Adaptive Query Image Searching Methods Using Improved Hash Coding and Unified Color and Intensity level Matrices
Year: 2016
Type: article
Abstract: In this thesis we had implemented Hash coding with bit wise hamming distance for similarity measurement between the feature vectors of query and database, which will be more secured and real time based application. By using the hash coding method the computational time is high and the accuracy will be much lower. Hence, hash coding will be extended with inclusion of gray scale matrices (GSM), which will represent the features of gray scale texture values in the image, but it will not consider the color information in the image. To incorporate both gray scale and color information, we proposed a novel scheme in which both color and intensity variations are represented in a single composite feature known as Unified Color and Intensity level Matrices (UCILM). The proposed scheme has been merged with hash coding to improve the efficiency of the retrieval system. This addition work had improved the system accuracy as well as the precision time with almost 88% of relevant image retrieval.
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W2420975673