On the performance of the Intel SR300 depth camera: metrological and critical characterization

In recent years, low-cost depth sensors played a key role in the development of many computer vision applications concerning different fields such as reverse engineering, robotics, construction industry and quality control. Each application, although, given the multiplicity of different problems, may need a specific device which works in a specific depth range (mid-to-long range, close range and nano/micro range). Nowadays, an incredible boost of the number of close range devices put on the market is witnessed and a number of different approaches for characterizing them have been developed by several researchers. Among close range devices, the RealSense™ SR300 promises to be among the best performing close range depth cameras. Given the potentiality of this new device, and since to date a deep investigation on its performances has not been assessed in scientific literature, the main aim of the present work is to characterize and to provide metrological considerations on the Intel® RealSense™ SR300 depth sensor. To this end, a methodology for close range depth cameras characterization based both on international standard and on best practices proposed in literature is here introduced. Such a methodology is straightforwardly applicable also to other close range devices. Additionally, the performance of such a device is critically compared against the performance of latest short-range devices, thus providing a useful guide, for researchers and practitioners, in an informed choice of the optimal device for their own application.