Related Experiment Video
Updated: Jul 9, 2026
11:34
High-resolution, High-speed, Three-dimensional Video Imaging with Digital Fringe Projection Techniques
Published on: December 3, 2013
SU-E-I-92: Accuracy Evaluation of Depth Data in Microsoft Kinect
K Kozono1,2,3, M Aoki1,2,3, M Ono1,2,3
1Department of Health Sciences, Graduate School of Medical Sciences, Kyushu University.
Medical Physics
|May 19, 2017
Summary
The Microsoft Kinect sensor offers accurate patient positioning in radiology and radiotherapy, achieving approximately 1 mm accuracy after calibration. This depth-sensing technology is suitable for clinical applications.
Area of Science:
- Medical Imaging
- Radiology
- Radiotherapy
Background:
- Real-time patient positioning is crucial in diagnostic radiology and radiotherapy.
- The Microsoft Kinect sensor presents a potential low-cost solution for this application.
Purpose of the Study:
- To evaluate the accuracy of Microsoft Kinect depth image data.
- To assess device-to-device variation of the Kinect sensor in simulated clinical settings.
Main Methods:
- Developed a computer program using OpenNI and OpenCV to measure quantitative distance data from the Kinect sensor.
- Calibrated distance data using the least squares method and evaluated variations across five Kinect sensors.
- Measured distances to a standardized object at various ranges (0-400 cm) from the sensor.
Main Results:
- Kinect sensor demonstrated an inability to measure distances below 50 cm.
- Calibrated Kinect sensors achieved a device-to-device variation error within 0.46% from 50 cm to 2 m.
- Maximum deviation after calibration was 1.1 mm (average 0.18 mm) between 50-150 cm.
Conclusions:
- The Microsoft Kinect sensor provides a distance accuracy of approximately 1 mm when properly calibrated.
- The sensor's accuracy and low variation make it suitable for patient positioning in radiology and radiotherapy.
