CONTENT BASED IMAGE RETRIEVAL USING MULTI-SVM TECHNIQUE FROM NOISY AND NON NOISY IMAGES
Journal: International Journal of Engineering Sciences & Research Technology (IJESRT) (Vol.9, No. 8)Publication Date: 2020-08-30
Authors : Pooja Rani; Asst Prof Sarbjit Kaur;
Page : 93-96
Keywords : : Content Based Image Retrieval; CBIR; Image Mining; Image Processing; Cluster Distance.;
Abstract
Image mining is a technique which handles the mining of information, image data association, or additional patterns not unambiguously stored in the images. It exploits methods from computer vision, image retrieval, image processing, data mining, machine learning, database, and artificial intelligence. In the proposed work, we have developed a new system that can retrieve the images from a dataset on the basis of contents of the query image. Here, ‘Content-Based' means that the search will analyze the actual contents of the image. The existing system does not evaluate the results upon attacks but in proposed system the results are also being evaluated on adding noise to the images and blurring the images. The overall average accuracy of the proposed system is 96% whereas that of existing system is 85%. Performance of the existing systems is checked on the maximum of 1000 images whereas the performance of the proposed system is checked on more than 5000 images.
Other Latest Articles
- HIDING TEXT DATA IN AUDIO SIGNALS USING 3BIT LSB TECHNIQUE AND DCT APPROACH
- ONION BULB CONSERVATION: CASE OF AN EXPERIMENTAL STUDY
- QUALITY OF SERVICE PARAMETERS EVALUATION IN CELLULAR NETWORKS
- COMPOUND FAULT DIAGNOSIS OF BEARINGS USING AN IMPROVED SPECTRAL KURTOSIS BY MAXIMUM CORRELATION KURTOSIS DECONVOLUTION (MCKD)
- DETERMINATION AND EVOLUTION OF PHOTOELECTRICAL PARAMETERS BASED ON CUINSE2 AND CUINS2 HETEROSTRUCTURE UNDER ILLUMINATION
Last modified: 2020-09-22 07:24:14