Is effectively reducing the amount of data that must be processed, while still accurately and completely describing original dataset. The main goal of feature extraction is to obtain the most relevant information from the original data and represent the information in a lower dimensionality space.
NEED FOR FEATURE EXTRACTION 1) To extract the core and unique information. 2) Reduce processing time. 3) Reduce the misclassification rate. 4) Right feature extraction technique reduce the complexity of the mathematical modeling.
TYPES OF FEATURE EXTRACTION TECHNIQUES a) Wavelet. b) Discrete Cosine Transformation (DCT) c) Harris Features – corner d) Histogram of Gradient (HoG) – object. e) Local Binary Pattern (LBP) – texture. f) Maximally stable external regions (MSER) – blob g) Surf features – object (speed up robust features) h) Eigen features – corner using eigen value algorithm. i) Fast feature – corner j) GLCM feature – texture. k) Gabor feature – texture analysis using Gabor filters.