To reduce the effect of these systematic errors, numerous efforts have been made in the area of software calibration. One of the first Kinect calibration was done by [10] in the popular software RGB Demo, where the intrinsic and extrinsic parameters of the infrared (IR) and RGB camera were calibrated based on the OpenCV calibration. Burrus [10] also proposed an algorithm for converting the disparity values to depth measurements; however no calibration routine for deriving these conversion parameters was suggested. This approach can estimate the alignment between the infrared image and the RGB image if libfreenect is used and can improve the alignment if OpenNI is used for data capture.