neuralet
neuralet copied to clipboard
Smart distancing on Jetson nano has very low fps
Hi, I solved the problem on my Jetson nano downgrade jetpack from 4.4 to 4.3 version and it is working, but I see thet in the video it should have 17 - 20 fps (I read this in the blue text) but seeing the video I suppose that fps is about 2-3, very slow. How is it possible? Where is the problem?
Thanks
Hi @grdsvt, Our video streaming method is a basic HTML method and this is why you see the output frame rate slower than the indicated fps in the blue text. The actual processing time is what the blue text indicates. This issue will be solved at #87 #67
Hi, I have a problem. I want to open camera on nano using gstreamer. So I changed to "VideoPath: v4l2src device=/dev/video0 ! video/x-raw, width=640, height=480 ! 'videoconvert ! appsink" in config-jetson.ini file. This gstreamer pipline is working in python file. And I changed "input_cap = cv.VideoCapture(video_uri, cv.CAP_GSTREAMER)" in libs/core.py. but I saw "failed to load video ". How can I solve it?
@mhsekhavat
Hi @wenzuolin-maker Was your problem solved ?
If yes then please help me
Hi, I have a problem. I want to open camera on nano using gstreamer. So I changed to "VideoPath: v4l2src device=/dev/video0 ! video/x-raw, width=640, height=480 ! 'videoconvert ! appsink" in config-jetson.ini file. This gstreamer pipline is working in python file. And I changed "input_cap = cv.VideoCapture(video_uri, cv.CAP_GSTREAMER)" in libs/core.py. but I saw "failed to load video ". How can I solve it?
more like it has been failed to load the default video sample as it has been removed, can you confirm the problem is with the VideoCapture for sure?
@codebugged What was your problem exactly?
Some work has been done around to support gstreamer input, but not well tested and merged yet. See #70