Home

🚧 Page under construction.

Welcome to the cvGPUSpeedup wiki!

Here we want to explain:

1. How is the code organized, to help anyone contribute to the code.
2. The difference between cvGPUSpeedup and FusedKernel library (at some point we may split the repository in two).
3. Details on how to compile the tests and benchmarks in both Windows and Linux.
4. Details on how to use cvGPUSpeedup, and how to use the FusedKernel library.

As of August 2024, I'm using this respository to:

1. Prepare for my PhD that I'm going to enroll in September, and which will be based on the current and future ideas developed in this library.
2. Provide speedups for the CUDA code used in AutomaticTV's product - a GRUP MEDIAPRO company. AutomaticTV uses cvGPUSpeedup to speedup the image preprocessing of computer vision codes along with OpenCV objects and types, and the FusedKernel library to speedup the image processing pipeline that processes the input cameras and generates the video outputs.

Additionally, two contributors are starting to work on some aspects of the library. Hopefully, we will get more contributors with time, which will force me to define contribution rules, code styles and so on. I'm actually open to discuss those details with anyone interested in contributing.