Stream processed video with OpenCV on Jetson TX2

[berkantay.5]

Hello everyone,

I have been searching for streaming processed video over network. I cannot stream anything using gstreamer not only for processed frames.

Steps:

1- Capture input stream (done)
2- Process video (done)
3- Display the processed image (done)
4- Stream processed image (NOT DONE)

I need help on 4th step. I tried to do it with opencv videowriter.

stream.py

import cv2

cap = cv2.VideoCapture("/home/nvidia/Downloads/s1.mp4")

out = cv2.VideoWriter("appsrc ! video/x-raw, format=BGR ! queue ! videoconvert ! video/x-raw, format=BGRx ! nvvidconv ! omxh264enc ! video/x-h264, stream-format=byte-stream ! h264parse ! rtph264pay pt=96 config-interval=1 ! udpsink host=127.0.0.1 port=5001", cv2.CAP_GSTREAMER, 0, 25.0, (1920,1080))

while cap.isOpened():
    ret, frame = cap.read()
    if ret:
        if out.isOpened():
            out.write(frame)
            print('writing frame')
        if cv2.waitKey(1) & 0xFF == ord('q'):
            break
    else:
        break

# Release everything if job is finished
cap.release()
out.release()

command for displaying the stream

gst-launch-1.0 -e udpsrc port=50001 ! application/x-rtp, encoding-name=H264, payload=96 ! queue ! rtph264depay ! h264parse ! nvv4l2decoder ! nvvidconv ! video/x-raw, format=BGRx ! videoconvert ! xvimagesink

I can not view the stream.

[DaneLLL]

Hi,
Probably the port is not set correctly. It should be 5001 instead of 50001. Might fail due to the typo.

[berkantay.5]

Tried that but didn’t work either.

[DaneLLL]

Hi,
We try to run the sample on JP4.4.1(r32.4.4) and it works well:

import sys
import cv2

def read_cam():
    cap = cv2.VideoCapture("filesrc location=/home/nvidia/sample_1080p_h264.mp4 ! qtdemux ! h264parse ! nvv4l2decoder ! nvvidconv ! video/x-raw, format=BGRx ! videoconvert ! video/x-raw,format=BGR ! appsink  ")

    w = cap.get(cv2.CAP_PROP_FRAME_WIDTH)
    h = cap.get(cv2.CAP_PROP_FRAME_HEIGHT)
    fps = cap.get(cv2.CAP_PROP_FPS)
    print('Src opened, %dx%d @ %d fps' % (w, h, fps))

    gst_out = "appsrc ! video/x-raw, format=BGR ! queue ! videoconvert ! video/x-raw,format=BGRx ! nvvidconv ! nvv4l2h264enc ! h264parse ! rtph264pay pt=96 config-interval=1 ! udpsink host=127.0.0.1 port=5001 "
    out = cv2.VideoWriter(gst_out, cv2.CAP_GSTREAMER, 0, float(fps), (int(w), int(h)))
    if not out.isOpened():
        print("Failed to open output")
        exit()

    if cap.isOpened():
        while True:
            ret_val, img = cap.read();
            if not ret_val:
                break;
            out.write(img);
            cv2.waitKey(1)
    else:
     print "pipeline open failed"

    print("successfully exit")
    cap.release()
    out.release()


if __name__ == '__main__':
    read_cam()

You may refer to it and give it a try.

If you use JP4.4(r32.4.3), please remove h264parse for a try:

gst-launch-1.0 -e udpsrc port=5001 ! application/x-rtp, encoding-name=H264, payload=96 ! queue ! rtph264depay ! nvv4l2decoder ! nvvidconv ! video/x-raw, format=BGRx ! videoconvert ! xvimagesink

[berkantay.5]

Hello @DaneLLL,

When I try to run above example that you shared with me all I see is a blank screen. Does not throws me any error. Since I use JP4.4 I tried to remove h264 parse command from the pipeline. Now I have failed to open output error.

[DaneLLL]

Hi,
Please check if you can run with

/opt/nvidia/deepstream/deepstream-5.0/samples/streams/sample_1080p_h264.mp4

Looks like the error is reported from qtdemux. Probably it is not h264 stream in s1.mp4. We try with sample_1080p_h264.mp4 and it runs fine.

[berkantay.5]

When I check the codec I see h264 for video that I want to stream.

[DaneLLL]

Hi,
We would suggest clarify if the failure is specific to decoding s1.mp4. Or you also fail in decoding sample_1080p_h264.mp4.

[berkantay.5]

Since I did not flashed deepstream during the installation on JetPack 4.4 I dont have that video. But I will try it.

[DaneLLL]

Hi,
You can try sample files in

Jellyfish Bitrate Test Files

Since it is mkv files, you would need to replace qtdemux with matroskademux.