MEMROC: Multi-Eye to Mobile RObot Calibration

TL;DR

A comprehensive set of experiments proves MEMROC’s efficiency, surpassing existing state-of-the-art methods in terms of accuracy, robustness, and ease of use, surpassing existing state-of-the-art methods in terms of accuracy, robustness, and ease of use.

Abstract

This paper presents MEMROC (Multi-Eye to Mobile RObot Calibration), a novel motion-based calibration method that simplifies the process of accurately calibrating multiple cameras relative to a mobile robot's reference frame. MEMROC utilizes a known calibration pattern to facilitate accurate calibration with a lower number of images during the optimization process. Additionally, it leverages robust ground plane detection for comprehensive 6-DoF extrinsic calibration, overcoming a critical limitation of many existing methods that struggle to estimate the complete camera pose. The proposed method addresses the need for frequent recalibration in dynamic environments, where cameras may shift slightly or alter their positions due to daily usage, operational adjustments, or vibrations from mobile robot movements. MEMROC exhibits remarkable robustness to noisy odometry data, requiring minimal calibration input data. This combination makes it highly suitable for daily operations involving mobile robots. A comprehensive set of experiments on both synthetic and real data proves MEMROC's efficiency, surpassing existing state-of-the-art methods in terms of accuracy, robustness, and ease of use. To facilitate further research, we have made our code publicly available at https://github.com/davidea97/MEMROC.git.

Figures (6)

• Figure 1: MEMROC Overview. The system is structured into three primary components: the Ground Plane Detection module, responsible for identifying the ground plane; the Ground Plane Validation module, which ensures the ground plane's alignment with the mobile robot; and the Motion-Based Calibration Process, the concluding stage that calibrates the system.
• Figure 2: Formulation of the motion-based calibration method. A mobile robot $R$ equipped with a camera $C$ moves from pose $i$ to $i+1$ towards a calibration pattern $P$ in the scene. $A_i$ represents the odometry of the mobile robot with respect to the fixed frame $W$ (i.e., the robot starting position), while $B_i$ represents the camera pose with respect to the calibration pattern. X denotes the camera to robot transformation to be estimated.
• Figure 3: Set of 4 images representing 4 different scenarios where the dataset was collected. The first two on the left show two different real-world scenarios, the other two simulated scenarios.
• Figure 4: Impact of image quantity on calibration methods, where the lines represent the average errors and the shaded areas indicate the standard deviation, reflecting the variability and consistency of each method.
• Figure 5: Evaluation of motion-based calibration method with incremental noise in measurements provided by odometry, with $\lambda\in[0,10]$. The lines represent the average errors and the shaded areas indicate the standard deviation.