MATLAB_KINECT_ Get image and physical data
dvid = VideoInput (‘Kinect’, 1) Define the video input object
ekinect has two sensor channels,
1 is the color RGB sensor,
深2 is deep sensor
cSrc = Getselectsource (VID) View Source Equipment specific attribute
vPreview (vid) Preview ClosepreView (VID) Close preview
src.EnableBodyTracking = ‘on’
相 Deep camera has a special attribute value Enablebodytracking,
明 indicates the tracking status. When setting to ON, it returns physical data.
认为 Deficiency OFF
tStart (VID2) Start the second video input object (depth flow).
[frame, ts, metaData] = getdata(vid2);
 Use GetData to get the body tracking data on the deep camera
 Function Return content:
aframe frame
 size is 512 × 424
 Monochrome 13bit format (Mono13)
16 -bit unsigned integer (UNIT16) data format
mTime Stamps timestamp
tMetadata metadata
IsBodyTracked
[1×6 logical]
 indicates whether to track the position of the six bodies.
1 indicates the tracking subject, 0 means not tracking.
BodyTrackingID
[1×6 double]
 means the body tracking ID.
ColorJointIndices
[25x2x6 double]
 2 (X, Y) coordinates of 25 joints in color images
 6 possible physical data
DepthJointIndices
[25x2x6 double]
 The 25 joints (x, y) coordinates of the depth images
 6 possible physical data
BodyIndexFrame
[424×512 double]
 means which pixels are tracked and which pixels are not.
 Use this metadata to obtain segmentation.
HandLeftState
[1×6 double]
 Used to identify the possible gesture status of the left hand of the body
0 means unknown
1 indicates not tracking
2 Open
3 close
4LASSO grip
HandRightState
 and handleftState
HandLeftConfidence
[1×6 double]
 Tracking Confidence
0 means low
1 means high
HandRightConfidence
 and handleftconfident
JointTrackingStates
[25×6 double]
 Used to identify the trace status of the joint
0 indicates not tracking
1 means inferred (inferred)
2 means tracking
JointPositions
[25x3x6 double]
 indicates the position coordinates of the 25 joints of the 6 body in the three -dimensional space
 Obtain sub -data in Metadata through index
metaData.IsBodyTracked
metaData.JointPositions(:,:,1)
imagesc(metaDataDepth.BodyIndexFrame)
 The physical data of the segmentation through the image display
colormap(jet)