- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi,
I need help on how to get face width using SR300. Is this possible?
Regards,
Ethan
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
One way to do this would be to look at a pair of facial landmark points on either side of the face and measure their distance between their coordinates. You can read individual face points with an instruction called QueryPoint.
https://software.intel.com/sites/landingpage/realsense/camera-sdk/v1.1/documentation/html/index.html?querypoint_landmarksdata_pxcfacedata.html Intel® RealSense™ SDK 2016 R2 Documentation
The values of particular landmark points are on a face chart.
https://software.intel.com/sites/landingpage/realsense/camera-sdk/v1.1/documentation/html/index.html?doc_face_face_landmark_data.html Intel® RealSense™ SDK 2016 R2 Documentation
Looking at the chart, you could use the furthest-out points of 53 and 69 at the left and right ears to make your face width calculation.
It is possible that you may also need to use another instruction called ProjectCameraToDepth to convert the image coordinates of the face that the camera is seeing into real-world XYZ coordinates.
https://software.intel.com/sites/landingpage/realsense/camera-sdk/v1.1/documentation/html/projectcameratodepth_pxcprojection.html ProjectCameraToDepth
Link Copied
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
One way to do this would be to look at a pair of facial landmark points on either side of the face and measure their distance between their coordinates. You can read individual face points with an instruction called QueryPoint.
https://software.intel.com/sites/landingpage/realsense/camera-sdk/v1.1/documentation/html/index.html?querypoint_landmarksdata_pxcfacedata.html Intel® RealSense™ SDK 2016 R2 Documentation
The values of particular landmark points are on a face chart.
https://software.intel.com/sites/landingpage/realsense/camera-sdk/v1.1/documentation/html/index.html?doc_face_face_landmark_data.html Intel® RealSense™ SDK 2016 R2 Documentation
Looking at the chart, you could use the furthest-out points of 53 and 69 at the left and right ears to make your face width calculation.
It is possible that you may also need to use another instruction called ProjectCameraToDepth to convert the image coordinates of the face that the camera is seeing into real-world XYZ coordinates.
https://software.intel.com/sites/landingpage/realsense/camera-sdk/v1.1/documentation/html/projectcameratodepth_pxcprojection.html ProjectCameraToDepth
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page