- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi guys,
I want only to use the r200 for scanning of feet, and wonder how I best go about it. Shall I use one of the present options (Face/head/body/object/full) or is there a convenient way to customize the scan function for scan of a foot?
Best regards
Per
Link Copied
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
You can use Face/head in the sample app to scan your foot. Thanks!
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Thanks, but isn't there some sort of face/head recognition pattern incorporated in those scan options?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I wasn't sure if you meant by the word "scanning" the scanning of a 3D model of the foot, or if you meant tracking the position of the foot and its joints.
If you meant tracking, the F200 camera can track the foot and knee joints with its hand tracking function (as it recognizes the foot as a hand and the toes as fingers). The R200 camera only has face tracking though, so foot tracking probably isn't viable with the R200, at least until Intel introduce hand tracking for it (they have suggested that F200-only features might eventually migrate to the R200 and vice versa.)
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi Marty,
I need to do a 3D scan of the foot, and I am in doubt about using the face or head scan options, in case they have recognition patterns for face or head, incorporated in the software. If they have such feature, can I get into their code, and change the pattern code?
Best regards,
Per
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Although the camera's library files are not currently open-source, Intel does support developers being able to create their own camera algorithms and use them commercially.
I don't have any experience of using the 3D model scanning features of RealSense, so it's difficult for me to comment in detail on the subject. Some googling revealed that a company called Volumental used the camera last year to scan feet to work out shoe sizes though.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi Per,
In the previous version of the SDK, head mode was best for scanning body parts other than the face (which only really worked on faces), but in the latest version, face mode is best for all body parts that I've tested on and head never seems to lock on to anything to start scanning.
So yes, face mode should work fine. Have you tried making foot scans and had problems?
James
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi James,
Thanks for the answer. I will try with Face, which is actually the only one I haven't tried. As you mention, the head option, never lock to the feet. That is why I thought they most have a "head" pattern to match with in the code.
Regards,
Per
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Yeah, the head option doesn't even lock on to heads any more - not sure what they've done to it, but hopefully it'll be fixed in the next version.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
The tracking of head tilting was removed from the SDK in the version before the current one because Intel felt it wasn't working as it should be. Perhaps that is related to lack of head lock-on in the scanning function.

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page