Home

Image Search Engine
Indexing and searching images by content

Introduction

Development of an image search engine that allows to extract from an image database, the most similar images (visual similarity) to a reference image.

In this work, I develop a technique based on color attributes to characterize images. This technique consists of calculating the histogram in order to have information relating to the frequency distribution of the colors of the image. The histogram is calculated for each of the images in the database. Once these images are characterized by their histograms, I use the Euclidean distance to measure the degree of similarity between them.

Steps for the application

1. Image search scenario
2. Representation of an image
3. Weighted color histogram
4. Quantification
5. Similarity measurement between images

Steps for the user

1. Select a target directory to index
2. Execute the indexing
3. Select a query image
4. Choose the target directory to search
5. Execute the search
6. Display the most similar images in order