Image Retrieval System Using Fusion of Texture Features with Various Distance Metrics

computer science engineering

Abstract- From last few years, the need of massive database used to store vast images has been developed rapidly and will also grow in the future. Content Based Image Retrieval System (CBIRS) provides us at most outcome to fetch the images from massive dataset. We present the proposed work which is “Fusion of CBIRS using Color and texture with various distance metrics”. In these systems, two features like Color and Texture are used to represent an image. For Color feature extraction, techniques like Color histogram (CH), Color Moment (CM) have been described. Grey level co-occurrence matrix (GLCM) is used for texture features extraction. In this paper, fusion of Color and Texture based retrieval system is described. By using the fusion approach, better results can be produced with the higher precision value. Retrieval time of images is more so we design a CBIR system using many distance metrics for similarity calculation and by using the best distance metric for reducing the Retrieval time. Euclidean distance, KL-Divergence distance, Manhattan distance, Jaccard, Cosine distance is used. GLCM technique which is used for texture features extraction obtained 91% average precision, LBP features obtained 92% average precision and fusion of both texture features extraction obtained 95% average precision.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.