ResearchBib Share Your Research, Maximize Your Social Impacts
Sign for Notice Everyday Sign up >> Login

Low Quality Image Retrieval System For Generic Databases

Journal: International Journal of Scientific & Technology Research (Vol.4, No. 12)

Publication Date:

Authors : ; ;

Page : 67-72

Keywords : Index Terms Content Based Image Retrieval; Low Quality Images; K means Clustering Technique; Query Based Image Content; Image Binarization;

Source : Downloadexternal Find it from : Google Scholarexternal


Abstract Content Based Image Retrieval CBIR systems have become the trend in image retrieval technologies as the index or notation based image retrieval algorithms give less efficient results in high usage of images. These CBIR systems are mostly developed considering the availability of high or normal quality images. High availability of low quality images in databases due to usage of different quality equipment to capture images and different environmental conditions the photos are being captured has opened up a new path in image retrieval research area. The algorithms which are developed for low quality image based image retrieval are only a few and have been performed only for specific domains. Low quality image based image retrieval algorithm on a generic database with a considerable accuracy level for different industries is an area which remains unsolved. Through this study an algorithm has been developed to achieve above mentioned gaps. By using images with inappropriate brightness and compressed images as low quality images the proposed algorithm is tested on a generic database which includes many categories of data instead of using a specific domain. The new algorithm developed gives better precision and recall values when they are clustered into the most appropriate number of clusters which changes according to the level of quality of the image. As the quality of the image decreases the accuracy of the algorithm also tends to be reduced a space for further improvement.

Last modified: 2016-04-24 17:48:47