OpenCV Image Analysis using Camera4Kivy.
2023-11-13 This repository is archived.
Uses OpenCV to edge detect the image stream, and replaces the usual Preview image with the detection result.
Available on most of the usual platforms.
On a desktop the example assumes a builtin camera facing the user, so the Preview is mirrored. If you want to point a freely mounted camera you may want to set the Camera4Kivy mirror option False.
The example demonstrates basic usage of the image analysis api. Analyzing image frames, annotating the results to the Preview (including when mirrored), and interacting with the annotation.
This example depends on Camera4Kivy. Read about Camera4Kivy because, depending on the platform you may need to install a camera provider.
pip3 install opencv-python camera4kivy
sudo apt-get install libatlas-base-dev
pip3 install opencv-python camera4kivy
Camera4Kivy depends on Buildozer 1.3.0 or later
pip3 install buildozer
The example includes a camera provider and a buildozer.spec.
This example is not available on iOS due to lack of an opencv recipe.