Cannot open camera by index

Question

I want to create a drowsines detector by using raspberry pi 4 model b, Picamera2 and alarm buzzer but I cannot open a camera preview after I run my python script. It keep show error to me, look like is the problem from index number, may I know how to check my Picamera2 index number?

below is the code:

[0:03:00.542245672] [2552]  INFO Camera camera_manager.cpp:284 libcamera v0.1.0+118-563cd78e
[0:03:00.571168714] [2562]  WARN RPiSdn sdn.cpp:39 Using legacy SDN tuning - please consider moving SDN inside rpi.denoise
[0:03:00.573974638] [2562]  INFO RPI vc4.cpp:444 Registered camera /base/soc/i2c0mux/i2c@1/ov5647@36 to Unicam device /dev/media0 and ISP device /dev/media1
[0:03:00.574054305] [2562]  INFO RPI pipeline_base.cpp:1142 Using configuration file '/usr/share/libcamera/pipeline/rpi/vc4/rpi_apps.yaml'
[0:03:00.578403164] [2552]  INFO Camera camera_manager.cpp:284 libcamera v0.1.0+118-563cd78e
[0:03:00.606613276] [2565]  WARN RPiSdn sdn.cpp:39 Using legacy SDN tuning - please consider moving SDN inside rpi.denoise
[0:03:00.608878048] [2565]  INFO RPI vc4.cpp:444 Registered camera /base/soc/i2c0mux/i2c@1/ov5647@36 to Unicam device /dev/media0 and ISP device /dev/media1
[0:03:00.608954253] [2565]  INFO RPI pipeline_base.cpp:1142 Using configuration file '/usr/share/libcamera/pipeline/rpi/vc4/rpi_apps.yaml'
[0:03:00.614843528] [2552]  INFO Camera camera.cpp:1183 configuring streams: (0) 640x480-XRGB8888 (1) 640x480-SGBRG10_CSI2P
[0:03:00.615262457] [2565]  INFO RPI vc4.cpp:608 Sensor: /base/soc/i2c0mux/i2c@1/ov5647@36 - Selected sensor format: 640x480-SGBRG10_1X10 - Selected unicam format: 640x480-pGAA
[ WARN:[email protected]] global cap_v4l.cpp:997 open VIDEOIO(V4L2:/dev/video0): can't open camera by index
[ WARN:[email protected]] global obsensor_stream_channel_v4l2.cpp:82 xioctl ioctl: fd=67, req=-2140645888
[ WARN:[email protected]] global obsensor_stream_channel_v4l2.cpp:138 queryUvcDeviceInfoList ioctl error return: 25
[ERROR:[email protected]] global obsensor_uvc_stream_channel.cpp:159 getStreamChannelGroup Camera index out of range

This is my python script:

import cv2
import numpy as np
import dlib
from imutils import face_utils
from picamera2 import Picamera2
import RPi.GPIO as GPIO

buzzer_pin = 16

picam2 = Picamera2()
picam2.configure(picam2.create_video_configuration(main={"format": 'XRGB8888', "size": (640, 480)}))
picam2.start()

GPIO.setmode(GPIO.BOARD)
GPIO.setwarnings(False)
GPIO.setup(buzzer_pin, GPIO.OUT)
GPIO.output(buzzer_pin, GPIO.LOW)

cap = cv2.VideoCapture(0)

detector = dlib.get_frontal_face_detector()
predictor = dlib.shape_predictor("shape_predictor_68_face_landmarks.dat")

sleep = 0
drowsy = 0
active = 0
status = ""
color = (0, 0, 0)


def compute(ptA, ptB):
    dist = np.linalg.norm(ptA - ptB)
    return dist


def blinked(a, b, c, d, e, f):
    up = compute(b, d) + compute(c, e)
    down = compute(a, f)
    ratio = up/(2.0*down)

    if(ratio > 0.25):
        return 2
    elif(ratio > 0.21 and ratio <= 0.25):
        return 1
    else:
        return 0


while True:
    frame = vs.read()
    frame = imutils.resize(frame, width=450)
    gray = cv2.cvtColor(im, cv2.COLOR_BGR2GRAY)

    faces = detector(gray)
    face_frame = frame.copy()

    for face in faces:
        x1 = face.left()
        y1 = face.top()
        x2 = face.right()
        y2 = face.bottom()

        
        cv2.rectangle(face_frame, (x1, y1), (x2, y2), (0, 255, 0), 2)

        landmarks = predictor(gray, face)
        landmarks = face_utils.shape_to_np(landmarks)

        left_blink = blinked(landmarks[36], landmarks[37],
                             landmarks[38], landmarks[41], landmarks[40], landmarks[39])
        right_blink = blinked(landmarks[42], landmarks[43],
                              landmarks[44], landmarks[47], landmarks[46], landmarks[45])


        if(left_blink == 0 or right_blink == 0):
            sleep += 1
            drowsy = 0
            active = 0
            if(sleep > 6):
                status = "SLEEPING !!!"
                color = (255, 0, 0)
                GPIO.output(buzzer_pin, GPIO.HIGH)

        elif(left_blink == 1 or right_blink == 1):
            sleep = 0
            active = 0
            drowsy += 1
            if(drowsy > 6):
                status = "Drowsy !"
                color = (0, 0, 255)
                GPIO.output(buzzer_pin, GPIO.HIGH)

        else:
            drowsy = 0
            sleep = 0
            active += 1
            if(active > 6):
                status = "Active :)"
                color = (0, 255, 0)
                GPIO.output(buzzer_pin, GPIO.LOW)

        cv2.putText(frame, status, (100, 100),
                    cv2.FONT_HERSHEY_SIMPLEX, 1.2, color, 3)

        for n in range(0, 68):
            (x, y) = landmarks[n]
            cv2.circle(face_frame, (x, y), 1, (255, 255, 255), -1)

    cv2.imshow("Frame", frame)
    cv2.imshow("Result of detector", face_frame)
    key = cv2.waitKey(1)
    if key == 27:
        break

I tried different index number already, but the result is still same. I cannot figure out what is the index number of Picamera2. I wish to get some help from you guys, thanks in advance

Answer 1

If you are using Picamera2 lib for Raspberry Pi camera access, Unlike the traditional OpenCV camera access, the Picamera library doesn't use an index for the camera. Instead, it directly interfaces with the Raspberry Pi camera.

So in order to fix the error, first remove the

cv2.CaptureVideo(0)

and replace the frame = vs.read() by

while True:
    frame = picam2.capture()
    frame = imutils.resize(frame, width=450)
    gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)

This should fix the errors. Adjust the code accordingly.