Skip to content

Latest commit

 

History

History
60 lines (31 loc) · 1.16 KB

README.md

File metadata and controls

60 lines (31 loc) · 1.16 KB
Raw

Bow-Retrival

BOW algorithm based on color histogram.

中文教程： https://blog.csdn.net/zxwsbg/article/details/97681139

个人网站： https://zxwsbg.cn

Introduction

An image retrival method based on Bag-of-words model using color histogram.

How to use

Just clone the repo to your own computer and then run it in Matlab. Follow the order run the .m program.

Method

step1. Extract Features

  • Convert RGB components to YUV components

  • save 1536*64 matrix for all images

step2. Select Features

  • Divide a picture into several 8*8 blocks, and tile each 8*8 block into a 1*64 vector

  • Choose a ratio, and then select some of those 1*64 vectors to cluster.

step3. Cluster

  • Use Kmeans to cluster features.

step4. Classify

  • Divide the original features into the specific class.

  • Generate the frequency histogram of each image.

step5. Calculate Distance

  • Calculate the distance between each two images.

  • Sort the distance matrix.

step6. Retrival

  • Use a image to test.

  • Calculate Presicion-Recall value.

How to contribute

  • Fork the reposity
  • Create Branch
  • Commit your code
  • Create Pull Request