Study on Image Retrieval based on Image Texture and Color Statistical Projection

Published: 13 July 2016| Version 1 | DOI: 10.17632/55b7ywy2sn.1
Contributors:
Bing Tang,
Xiaofei Zheng

Description

This article presents an image texture and hue statistical projection based retrieval. First the image is converted to HSI color model, the gray value of the image extraction, and Robert algorithm to extract the texture, then the image is divided into blocks and extracts the main color block, the main color image blocks are respectively projected in the horizontal and vertical direction of 2, get 2 projection histogram, the 2 projection histograms of the first three order center extraction distance and Robert algorithm as the features of texture, image similarity calculation. Make a very full pave the way for future Canny edge processing algorithm research of image retrieval.

Files

Steps to reproduce

Firstly, the image is converted to HSI color model, the gray value of the image extraction, and Robert algorithm to extract the texture; then, the image is divided into blocks and extracts the main color block, the main color image blocks are respectively projected in the horizontal and vertical direction of 2, get 2 projection histogram, the 2 projection histograms of the first three order center extraction distance and Robert algorithm as the features of texture image similarity calculation; finally,make a very full pave the way for future Canny edge processing algorithm research of image retrieval.

Institutions

Chinese Academy of Tropical Agricultural Sciences

Categories

Information Economics

License