OpenCV.jl is a Julia package that provides an interface to the popular computer vision library OpenCV. It allows Julia users to leverage the extensive functionalities and algorithms offered by OpenCV for various computer vision tasks, such as image and video processing, object detection, feature extraction, and more.

• Comprehensive OpenCV bindings: OpenCV.jl provides comprehensive bindings to the OpenCV library, enabling Julia users to access a wide range of computer vision algorithms and functionalities.
• Efficient and performant: The package is designed to leverage the speed and efficiency of both Julia and OpenCV, ensuring high-performance computer vision tasks.
• Interoperability: The package supports interoperability between Julia arrays and OpenCV's data structures, facilitating data exchange and manipulation between the two.

You can install OpenCV.jl using Julia's package manager. Open a Julia REPL and run the following command:

] add OpenCV

This will download and install the package and its dependencies.

To start using OpenCV.jl in your Julia code, you first need to import the package:

using OpenCV

Once imported, you can call OpenCV functions and use its data structures. Here's a simple example that loads an image and applies a Gaussian blur:

List of wrapped methods is available here: wrapped methods

using ImageCore, OpenCV, TestImages, MosaicViews, ImageView
img_orig = testimage("Mandrill")
img_raw =  collect(rawview(channelview(img_orig)))
img_gray = OpenCV.cvtColor(img_raw, OpenCV.COLOR_RGB2GRAY)
img_blur = OpenCV.blur(img_gray, OpenCV.Size(Int32(9), Int32(9)))
img_back =  colorview(Gray, float.(img_blur)./255)[1,:,:]
imshow(mosaic(img_orig, img_back; nrow=1))

For more information on the available functions and usage examples, please refer to the documentation.

OpenCV.jl welcomes contributions from the community. If you encounter any issues, have suggestions for improvements, or would like to contribute code, please check the GitHub repository.

OpenCV.jl currently just directly provides OpenCV_jll without a higher level interface written here. It provides an avenue for further development using OpenCV bindings. For contribution, there are two very important components that must be maintained i.e. YggDrasil build_tarballs.jl which build OpenCV_jll which is available here and the Julia Bindings of OpenCV which are available here. To build the Julia Bindings, this blog post will be of help. Also, @archit120 's blog posts might be of interest.

OpenCV.jl is licensed under the MIT License. Please refer to the license file for more information.

OpenCV.jl is built upon the excellent work of the OpenCV community. We are grateful for their continuous efforts in advancing the field of computer vision. @archit120 has been instrumental in the development of this package during GSOC 2021 and GSOC 2020.