Low FPS with Python, OpenCV on Raspberry Pi

Question

I'm trying to do some image processing on a Raspberry Pi with Python and OpenCV. It works well so far except a low FPS rate. Even without any image processing and just with the code below I get only 10 FPS with 640x480 resolution. Is there a faster way to capture the video stream? Do I something wrong here?

import numpy as np
import cv2
import time
from picamera.array import PiRGBArray
from picamera import PiCamera

# initialize the camera and grab a reference to the raw camera capture
camera = PiCamera()
camera.resolution = (640, 480)
camera.framerate = 32
rawCapture = PiRGBArray(camera, size=(640, 480))
# allow the camera to warmup
time.sleep(0.1)
start = time.time()

for img in camera.capture_continuous(rawCapture, format="bgr", use_video_port=True):

    frame = img.array
    rawCapture.truncate(0)
    end = time.time()
    print 'fps:', int(round(1 / (end - start)))
    start = time.time()

Thank you so far.

regards

Answer 1

The hardware always produces YUV (I420) and conversion to BGR or RGB is done as an extra vector sw stage, thus reducing your frames per second.

I would suggest creating a thread dedicated solely to your IO pipeline, reducing latency and potentially increasing your fps, however I highly doubt you will be able to achieve the glorious 90fps (at 640x480) with the BGR model.

Check these two posts for a more detailed explanation: limited framerate picamera v2

https://raspberrypi.stackexchange.com/questions/22040/take-images-in-a-short-time-using-the-raspberry-pi-camera-module