I am working on a tracking solution for an AR engine that needs to be deployed and work at minimum of 25 FPS on IOS and Android devices.
So far i have been planning to optimize performance by multi threading and running on GPU pipeline, but i found out that when using the gpu.hpp implementation, it will only work on Nvidia GPUs (it is using CUDA in the background).
Is there a snapdragon equivalent implementation for OpenCV that provides similar performance boost as the Nvidia CUDA implementation?
This forum is disabled, please visit https://forum.opencv.org