You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
While adding pattern projectors like in #20 affords disparity depth to work in absence of any ambient lighting, neural inference likely will not (as there will be little dots of illumination, not overall illumination).
So to allow 3D object localization (i.e. an object detector like YOLO or MobileNet-SSD fused with spatial information), having overall illumination (and not just pattern projection) is required.
Move to the how:
So there are 3 options for supporting the IR-capable camera modules:
Use IR-only camera modules (i.e. bandpass around IR).
Use Visible-light + IR-light capable camera modules (i.e. a single bandpass covering all visible + IR spectra)
Use a mechanical IR-cut filter that engages during high ambient light and disengages during low-light conditions when the IR illuminators are active.
We have tested both 1 and 2 with DepthAI (example images below) and found that 1 produces much sharper images when IR illumination is being used (this IR flashlight was used). See the quick test results below:
Examples image or IR-bandpass-only camera modules with IR-flashlight below:
Example of Visible-light + IR-light capable camera modules with IR-flashlight below:
We have not yet tested option 3, but it could be an interesting solution - but with the cost that usually such mechanical moving parts are a point of failure.
Move to the what:
Support IR illumination for object detection and disparity depth in total (ambient) darkness conditions.
The text was updated successfully, but these errors were encountered:
Please not that 1 above can be used directly on our designs, but keep in mind that the existing modules that come pre-mounted on our boards are actually glued down. So these are recommended for new designs, or fresh DepthAI boards for which cameras have not yet been installed. Otherwise, if you are trying to replace cameras on a DepthAI board with cameras already mounted, you'd have to do some surgery with isopropyl alcohol and a xacto-knife to catastrophically-remove the existing camera module (which might be a terrible idea).
And 2 is recommended for prototyping. It can be used with the BW1098FFC or the DM090FFC (here). It also has M12 lens mounting, so it allows swapping in a variety of lenses for a variety of filtering, field of view, etc. permutations.
Start with the
why
:While adding pattern projectors like in #20 affords disparity depth to work in absence of any ambient lighting, neural inference likely will not (as there will be little dots of illumination, not overall illumination).
So to allow 3D object localization (i.e. an object detector like YOLO or MobileNet-SSD fused with spatial information), having overall illumination (and not just pattern projection) is required.
Move to the
how
:So there are 3 options for supporting the IR-capable camera modules:
We have tested both
1
and2
with DepthAI (example images below) and found that1
produces much sharper images when IR illumination is being used (this IR flashlight was used). See the quick test results below:We have not yet tested option
3
, but it could be an interesting solution - but with the cost that usually such mechanical moving parts are a point of failure.Move to the
what
:Support IR illumination for object detection and disparity depth in total (ambient) darkness conditions.
The text was updated successfully, but these errors were encountered: