Closed beniroquai closed 2 years ago
Thank you for your question. The CSI-2 driver supports the V4L2 interface for capturing frames. Any Python package that uses this interface should work, e.g. python-v4l2capture
Thanks for your quick reply! We managed to get the camera working with openCV+Gstreamer. You can find our working example here
I was wondering if there is a way to get the full 12-bit dynamic range using the Allied Vision Alvium 1800 C-158 camera using the above mentioned script. It seems that openCV cv.VideoCapture only receives 8Bit.
Thank you very much for your reply!
Nice to hear that you got it running! :) Currently, there is no 12-bit pixelformat available.
The following V4L2 pixel formats are supported on the NVidia boards with the Alvium camera: 8-bit Mono: V4L2_PIX_FMT_GREY (GREY) (mono cameras only) 8-bit Bayer: V4L2_PIX_FMT_SGRBG8 (color cameras only) 24-bit RGB: V4L2_PIX_FMT_XRGB32 (BX24) 24-bit BGR: V4L2_PIX_FMT_XBGR32 (XR24) 16-bit YUV: V4L2_PIX_FMT_VYUY
Thanks for the prompt answer!
Currently we call the camera using this: 'v4l2src device='+ mycamera+ ' extra-controls='+'"'+'c,exposure='+str( myexp_time)+',exposure_auto=1'+'"' +' ! video/x-raw, format=BGRx ! videoconvert ! appsink'
How could we change this line to get the 24bit or 16 bit output?
the cap.read()
is outputting a 3-channel RGB frame, where each channel
has the same information.
Any ideas how to solve this?
The format "BGRx" refers to the above-mentioned 24 bit BGR format using 8 bit per channel. At the moment, there is no pixel format with a bit depth greater than 8 available.
Thanks for your quick reply! We managed to get the camera working with openCV+Gstreamer. You can find our working example here
I was wondering if there is a way to get the full 12-bit dynamic range using the Allied Vision Alvium 1800 C-158 camera using the above mentioned script. It seems that openCV cv.VideoCapture only receives 8Bit.
Thank you very much for your reply!
HI @beniroquai Hope the things are going well. I tried to open the link to see the python based code for accessing frames from camera but the link is not working. I have agx XAVIER and trying to work on python with this camera. Can you please share the python bindings so that I can try. Please also share your image quality and frame rate experience as I want to use it on drone for aerial scene. So is that camera will be good enough to work with? will be very thankful to you .... Thanks
Hey, you can check the official Vimba Python which runs on the jetson. It's not straightforward to make it work with openCV, but I give a tutorial for the Jetson Nano here. Hope that helps.
Is it possible to access the frames from a python routine as it's done in picamera for the raspberry pi?