Open AccessSPATIAL-SPECTRAL FUZZY K-MEANS CLUSTERING FOR REMOTE SENSING IMAGE SEGMENTATION
Author(s) -
Dinh Sinh,
Ngo Thanh Long,
Trinh Le Hang
Publication year - 2018
Publication title -
vietnam journal of science and technology/science and technology
Language(s) - English
Resource type - Journals
eISSN - 2815-5874
pISSN - 2525-2518
DOI - 10.15625/2525-2518/56/2/10785
Subject(s) - cluster analysis , fuzzy clustering , cure data clustering algorithm , flame clustering , correlation clustering , pattern recognition (psychology) , spectral clustering , data stream clustering , artificial intelligence , mathematics , canopy clustering algorithm , single linkage clustering , data mining , k medians clustering , dbscan , computer science , fuzzy logic
Spectral clustering is a clustering method based on algebraic graph theory. The clustering effect by using spectral method depends heavily on the description of similarity between instances of the datasets. Althought, spectral clustering has been significant interest in recent times, but the raw spectral clustering is often based on Euclidean distance, but it is impossible to accurately reflect the complexity of the data. Despite having a well-defined mathematical framework, good performance and simplicity, it suffers from several drawbacks, such as it is unable to determine a reasonable cluster number, sensitive to initial condition and not robust to outliers. In this paper, we present a new approach named spatial-spectral fuzzy clustering which combines spectral clustering and fuzzy clustering with spatial information into a unified framework to solve these problems, the paper consists of three main steps: Step 1, calculate the spatial information value of the pixels, step 2 applies the spectral clustering algorithm to change the data space from the color space to the new space and step 3 clusters the data in new data space by fuzzy clustering algorithm. Experimental results on the remote sensing image were evaluated based on a number of indicators, such as IQI, MSE, DI and CSI, show that it can improve the clustering accuracy and avoid falling into local optimum.