How to set D435's depth effective range

[cxinl]

Hi

Recently, I use d435 to reconstruct 3D face model, I set the depth image to 640x480, and I set peple's face 80cm to 200cm, the model reconstructed

I foud that it can't see the eyes area, the nose can be reconstruct. Does there are some API that can set the effective reange of depth, and soto make

depth resolution more accurate.Thanks

[MartyG]

Ideally depth resolution of 848x400 or greater should be used, otherwise the accuracy of the results can reduce. I remember somebody found though that if a lower resolution such as 640x480 was used, like you are doing, it enabled the camera to be placed closer to objects when scanning (i.e the minimum distance of the camera was reduced).

If you are using the RealSense Viewer to do your scanning then you can set it to a 'High Accuracy' configuration, which gives better accuracy at the expense of a lower 'fill rate'.If you look for the drop down menu that is set by default to 'Custom' near the top of the RealSense Viewer's options side-panel (it is just below the camera name) then you can select 'High Accuracy' from the menu instead.

Also, you could try altering the scaling factor of the camera to get the best results.

[cxinl]

Thanks.

I used SDK2.0, and I create a project to do face scan, is there some APIs I can use or I must set a large depth image size? thanks

[MartyG]

3D model scanning is not as well supported on the 400 Series cameras at the moment as it was with older RealSense camera models, because the 400 Series is still new. So other than the methods that I suggested, there may not be another way to produce your model and increase its accuracy.

[cxinl]

Thanks