How to analyze IR (850nm) wavelength with the Intel RealSense D435 and IR-pass filter
In our actual project we have multiple IR lights and Intel D435 cameras. We need to capture the IR video stream with the RealSense cameras.
Our Region of Interest has a size of about 100 x 100 mm and is about 700 mm away from our camera. We want to highest possible resolution for the ROI.
We have multiple hardware IR-pass filters, which we can attach to the camera. We are using pyrealsense2 (python and openCV).
Do you have any suggestions on how to achieve this?
-
Hi Michael Harder It sounds as though you want to use the IR Pass filter to exclude visible light wavelengths but allow Near IR (NIR) wavelengths to pass through. Is that correct, please?
Intel's new RealSense D435f model adds such filters to the lenses of the D435 or you could potentially retrofit your existing D435 cameras externally with the same filter used on the D435f. The link below provides more information about this.
https://support.intelrealsense.com/hc/en-us/community/posts/7030571809299
The D435 will support an infrared resolution of 1280x720 at 30 FPS on a USB 3 connection.
Traditionally in terms of RealSense, a Region of Interest (ROI) refers to an auto-exposure region of interest, where a defined area of the screen has its pixels maintained at a defined intensity setpoint value. As you provide height, width and distance values, I wonder whether you are seeking to instead define a 3D bounding box and only do processing work on the coordinates within that box, like in the case at the link below.
https://github.com/IntelRealSense/librealsense/issues/4838#issuecomment-530404166
If it is an auto-exposure region of interest that you are seeking to implement then a Python code snippet for doing so is here:
https://github.com/IntelRealSense/librealsense/issues/2681#issuecomment-436391721
Please sign in to leave a comment.
Comments
1 comment