Font Size: a A A

Research On Key Technologies Of Content-based Natural Landscape Image Retrieval

Posted on:2018-08-07Degree:MasterType:Thesis
Country:ChinaCandidate:F Y HuFull Text:PDF
GTID:2358330518460496Subject:Signal and Information Processing
Abstract/Summary:PDF Full Text Request
This paper focuses on the landscape image retrieval technology research with Corel database as the data source.The main work is as follows:l)image retrieval based on color and texture features of images;2)image retrieval based on K-means and improved BOF word bag model;3)design and implementation of the image retrieval system.In detail,the research work includes:1.Image retrieval based on color and texture features of images.First of all,this paper introduces the principle and process of extracting color features based on HSV color space,and extracting texture features based on gray level co-occurrence matrix and color texture co-occurrence matrix.Then,the image retrieval application of the fused color histogram and gray level co-occurrence matrix,as well as the fusion of color histogram and color co-occurrence matrix are analyzed and compared.Finally,the advantages and disadvantages of different fusion methods are compared through experiments.2.Image retrieval based on K-means and improved BOF word bag model Considering the advantage of color word bag(BOF)based on SIFT feature in image retrieval,it is introduced into this paper.This paper first proposes a HSV color space based on bag of words(BOC)method,then combines them into the BOF vector.The improved BOF bag model based on SIFT feature fusion are gained for image retrieval,finally verifies before and after the improvement of retrieval results by experimental data.3.Design and implementation of the image retrieval system.In order to facilitate the subsequent retrieval,the corresponding image retrieval system is developed on the basis of the previous methods.And it mainly discusses the development environment,system structure design and functional modules.
Keywords/Search Tags:Image retrieval, color feature, texture feature, K-means, BoF, Feature fusion
PDF Full Text Request
Related items