Depth analysis of kinect v2 sensor in different mediums

From the last few years, RGB-D cameras are widely used by researchers in various fields. Their reasonable cost and the ability to estimate distances at a high frame rate have made these sensors recommendable for applications in gaming accessories, robotics, computer vision, etc. In addition to color, these sensors also provide depth information. Aspects like the stability, accuracy and reliability of depth-sensing cameras like Kinect v2 must also be considered before using the device for applications like that of 3D space modelling. In this paper, an analysis of the error in the depth measurement as well as calculation of Depth Entropy given by Kinect v2 sensor in different mediums viz. air, glass and water has been done. We have validated our findings using the theories of optics. The findings from error analysis are used to make an error compensation model which can correct depth at each pixel of the image. The error analysis and error compensation model proposed herewith will help in improving the accuracy of present and future depth sensing devices.