Other
JCBIR is a Content Based Image Retrieval System using Wavelet Transform and K-Means Clustering developed in Java.
JCBIR have developed as a Final Project in Information Retrieval Lecture at Institute Technology of Sepuluh Nopember
Algorithms of this application based on this paper (1) , but we've added K-Means for clustering with euclidean distance to improve query speed.
Dataset for evaluation has been taken from http://wang.ist.psu.edu/docs/related/.
(1) Latha, Jinaga. Content Based Color Image Retrieval via Wavelet Transforms, IJCSNS International Journal of Computer Science and Network Security, 2007. http://paper.ijcsns.org/07_book/200712/20071205.pdf
Libraries Used :
Apache Lucene (http://lucene.apache.org)
Efficient Java Matrix Library (http://code.google.com/p/efficient-java-matrix-libra
图像
基于
内容
检索
No comment