Jetson nano python program freezes on the line cv2.VideoWriter.write(frame)

Autonomous Machines Jetson & Embedded Systems Jetson Nano
,
1

Hello!
I use the latest Jetson Nano sd-card image for now, February 24 (downloaded it here: Jetson Download Center | NVIDIA Developer), camera IMX219-160, python 3.6.9, OpenCV library 4.1.1 (it was already installed in image). In my python program I use the next code to get images from camera and to write them to videofile:

def get_gstreamer_output_pipeline(video_file_name, path_to_video_folder):
    return ('appsrc ! '
            'video/x-raw, format=BGR ! '
            'queue ! '
            'videoconvert ! '
            'video/x-raw,format=RGBA ! '
            'nvvidconv ! '
            'omxh264enc ! '
            'h264parse ! '
            'qtmux ! '
            'filesink location=%s%s ' % (path_to_video_folder, video_file_name))

gstreamer_pipeline = (
    'nvarguscamerasrc ! '
    'video/x-raw(memory:NVMM), width=(int)1080, height=(int)720, format=(string)NV12, framerate=(fraction)30/1 ! '
    'nvvidconv ! '
    'video/x-raw, format=(string)BGRx ! '
    'videoconvert ! '
'video/x-raw, format=(string)BGR ! '
'appsink')
video_capturer = cv2.VideoCapture(gstreamer_pipeliPreformatted textne, cv2.CAP_GSTREAMER)



counter = 0
video_name_1 = str(counter)
video_name_2 = str(counter + 1)
gst_out = get_gstreamer_output_pipeline(video_name_1, path_to_video_folder)
output_1 = cv2.VideoWriter(gst_out, cv2.CAP_GSTREAMER, 0, 30.0, (int(1080), int(720)))
gst_out = get_gstreamer_output_pipeline(video_name_2, path_to_video_folder)
output_2 = cv2.VideoWriter(gst_out, cv2.CAP_GSTREAMER, 0, 30.0, (int(1080), int(720)))

current_date = datetime.datetime.now()
while (video_capturer.isOpened()):
    if (datetime.datetime.now() - current_date).total_seconds() > 60:
        output_1.release()
        output_2.release()

        current_date = datetime.datetime.now()

        counter += 2
        video_name_1 = str(counter)
        video_name_1 = str(counter + 1)

        gst_out = get_gstreamer_output_pipeline(video_name, path_to_video_folder)
        output_1 = cv2.VideoWriter(gst_out, cv2.CAP_GSTREAMER, 0, 30.0, (int(1080), int(720)))
        gst_out = get_gstreamer_output_pipeline(video_name_m, path_to_video_folder)
        output_2 = cv2.VideoWriter(gst_out, cv2.CAP_GSTREAMER, 0, 30.0, (int(1080), int(720)))

    _, frame = video_capturer.read()
    if (not _):
        break

    # ...Some frame processing...
    
    output_1.write(frame)
    output_2.write(image)

It seems that in some cases after a few hours of this program work, the program freezes when writing frame to a videofile. It just stops on the line output_1.write(frame) or output_2.write(image) and doesn’t move on. From the moment of the freeze, syslog is spammed with messages:

Feb 22 19:52:14 jetson-desktop nvargus-daemon[5375]: (Argus) Error OverFlow: Too many pending events, ignoring new events (in src/api/EventProviderImpl.cpp, function addEvent(), line 156)                      
Feb 22 19:52:35 jetson-desktop nvargus-daemon[5375]: message repeated 1255 times: [ (Argus) Error OverFlow: Too many pending events, ignoring new events (in src/api/EventProviderImpl.cpp, function addEvent(), line 156)]
Feb 22 19:52:35 jetson-desktop nvargus-daemon[5375]: (Argus) Error OverFlow: Too many pending events, ignoring new events (in src/api/EventProviderImpl.cpp, function addEvent(), line 156)
Feb 22 19:52:35 jetson-desktop nvargus-daemon[5375]: (Argus) Error OverFlow: Too many pending events, ignoring new events (in src/api/EventProviderImpl.cpp, function addEvent(), line 156) ...

Can you please help me to solve this problem? Thank you.

3

Hi,
Could you check if it stops due to insufficient freespace on storage? For writing to video files for hours, it may consume significant storage space.

4

I checked, at the time of the freeze, there is enough space on the sd card.

5

Hi,
Looks like you use omxh264enc plugin in the pipeline. We have deprecated omx plugins. Please try v4l2 plugins.

You can refer to this python sample:
Displaying to the screen with OpenCV and GStreamer - #9 by DaneLLL

6

Thank you. I will try it and report the results later.

8

Thank you. It seems that with the next pipeline video writing is much more stable and program does not freeze after several hours.

def get_gstreamer_output_pipeline(video_file_name, path_to_video_folder):
    return ('appsrc ! '
            'video/x-raw, format=BGR ! '
            'queue ! '
            'videoconvert ! '
            'video/x-raw,format=RGBA ! '
            'nvvidconv ! '
            'nvv4l2h264enc ! '
            'h264parse ! '
            'qtmux ! '
            'filesink location=%s%s ' % (path_to_video_folder, video_file_name))
9