Temu Kembali Citra Berbasis Konten Pada Citra Lintas Domain

Autor:	Yaya Wihardi, Herbert Siregar, Ali Mulyawan
Jazyk:	indonéština
Rok vydání:	2018
Předmět:	Image Retrieval CBIR Gaussian Approximation OASIS Technology Information technology T58.5-58.64 Business HF5001-6182
Zdroj:	Jurnal Computech dan Bisnis, Vol 12, Iss 1, Pp 34-46 (2018)
Druh dokumentu:	article
ISSN:	1978-9629 2442-4943
Popis:	This research aimed to develop a cross domain content based image retrieval system by using Histogram of Oriented Gradient (HoG) features. To compute visual similarity, we use Gaussian Approximation for Fast Image Similarity (GAFIS), Online Algorithm for Scalable Image Similarity Learning (OASIS), and Hybrid method that combine both of GAFIS and OASIS. The results show that the hybrid method outperforms the others, it is because the methods could extract uniqueness of each image query and image database by utilizing Gaussian parameters from negative-set and combine it with OASIS parameters
Databáze:	Directory of Open Access Journals
Externí odkaz:	https://doaj.org/article/c918d5d48ca445b1826dc60e99a2f9e5 Zobrazit plný text záznamu View record in DOAJ