Image acquisition with libargus

Hi,

I use IMX304m camera (MIPI camera), which is monochrome, encoded on 12 bit.
I need to implement the process below :

1- take x times the same image of an object ;
2- change the position of the object ;
1- take x times the same image of the object in its new position ;
2- change the position of the object ;
[…]
1- take x times the same image of the object in its new position ;
2- change the position of the object ;
1- take x times the same image of the object in its new position ;

To implement it, I would like to use libargus.
For all images, I need to extract pixels’ value from the raw pointer to an unsigned short . I didn’t find in the samples /usr/src/tegra_multimedia_api/samples how access to the raw pointer, and how access pixels. Maybe is there a process already implemented in libargus ?
Moreover, do you provide a sample close to my application ? I didn’t find anything.

I have an other question. I have a monochrome camera, and the sample write .jpg as rgb image. Why ? Notice that is RV12 format.

Thank you

Acquire the frame by PIXEL_FMT_RAW16, have a reference to below sample code.

Check the tegra_multimedia_api/argus/samples/cudaBayerDemosaic

Actually, I don’t want to debayer the image because it is monochrome sensor, I will lose details

If you just want to get the raw data you can use v4l2 framework without argus.

Hi,

Thank you for your answer. Do you provide sample with v4l2 to access raw data ?

Have a reference to the multimedia sample code. 12_camera_v4l2_cuda or v4l2cuda