# Feature Extraction Essay

Better Essays

In many research works, the various feature extraction algorithms are used for remotely sensed imagery. In this proposed work, feature extraction is mainly divided into three steps, viz., Discrete Wavelet Transformation of Pixel Information, Principle Component Analysis using DWT coefficients, and then extraction of Gray Level Co-occurrence Matrix derived statistical features. All the steps are discussed here in detail. DWT Discrete Wavelet Transform or DWT is technique comprising of important features like localization of space frequency and multi-resolution. There is a great flexibility in DWT for choosing varying window size, bases and the low computational complexity [16]. Here in this work the complex signals are decomposed into sum …show more content…

ϕ=1/M ∑_(n-1)^M▒T_n … (4.9) Difference between the Input Matrix (Ti) with the Previous Average Input Matrix (ψ) is calculated. ϕ_i=T_i-ψ … (4.10) Calculation of the covariance matrix of the variable (Фi). C= 1/M ∑_n^M▒〖ϕ_n-ϕ_n^T 〗 … (4.11) Calculation of the Eigenvectors and Eigenvalues of the covariance matrix. The eigenvectors that obtained from the covariance matrix sorted according to the largest eigenvalues. Selecting the principal component, from M eigenvectors (Eigen faces), only M is chosen, which have the highest eigenvalues. The higher the eigenvalue, the more characteristic features of a face does the particular eigenvector describe. In MATLAB 2016a, pca is given as a function to obtain coefficients from principle components analysis of a given raw data vector. GLCM GLCM or Gray Level Co-occurrence Matrix is most widely used method of Texture Based Feature Extraction. GLCM, first proposed by Haralick in 1973, characterize texture using a variety of quantities derived from second order image statistics [21]. Co-occurrence texture features from an image are extracted in two steps. First, “the pair-wise spatial co-occurrences of pixels separated by a particular angle and/or distance are tabulated using a gray level co-occurrence matrix” (GLCM). Second, “the GLCM is used to compute a set of scalar quantities that characterize different aspects of the underlying texture”. These steps