- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hello, i have developed with realsense camera with 3 months.
I want to build accurate face model for myself. but I have found 2 problems:
1. when my face is close to the camera (10~20cm), the depth image is stable (collecting 10 times, vary little),but when my face is 30 cm or more far way from the camera, the depth image is very unstable , the point cloud looks not like my face, even worse ,not like a face, the nose is not in nose shape. and the shape vary largely time to time (I am not moving ,but the depth image vary due to the measurement is not stable).
2. when my face is close to the camera (10~20cm), the reconstructed face shape is distorted since too close.
I want to ask: How can I build my accurate 3D face shape with this camera.
I have tried kinect2, it is very stable when I am 30-80 cm far away, but it do not work at 10~20 cm.
guys, help me!
Link Copied
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I am using world coordinates.
How to obtain a stable depth image??
I know there must be pixel unstable.
but if the pixels is measured for more times, the depth value should not vary too large ( as in my project, it vary 3~5 mm at 30cm far away)
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
You can try changing the IVCam parameters...
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
thanks ,but which parameters are realated to the depth measurement? there is no comment in SDK document.
would you pls tell me more details?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content

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