A novel technique for content based image retrieval using color, texture and edge features

Abstract

The massive growth of digital technology along with use of internet has increased the use of audio-visual data such as images and videos in many domains like digital museums, commercial use, crime prevention, medical images, remote sensing and so on. With increasing volume of digital data, search and retrieval of relevant images from large data sets in accurate and efficient way is a challenging problem. CBIR combines the contents or features of image like color, texture, edges rather than keywords, labels related to an image. This paper presents systematic literature review of various image retrieval techniques presenting the basic concepts and available methods with their research gaps. In this study, retrieval techniques based on features like HSV, Color Moment, HSV and Color Moment, Gabor Wavelet and Wavelet Transform, Edge Gradient are studied and implemented. An approach is proposed for retrieval based on combination of color, texture and edge features of image. Performance evaluation of studied image retrieval techniques and proposed technique is done using parameters like Sensitivity, Specificity, Retrieval score, Error rate and Accuracy. Experimental results of performance evaluation demonstrate that proposed technique outperforms other techniques.