Setting Up ESP32-Cam

This only works for the ESP32-Cam-MB, not the ESP32-Cam itself. The difference is that the ESP32-Cam-MB has a second layer & Micro-b port, as shown below.

Follow this tutorial to use the ESP32-cam-webserver library: Link

• The above uses this github library for Arduino: github
• Do not uncomment // #define URL_HOSTNAME "esp-cam" or // #define WIFI_AP_ENABLE
• If done correctly, navigating to http://192.168.1.168:81/view should open the camera view. Double-click to enter fullscreen mode.

Converting ESP32-Cam output to Webcam

The ESP32-Cam is an IP camera, and the code I used to detect ArUco markers requires a webcam output. To solve this I use OBS Studio & Virtual Cam to stream the ESP32-Cam's browser output as a virtual webcam.

• Install OBS Studio here: Link
• Install OBS Virtual Cam here: Link

Install Code

1. Install the code: git clone https://github.com/danielchandg/aruco-markers.git
2. Run opencv_generate_aruco/opencv_generate_aruco.py and follow the example usage at the top to generate ArUco markers

• Note that ArUco marker detection works best when ArUco markers have a wide white border on all 4 sides.

3. Run opencv-detect-aruco/detect_aruco_image.py and follow the example usage at the top to detect ArUco markers in an image
4. Run opencv-detect-aruco/detect_aruco_video.py to detect ArUco markers using webcam output, or opencv-detect-aruco/detect_aruco_video_guess_type.py to detect any ArUco markers and print the ArUco dictionary.

• Examples from detect_aruco_video.py: Video demonstration: Video