cv-opencv

Context: $ARGUMENTS

OpenCV 4.x via the Python cv2 binding. Python is the first-class surface — the C++, Java (javadoc), and JS (opencv.js) APIs are parallel and the same concepts port over. Canonical docs: https://docs.opencv.org/4.x/.

Quick start

• Inspect an image (shape/dtype/channels): → Step 2 (cv.py info)
• Grab N frames from a webcam / file: → Step 3 (cv.py capture)
• Detect faces with the modern YuNet model: → Step 4 (cv.py detect-faces)
• Track a bounding box across frames: → Step 5 (cv.py track)
• Run an ONNX model via the dnn module: → Step 6 (cv.py yolo)
• Calibrate a camera from a chessboard: → Step 7 (cv.py calibrate)
• Stitch a panorama: → Step 8 (cv.py stitch)

When to use