper, a novel superpixel-based approach is introduced for unsupervised change detection using remote sensing images. The proposed approach contains three steps: 1) Superpixel segmentation. The simple linear iterative cluster (SLIC) algorithm is applied to obtain lattice-like homogenous superpixels. To avoid discordances of the superpixel boundaries obtained from bi-temporal images, the two images are firstly fused using principle component analysis. And then, the SLIC algorithm is applied on the first three principle components, which contain the main information of the two images. 2) For each superpixel, which is considered as the basic unit of the image space, the multi-dimensional change vector is computed from spectral, textural and structural features. 3) The superpixels are classified into two type: changed and unchanged through two progressive classification processes. The superpixels are firstly cataloged into three types: changed, unchanged and undefined by thresholding the change vectors and a voting process. And then the undefined superpixels are further classified into two classes: changed and unchanged, using a SVM-based classifier, which is trained by the derived changed and unchanged superpixels from the former step. The experiment using Indonesia data set has confirmed that the proposed approach is able to detect the changes automatically, by exploiting multiple change features.
Highlights
Change detection is the process of identifying differences in the state of an object or phenomenon by analyzing a pair of images acquired on the same geographical area at two different instants (Singh, 1989)
The work flow of the method can be simple divided into three steps: superpixel segmentation, multi-dimensional change vector extraction and unsupervised progressive classification, where the last step can be divided into two sub-steps
The details of the simple iterative cluster (SLIC) algorithm is described in Algorithm.1: Algorithm 1 The SLIC algorithm 1: Initialize cluster centers by sampling pixels at regular grid steps S; 2: Perturb cluster centers in an n × n neighborhood, to the lowest gradient position; 3: For each cluster center Ck, assign the best matching pixels from a 2S × 2S square neighborhood around the cluster center according to the distance measure in Equation(3); 4: Compute new cluster centers and residual error E {Distance between previous centers and recomputed centers}; 5: Repeat step (3) and (4) until E ≤ threshold; 6: Enforce connectivity; 2.2 Superpixel-based Change Vector Extraction
Summary
Change detection is the process of identifying differences in the state of an object or phenomenon by analyzing a pair of images acquired on the same geographical area at two different instants (Singh, 1989). Compared with pixel-based method, the object-oriented approach is a recently developed knowledge-based technique It considers landscapes as aggregations of meaningful objects corresponding to ground entities and patches of surface cover (Dronova et al., 2011). It starts with a segmentation process, and is followed by a successive analysis with the help of expert knowledge It represents a obvious advantage in analyzing high-resolution data because image pixels are meaningfully grouped into networked homogeneous objects, and the noises are reduced (Lu et al, 2011). It enable us to use the statistical tools for automatical analysis, which have been widely used in pixel-based approaches It considers the superpixel as the smallest unit, which enables us to model the changes in an object (region) -based manner, where region-based spectral, textural and structural differences can be utilized
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
