TemplateMatching

TemplateMatching is a Julia package designed to offer a native Julia implementation of template matching functionalities similar to those available in OpenCV. This package aims to provide an easy-to-use interface for image processing and computer vision applications, allowing users to leverage the high-performance capabilities of Julia for template matching operations. The package offers performance slightly below that of OpenCV but significantly better than a naive implementation.

Features

Masks are not yet supported in the current version of the package. Unlike OpenCV, TemplateMatching.jl supports n-dimensional arrays¹.

Below is a table summarising available methods and their equivalent in opencv.

TemplateMatching.jl	Mask	OpenCV equivalent
SquareDiff	Not yet supported	TM_SQDIFF
NormalizedSquareDiff	Not yet supported	TM_SQDIFF_NORMED
CrossCorrelation	Not yet supported	TM_CCORR
NormalizedCrossCorrelation	Not yet supported	TM_CCORR_NORMED
CorrelationCoeff	Not yet supported	TM_CCOEFF
NormalizedCorrelationCoeff	Not yet supported	TM_CCOEFF_NORMED

Installation

To install TemplateMatching, use the Julia package manager. Open your Julia command-line interface and run:

using Pkg
Pkg.add("TemplateMatching")

Usage

Almost everything you need is the function match_template and its inplace counterpart match_template!.

Below is a quick start example on how to use the TemplateMatching package to perform template matching:

using TemplateMatching
using Images

# Load your source image and template
source = rand(1000, 1000)
template = source[400:500, 100:150]

# Perform template matching using square difference
result = match_template(source, template, SquareDiff())

# Get the best match
argmin(result) # CartesianIndex(400, 100)

# Perform template matching using normalized correlation coefficient
result = match_template(source, template, NormalizedCorrCoeff())

# Get the best match
argmax(result) # CartesianIndex(400, 100)

Documentation

For more detailed information on all the functions and their parameters, please refer to the full documentation.

Possible improvements

• Support for template mask (planned)
• Support for GPU
• Improve performances
• More tests and examples in documentation
• Better errors

License

TemplateMatching is provided under the MIT License. Feel free to use it in your projects.

Footnotes

1. Up to 64 dimensions because of an implementation detail, but this shouldn't be a problem in most cases. ↩