An Innovative Method for Image Retrieval using Weighted Edge Matching Algorithm

  • R.Tamilkodi and G.Rosline Nesakumari

Abstract

With enormous development in diverse kinds of images through electronic communication networks, it becomes a demanding chore to retrieve the efficient result from a wide collection of images. CBIR i.e. content based image retrieval answers the trouble as it selects visual image contents or features to deal with images. Since images are represented by definite features to create possible exact retrieval of the essential images here this article proposed a method to extract certain features from the image based on color, texture, and shape. This manuscript has proposed an move towards an weighted edge matching information retrieval (WEMIR) to execute content based image retrieval. It is a fusion move towards to extract the color, texture and shape features from images. With the sole feature extraction, acceptable outcomes are not formed. Hence more than one feature extraction is developed to carry out retrieval of images. To extract the color feature, the higher order of confined mean is used to improve the lower contrast to gain high contrast. To extract the texture features multi optimization techniques are used and for shape feature extraction weighted edge matching technique is used. Pixel content is extracted from each image present in the database as well as for the test image provided. Based on the proposed method WEMIR the optimal features are obtained for query and image database. By using image distance measure corresponding images are retrieved from the database. The competence of the proposed WEMIR is calculated using precision and recall. The proposed method shows better retrieval results while compared with the traditional methods.

Published
2020-06-04
How to Cite
R.Tamilkodi and G.Rosline Nesakumari. (2020). An Innovative Method for Image Retrieval using Weighted Edge Matching Algorithm. International Journal of Advanced Science and Technology, 29(11s), 1256 - 1267. Retrieved from https://sersc.org/journals/index.php/IJAST/article/view/20819
Section
Articles