2 Replies Latest reply on Jan 24, 2018 10:54 PM by isaac_champ

How to extract the blob data, segmentation image, while using SDK 2.0

isaac_champ Jan 22, 2018 8:32 AM

In SDK 2016 R3, we can use QuerySegmentationImage get segmentation image data on SR300.

Since we decide to switch to D435 and we also need to upgrade the SDK to 2.0.

However, the SDK 2.0 does not have this function and we can not find any sample codes to implement it.

Do anyone know how to get segmentation image by using SDK 2.0? Or when will the intel add this feature into SDK 2.0?

1. Re: How to extract the blob data, segmentation image, while using SDK 2.0

Jan 23, 2018 4:20 PM (in response to isaac_champ)

This message was posted on behalf of Intel Corporation

hHello Isaac_champ,

It is possible to do segmentation using SDK 2.0--refer to the align example at https://github.com/IntelRealSense/librealsense/tree/master/examples/align. This example is not distributed with the Windows installer, only on Ubuntu, because it depends on the IMGUI tool. But you can compile it on Windows if you get the tool or just extract what you need from it and use it in your own application.

For best segmentation results, we recommend configuring the depth camera for
“High Density” + post-processing + hole-filling

See the measure example for how to use hole-filling: https://github.com/IntelRealSense/librealsense/blob/ba01147d65db16fdf4da36a3e718fe81c8421034/examples/measure/readme.md.

Regards,
Jesus
Intel Customer Support
Like Show 0 Likes(0)

Actions
2. Re: How to extract the blob data, segmentation image, while using SDK 2.0

isaac_champ Jan 24, 2018 10:54 PM (in response to isaac_champ)

Ok, thanks!
We will try this.
Like Show 0 Likes(0)

Actions