Since the FOV (Field Of View) of Kinect V2 is horizontal 70 degrees, only the frontal sides of the objects can be scanned by a single Kinect sensor.

The ToF scanning method used in Kinect V2 measures the depth by calculating the time that the infrared light is emitted and then reentered the camera sensor. The second version of it, Kinect V2, adopts the Time of Flights (ToF) method instead of the Light Coding™ and produces a 512 × 424 depth map as well as a 1920 × 1080 full HD size color image. The first version of Kinect generates a color image of 640 × 480 resolution and a depth map of 320 × 240 with 30 frames per second. Kinect sensor developed for the gesture recognition for indoor video games has the IR sensor, which can be used to measure the depth and to construct 3D point clouds. Although the depth measurements through the IR sensor do not have the same accuracy and distance coverage compared to the laser scanners, its low price and easy availability are definitely the merits. Unlike the stereo camera, the IR sensor can be also used for the distance measurements even with no visible light. Alternatively, 3D point cloud can be obtained by measuring the depth from the sensor to the target object by using a stereo camera or an IR sensor. The high price of the LIDAR equipment, however, is a major obstacle in adopting the laser sensor for small-scale 3D object scanning applications. LIDAR is a 360° omnidirectional scanning device that can estimate the distance by measuring the reflected laser pulses from the target object and is widely used in autonomous vehicles. The performance of the proposed method is evaluated by constructing 3D point clouds for human bodies and by making geometric measurements for cylindrical testing objects. That is, a set of correspondence pairs in skeleton joints of human body detected by Kinect V2 sensors is used to determine the calibration matrices, then Iterative Closest Point (ICP) algorithm is adopted for finely tuning the calibration parameters. The basic idea of the proposed registration method is to exploit the skeleton information of the human body provided by the two Kinect V2 sensors. Therefore, it is suitable in portable camera setting environments with frequent relocations. Our approach aims at achieving a handy registration without the calibration-assisting devices such as the checker board. This paper proposes a registration method for two sets of point clouds obtained from dual Kinect V2 sensors, which are facing each other to capture omnidirectional 3D data of the objects located in between the two sensors.

Kinect random dot pattern