Robust Intrinsic and Extrinsic Calibration of RGB-D Cameras

Abstract

Color-depth cameras (RGB-D cameras) have become the primary sensors in most robotics systems, from service robotics to industrial robotics applications. Typical consumergrade RGB-D cameras are provided with a coarse intrinsic and extrinsic calibration that generally does not meet the accuracy requirements needed by many robotics applications (e.g., highly accurate 3D environment reconstruction and mapping, high precision object recognition and localization, . . . ). In this paper, we propose a human-friendly, reliable and accurate calibration framework that enables to easily estimate both the intrinsic and extrinsic parameters of a general color-depth sensor couple. Our approach is based on a novel two components error model. This model unifies the error sources of RGB-D pairs based on different technologies, such as structured-light 3D cameras and time-of-flight cameras. Our method provides some important advantages compared to other state-of-the-art systems: it is general (i.e., well suited for different types of sensors), based on an easy and stable calibration protocol, provides a greater calibration accuracy, and has been implemented within the ROS robotics framework. We report detailed experimental validations and performance comparisons to support our statements.

Multi-Cue-Based Circle Detection and Its Application to Robust Extrinsic Calibration of RGB-D Cameras

“…Many researchers have developed methods for measuring the systematic bias existing in the depths acquired with ToF or SL 3D cameras [28,29,30,31]. Because the bias depends on various factors such as pixel location, IR intensity, and measured distance, it is difficult to model the bias without an external distance measuring device.…”

“…Because the bias depends on various factors such as pixel location, IR intensity, and measured distance, it is difficult to model the bias without an external distance measuring device. Thanks to Zhang’s calibration method [8], a color camera is ready to play the role [28,29,30,31]. Once the color camera is calibrated by using a planar checkerboard [8], it is possible to estimate the plane parameters of the checkerboard in an image.…”

Intrinsic Calibration of Depth Cameras for Mobile Robots Using a Radial Laser Scanner

“…Both approaches employ planar surfaces for depth compensation, tendency that still holds in recent works. An example of this is the work by Basso et al [19], which proposed a calibration method based on the observation of a planar pattern with a regular camera, while the extrinsic calibration is more a "side effect".…”

“…In order to evaluate the undistortion performance, we follow a similar approach as described in [19]. Since the local distortion errors deform the reconstructed 3D structure, the evaluation method consists of fitting a plane to the point cloud acquired while observing a wall, and then computing the Root Mean Square (RMS) perpendicular distance to the extracted plane for each point belonging to the planar surface.…”

Research on Robot Control Technology Based on Vision Localization

“…It is the basis for subsequent coordinate system transformation and determination of any target workpiece [Tan and Sun (2018)]. The camera coordinates [Xc, Yc, Zc] can be obtained from any point [Xw, Yw, Zw] in the world coordinate system space by the transformation of the camera's external parameter matrix [Basso, Menegatti and Pretto (2018)]. Therefore, M2 can be changed to , and the transformation process is like formula 3.…”