Accuracy improvements of consumer-grade 3D sensors for robotic applications

Abstract

Recent advances in development of low-cost 3D sensors, such as Microsoft Kinect, bring attractive opportunities to robot system integrators. The accuracy provided by such sensors is generally unsatisfactory for many robotic applications, but it may be improved through calibration. This paper presents a calibration case study that is based on the sensor calibration procedure involving only a use of a simple checkerboard. It is shown that the calibration enables improving sensor accuracy 3 to 5 times, depending on the anticipated use of the sensor. Additionally, results obtained using different levels of complexity of calibration models reveal that depth measurement correction is an important component of calibration as it may reduce by 50% the errors in sensor reading.