CBIR on Biometric Application using Hough Transform with DCD ,DWT Features and SVM Classification

Content based image retrieval (CBIR) has been possibly the greatest significant enquiry areas in computer science for the last decade. A retrieval way which mix texture, color and shape feature is future in this paper. In this research, implemented a novel method for CBIR using Hough Transform ,DCD and DWT feature with Support vector machine (SVM) as a classifier. In the process of feature extraction, firstly extract texture feature using discrete wavelet transform (DWT), extract color feature using dominant color descriptor (DCD) on RGB and HSV color space for improving computation and efficiency and for line detection use Hough Transform of images. The experimental dataset contain 444 images including facial images. The match size is considered utilizing weighted Euclidean distance (WED). For improving effectiveness of the system, classify data using RBFSVM. Performance analysis is depend on precision, accuracy and F-measure. Keywords—CBIR; DWT; Hough Transform;DCD;RGB and HSV color.