OpenCV.js: Computer Vision Processing for the Web

The Web is the most ubiquitous computing platform. There are already billions of devices connected to the web that have access to a plethora of visual information. Understanding images is a complex and demanding task which requires sophisticated algorithms and implementations. OpenCV is the defacto library for general computer vision application development, with hundreds of algorithms and efficient implementation in C++. However, there is no comparable computer vision library for the Web offering an equal level of functionality and performance. This is in large part due to the fact that most web applications used to adopt a clientserver approach in which the computational part is handled by the server. However, with HTML5 and new client-side technologies browsers are capable of handling more complex tasks. This work brings OpenCV to the Web by making it available natively in JavaScript, taking advantage of its efficiency, completeness, API maturity, and its community’s collective knowledge. We developed an automatic approach to compile OpenCV source code into JavaScript in a way that is easier for JavaScript engines to optimize significantly and provide an API that makes it easier for users to adopt the library and develop applications. We were able to translate more than 800 OpenCV functions from different vision categories while achieving near-native performance for most of them.