Detection and Analysis of Urban Change in Remotely Sensed Imagery by Principal Component Analysis of Image Data
Journal: International Journal of Science and Research (IJSR) (Vol.3, No. 11)Publication Date: 2014-11-05
Authors : Ramesh A; Rubina Parveen; Priya Narayanan;
Page : 812-818
Keywords : Principal Component Analysis PCA; Multispectral Images; Eigen values; eigenvector; Covariance Matrix; Correlation Matrix; Urban changes detection;
Abstract
This paper exhibits the time variant multispectral transformation of image data for detection and analysis of urban changes across two or more images over time. The technique covered, which appeal directly to the vector nature of the image, includes, the principal component analysis transformation (PCA). The usefulness of PCA in processing of multispectral satellite images has been highlighted. PCA reduces image dimensionality by defining new, uncorrelated bands composed of the principal components (PCs) of the input bands. Thus, this method transforms the original data set into a new dataset, which captures better the essential information. IRS IC LISS III images of 2002 and 2011 of Gulbarga area were geometrically co-registered on which PCA for the urban change Detection was undertaken. A colour composite of Eigen images from the resulting PCA was used for analysis of urban change. It has been experienced that PCA effectively summarize the dominant modes of spatial, spectral and temporal variation in data in terms of linear combinations of image frames. It provides maximum visual separability of image features thus improving the accuracy of urban change detection and analysis.
Other Latest Articles
- Contra-Cyclical Capital Buffering Solutions for Banks
- Bayesian Approach for Spectrum Sensing in Cognitive Radio
- A Review of Web Service Recommendation Systems
- Study of Physico-Chemical Properties of Glycerol Ester Based Non-Ionic Gemini Surfactant
- Effect of Sperm Quality of Beef Cattle on Fertility Percentage
Last modified: 2021-06-30 21:12:54