Align and merge zdfs (or pcds)

[adrianxsalazar]

Hi everyone,

I would like to create a point cloud of an entire object, similar to what is shown here (https://support.zivid.com/en/latest/reference-articles/settings/vision-engine/omni-engine.html). To achieve this, I am capturing images with the Zivid 2+ M60 from different points of view. Subsequently, I obtain several ZDF point clouds (in PLY format as well). Currently, I am attempting to align and merge these point clouds from different viewpoints into a single, large, and detailed point cloud. However, I have not been able to find anything in the documentation to assist with this. The closest resource I have found involves using several cameras (https://support.zivid.com/en/latest/academy/camera/multiple-zivid-cameras.html). Is there any method for merging point clouds that you would recommend, either with the Zivid SDK or without it?

Thanks in advance.

Kind regards,

[eskaur]

Hi @adrianxsalazar !

First of all, the point cloud you link to here (https://support.zivid.com/en/latest/reference-articles/settings/vision-engine/omni-engine.html) is from a single Zivid capture. It is not stitched from multiple captures at different angles.

Anyway, if you want to stitch together multiple captures we have some help for you in the "Multi-camera calibration" API: https://github.com/zivid/zivid-python/blob/master/modules/zivid/_calibration/multi_camera.py

Here's the appropriate section in our Knowledge Base: https://support.zivid.com/en/latest/academy/applications/multi-camera-calibration.html

Here's a C++ tutorial (which should be quite equivalent to how it's done in Python): https://github.com/zivid/zivid-cpp-samples/blob/master/source/Applications/Advanced/MultiCamera/MultiCameraTutorial.md

Since you don't have multiple cameras, but rather a single camera that you move around, you will need to keep the Zivid calibration board in the scene and in view from all angles.