fariadiego on Wed, 29 Jun 2011 09:40:21


Does anyone know about the sensor model, how it is computed (how the point cloud is achieved, of course, the intrinsic parameters of both cameras, and extrinsic mapping between them have to be known usign calibration methods), or at least the error measurement (this is the most important). Does anyone know the range error (depth info) provided by the sensor? Inside the 2m of range of the sensor which is the accuracy/resolution?

Thank you in advance



ChocheR on Wed, 29 Jun 2011 11:44:12

Hi Diego,

Primesense is the developing Company of Kinect Technology. Here you have an datasheet with similar Sensor.


Todd Ferkingstad on Wed, 29 Jun 2011 21:50:32

FYI, the SDK clamps depth at .8 meters and 4 meters.

newricky on Thu, 25 Aug 2016 15:00:56

Yes Prime Sense was the technology used in the Kinect 1, although Canesta was bought by MS and is used in the Kinect 2. The Kinect 2 is more accurate. For the precise numbers you would need to ask MS. I worked with the Canesta Sensor technology when I developed applications for the Canesta sensor I created. The Prime Sense technology has certain application concerns, based on my testing of the hardware 7 years ago and those issues may have been improved. Again you would need to go to the present day source (MS) to get the precise advertised numbers.