Jetson - 2 lane CSI camera IMX274 - setting resolution, exposure time

Hi Experts,

I am using 6 IMX274 cameras with Jetson TX2 using CSI interface. I am using nvgstcapture-1.0 to capture images from the sensor. The exposure time for the captured images is 0.03 seconds which should be reasonable to get high dynamic range images. But I am not able to get high dynamic range images unless my exposure time is high( 0.1 sec or higher).

My questions are:

  1. Is it the limitation of the camera sensor or nvgstcapture-1.0 application? If it’s problem with the nvgstcapture-1.0 application, what other options do I have ?

  2. I cannot set the resolution to highest possible resolution possible by imx275 sensor(i.e to 3864 x 2196) Is it limitation of using nvgstcapture-1.0 application? or is it limitation of using only 2 lanes for each camera ?

  3. How to create efficient camera pipelines when using 6 cameras? Which application or camera service should I be using?

Sridhar Gunnam

I think you can use argus API to implement your APP as your request.
You should be able to see a sample in the …/64_TX2/tegra_multimedia_api/argus/ after you flash and download the tools by run the

Thanks. I will give argus API a try and see if it solves camera configuration issue.


When I try to run the sample application, I get the following error. Can you let me know how to get these libraries? (I am using the Tegra R28 (release), REVISION: 1.0 )

tegra_multimedia_api/samples/09_camera_jpeg_capture$ make
Linking: camera_jpeg_capture
/usr/bin/ld: cannot find -lv4l2
/usr/bin/ld: cannot find -lcudart
/usr/bin/ld: cannot find -lnvinfer
/usr/bin/ld: cannot find -lnvparsers
collect2: error: ld returned 1 exit status
Makefile:60: recipe for target ‘camera_jpeg_capture’ failed
make: *** [camera_jpeg_capture] Error 1


Please check the readme

Just FYI, JetPack installs those somewhere under “/usr/local”, e.g., you might have “/usr/local/cuda-8.0/” for a CUDA 8 library.

If the camera is V4L2 compatible you could set exposure time through that interface.

sudo apt-get install v4l-utils
v4l2-ctrl --list-ctrls