Index

1. Introduction
2. Features

Introduction

Colour histogram matching is a simple method for comparing the visual similarity of images or image regions. Despite its simplicity, it can be quite effective, especially if an appropriate colour space is chosen. The process consists of two parts. First, the colour histograms of the images or image regions, which shall be compared, are computed. Then the matching score is computed based on a histogram distance metric. The histogram matching module implements the computation of the colour histograms as well as different distance metrics to compute the distance score. Currently, the histogram intersection, the Hellinger distance and the earth mover’s distance are implemented.

Features

HistogramMatching library provides functions to:

• Convert images into different colour spaces.
• Compute colour histograms.
• Compute histogram matching scores.