The Accuracy of the Microsoft Kinect V2 Sensor for Human Gait Analysis. A Different Approach for Comparison with the Ground Truth

Abstract

Several studies have examined the accuracy of the Kinect V2 sensor during gait analysis. Usually the data retrieved by the Kinect V2 sensor are compared with the ground truth of certified systems using a Euclidean comparison. Due to the Kinect V2 sensor latency, the application of a uniform temporal alignment is not adequate to compare the signals. On that basis, the purpose of this study was to explore the abilities of the dynamic time warping (DTW) algorithm to compensate for sensor latency (3 samples or 90 ms) and develop a proper accuracy estimation. During the experimental stage, six iterations were performed using the a dual Kinect V2 system. The walking tests were developed at a self-selected speed. The sensor accuracy for Euclidean matching was consistent with that reported in previous studies. After latency compensation, the sensor accuracy demonstrated considerably lower error rates for all joints. This demonstrated that the accuracy was underestimated due to the use of inappropriate comparison techniques. On the contrary, DTW is a potential method that compensates for the sensor latency, and works sufficiently in comparison with certified systems.

Evaluation of lower extremity gait analysis using Kinect V2® tracking system

“…Regarding the hardware and software, the mechanism of posture recognition is based on projecting a pulsed infrared pattern and simultaneously measuring the reflection with an infrared sensor and a red/green/black (RGB) sensor. Kinect measures the time of flight (TOF) for each pattern and then matches the data to the RGB image [ 16 ]. Our preliminary study showed that the detection of heel strike judged from the position of the ankle was not precisely recognized because infrared was blocked in the case where the forefoot covers the ankle.…”

“…Thus, to recognize exact ankle detection, we adopted a hip-knee cyclogram in this study [ 12 ]. As for the number of Kinect installed for gait analysis, multiple Kinect systems have been used for gait analysis [ 8 , 16 ] ( Table 5 ). However, the use of multiple Kinect systems requires relatively high cost and technical skills to integrate each data.…”

Supervised learning for improving the accuracy of robot-mounted 3D camera applied to human gait analysis

“…Latency compensation allows for the suppression of time delays due to data processing time in 3D cameras. In our previous study [18], the latency of the Kinect V2 sensor was measured and corrected using the Dynamic Time Warping (DTW) algorithm. The latency compensation process improves the estimation of joint angles during gait analysis.…”

Effect of vestibular rehabilitation therapy on spatio-temporal gait parameters in elderly patients with post-stroke hemineglect

“…All patients in the experimental and control groups had their spatio-temporal gait parameters evaluated at baseline, four and eight weeks after the intervention. Spatio-temporal gait parameters were identified by using a markless 3D motion analysis sensor, the Microsoft Kinect V2 [20]. Each patient was asked to walk toward the Kinect sensor (Microsoft Kinect™ V2) maintaining a comfortable speed without being instructed to correct their gait.…”