Experimental evaluation of Visual‐Inertial Odometry systems for arable farming

Abstract

The farming industry constantly seeks the automation of different processes involved in agricultural production, such as sowing, harvesting and weed control. The use of mobile autonomous robots to perform those tasks is of great interest. Arable lands present hard challenges for Simultaneous Localization and Mapping (SLAM) systems, key for mobile robotics, given the visual difficulty due to the highly repetitive scene and the crop leaves movement caused by the wind. In recent years, several Visual-Inertial Odometry (VIO) and SLAM systems have been developed.They have proved to be robust and capable of achieving high accuracy in indoor and outdoor urban environments. However, they were not properly assessed in agricultural fields. In this study we assess the most relevant state-of-the-art VIO systems in terms of accuracy and processing time on arable lands to better understand how they behave on these environments. In particular, the evaluation is carried out on a collection of sensor data recorded by our wheeled robot in a soybean field, which was publicly released as the Rosario data set. The evaluation shows that the highly repetitive appearance of the environment, the strong vibration produced by the rough terrain and the movement of the leaves caused by the wind, expose the limitations of the current state-of-the-art VIO and SLAM systems. We analyze the systems failures and highlight the observed drawbacks, including initialization failures, tracking loss and sensitivity to Inertial Measurement Unit saturation. Finally, we conclude that even though certain systems like ORB-SLAM3 and stereo Multi-State Constraint Kalman Filter show good results with respect to others, more improvements should be done to make them reliable in agricultural fields for certain applications such as soil tillage of crop rows and pesticide spraying.

GNSS‐stereo‐inertial SLAM for arable farming

“…This indicates that it is important not only to focus on global measurements, but also to have a robust visual-inertial fusion. In our case, we use ORB-SLAM3 as the underlying system, as a result of having analyzed the performance of different visual-inertial systems in previous research (Cremona et al, 2022). As a conclusion, in addition to a tight coupling of the sensor data, the robustness of the visual-inertial estimates are also relevant for practical implementations in agricultural applications.…”

“…This work presents a GNSS-stereo-inertial SLAM framework that fuses in a tightly-coupled manner the information from a stereo camera, an IMU and a conventional GNSS sensor. In order to report the most competitive results, we implement our GNSS factor on top of the ORB-SLAM3 framework, the top performer in the evaluation of (Cremona et al, 2022). As we are motivated by long-term autonomous navigation in arable farms, we present results in the Rosario Dataset and in-house sequences from an agricultural robot.…”

“…The proposed framework extends the Visual-Inertial SLAM system ORB-SLAM3 (Campos et al, 2021) with GNSS measurements. We built on top of ORB-SLAM3 since it has a fair performance in agricultural environments (Cremona et al, 2022). Our implementation is publicly released as open source to facilitate its use, extension and reproduction of the results by the robotics community.…”

“…From the results in (Cremona et al, 2022), ORB-SLAM3 presents an acceptable accuracy in arable lands for short camera trajectories, but long-term navigation is still challenging. The authors propose a novel loop closure algorithm to correct the drift.…”

“…Adding inertial measurements provides a slight improvement in the estimation. Nevertheless, as shown in (Cremona et al, 2022), state-of-the-art visual-inertial systems accumulate significant errors after navigating a few minutes on arable lands. Robust SLAM systems such as ORB-SLAM3 (Campos et al, 2021) can eliminate drift when revisiting already mapped places, but the so-called loop closing offers a poor performance on agricultural fields due to insufficiently discriminative visual appearances.…”