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.

A Linear Approach for Depth and Colour Camera Calibration Using Hybrid Parameters

“…Many recent studies [8][9][10][13][14][15][16][17][18][19] proposed a depth measuring model and the corresponding calibration algorithms. Most of them are capable of calibrating depth and colour cameras at the same time.…”

“…Raposo et al [10] proposed several modifications of the optimization pipeline of [9] to improve the stability and the runtime. Follow-on studies [17][18][19] proposed the improved depth measuring models for depth camera. Kim et al [17] proposed a new depth correction method based on the analysis of depth measurement principle of Kinect.…”

“…Kim et al [17] proposed a new depth correction method based on the analysis of depth measurement principle of Kinect. A reformulated depth correction model was proposed in [18][19], which directly rectifies the sensor that provides depth data.…”

“…Our implementation is based on previous studies which improved calibration accuracy by introducing the specific intrinsic parameters for both depth and colour cameras [8][9][10] and the novel depth measurement models [17][18][19] . In this paper, we propose a matrix of hybrid parameters as an alternation to the specific intrinsic parameters and propose a corresponding calibration algorithm, which solves the hybrid parameters with linear equations.…”

Calibration of Kinect-type RGB-D sensors for robotic applications

“…First, it presents a calibration model suitable for software calibration of Kinect-type sensors. Second, it describes a two-step calibration procedure recently proposed by this author [5,6] for identification of the model parameters. The procedure is simple in the sense that it makes use of only a commonly adopted camera calibration tool (i.e., a checkerboard) and it does not require additional specific calibration objects or external measurement devices.…”