A collection of hands-on OpenCV tutorials, focused on the parts of the library that turn a camera stream into something a machine can actually act on: deep learning inference, GPU acceleration, and real-time pipelines.
| Tutorial | Description | |
|---|---|---|
 |
Installing OpenCV 5 with CUDA and DNN Support on Ubuntu 24.04 | Build OpenCV 5.0.0 from source with the CUDA DNN backend. The NVIDIA driver, CUDA and cuDNN stack, every CMake flag that decides whether GPU inference works, the NumPy 2 trap Ubuntu 24.04 sets for you, and a five-step verification ladder that catches the failures OpenCV never reports. |
 |
Detecting Everyday Objects with YOLO, OpenCV DNN and CUDA | Run a COCO-trained YOLO model in real time using nothing but OpenCV and an NVIDIA® GPU. Complete Python and C++ code, the letterbox and output-tensor maths explained, CUDA build instructions, and the OpenCV 5 engine trap that silently sends your inference back to the CPU. |