An Overview of Feature Extraction Techniques in Content-Based Image Retrieval Systems

Abstract

The collection of images has grown rapidly and continues to increase in future due to widespread use of internet. Content based Image Retrieval (CBIR) system aims at retrieval of images of relevance to the query image input by the user from an enormous image database by low-level feature (such as color, texture and shape) extraction from the image. Some common applications of CBIR and existing commercial systems employing CBIR have been highlighted. General algorithm for explaining the working of CBIR is presented. This article presents a review on methods for low level feature extraction like extracting color, texture and edge features. CBIR still faces some challenges like judgement of human perception of visual content, less appropriate selection of similarity measure, semantic gap and other factors.