Camera (Python)

Import

import doly_camera

Minimal working example

Below is the module’s example.py as a starting point.

import cv2
import doly_camera

def main():
    print("Initializing camera")

    cam = doly_camera.PiCamera()

    cam.options.photo_width = 3280
    cam.options.photo_height = 2464
    cam.options.video_width = 1280
    cam.options.video_height = 960
    cam.options.framerate = 15
    cam.options.verbose = False

    # -----------------------------
    # Single frame capture example
    # -----------------------------
    if cam.start_photo():
        frame = cam.capture_photo()  # numpy array or None
        if frame is None:
            print("[error] Camera read buffer failed at capture!")
            return -3

        cv2.imwrite("capture.jpg", frame)
        print("Frame saved!")
        cam.stop_photo()

    # -----------------------------
    # Video frame capture example
    #
    # This example uses an OpenCV imshow() window to display captured frames. (X11 forwarding over SSH)
    # However, you have several options to view the camera output.
    # Here are a few examples:
    # 1) Stream frames over the network (probably fast and best option, requires addditional coding)
    #       inside your capture loop:
    #       std::vector<uchar> jpeg;
    #       cv::imencode(".jpg", frame, jpeg); (optional: reduce JPEG quality for small data)
    #       -> send jpeg over socket / HTTP
    # 2) X11 forwarding over SSH (simple, but slow)
    #
    #
    # This example uses an OpenCV imshow() window to display captured frames (X11 forwarding over SSH).
    #
    # Note:
    # - The image window may NOT be displayed when running with root privileges
    #   (this can cause GTK/X11 permission errors).
    # - X11 forwarding is required to display the window over SSH.
    # - Make sure your SSH client supports X11 forwarding.
    #
    # To test whether X11 forwarding is working correctly:
    #   sudo apt install -y x11-apps
    #   xclock
    # If X11 is working, a clock window should appear.
    #
    # Also verify that DISPLAY is set:
    #   echo $DISPLAY
    # It should show something like "localhost:11.0"
    # -----------------------------
    if cam.start_video():
        print("Video window is live! Press ESC to quit!")

        cv2.namedWindow("Camera", cv2.WINDOW_NORMAL)
        cv2.resizeWindow("Camera", 640, 480)

        time_out_ms = 1500
        total_frame = 300 #test frame count

        while total_frame > 0:
            total_frame -= 1

            frame = cam.get_video_frame(time_out_ms)  # numpy array or None
            if frame is not None: 
                # resize frame for fast X11 transfer
                resized = cv2.resize(frame, (640, 480), interpolation=cv2.INTER_AREA)         
                cv2.imshow("Camera", resized)
                k = cv2.waitKey(1) & 0xFF
                if k == 27:  # ESC
                    break
            else:
                print("[warn] getVideoFrame() timeout/fail")

        cam.stop_video()

    cv2.destroyAllWindows()
    return 0

if __name__ == "__main__":
    raise SystemExit(main())

Common tasks

• Initialize
• Capture frame(s)
• Process frame
• Dispose / cleanup

Notes

• Platform: Raspberry Pi OS
• Python: 3.11
• Installed on robot: Yes (preinstalled)

API reference

See: API reference