qt.qpa.xcb: could not connect to display when using yolov4-custom-functions-CodePudding

I am using https://github.com/theAIGuysCode/yolov4-custom-functions this repository. I want to crop and save pictures while running on webcam. But when I run the following code:

python detect_video.py --weights ./checkpoints/yolov4-416 --size 416 --model yolov4 --video ./data/video/bobinmi.mp4 --output ./detections/results.avi --crop

I can just detect the first frame of the video. Then it gives error and not detect. What should I do?

I'm using CUDA 10.1, Python 3.7 and Tensorflow 2.3. The error is same in CPU and GPU working environments.

The error message is:

FPS: 1.50
qt.qpa.xcb: could not connect to display 
qt.qpa.plugin: Could not load the Qt platform plugin "xcb" in "/usr/local/lib/python3.7/dist-packages/cv2/qt/plugins" even though it was found.
This application failed to start because no Qt platform plugin could be initialized. Reinstalling the application may fix this problem.

Available platform plugins are: xcb.

Fatal Python error: Aborted

Current thread 0x00007fbeed09b780 (most recent call first):
  File "detect_video.py", line 165 in main
  File "/usr/local/lib/python3.7/dist-packages/absl/app.py", line 254 in _run_main
  File "/usr/local/lib/python3.7/dist-packages/absl/app.py", line 308 in run
  File "detect_video.py", line 178 in <module>

CodePudding user response：

The error message is only caused in Colab at line 165, that is with the following instruction:

cv2.namedWindow("result", cv2.WINDOW_AUTOSIZE)

From the documentation this function is used to create a window. The reason is that Colab does not have a screen and some cv2 functions need a screen to perform the operations. Creating a windows is definitely one of them.

To fix the error in Colab should be enough to install a virtual screen. The code (from this stack answer):

!apt-get install -y xvfb # Install X Virtual Frame Buffer
import os
os.system('Xvfb :1 -screen 0 1600x1200x16  &')    # create virtual display with size 1600x1200 and 16 bit color. Color can be changed to 24 or 8
os.environ['DISPLAY']=':1.0'    # tell X clients to use our virtual DISPLAY :1.0