![]() ![]() The PL-VIO improves positioning accuracy, especially in uncontrolled outdoor scenarios. propose a tightly-coupled monocular VIO system, PL-VIO, by integrating more structured line features. However, the use of the RGB-D sensor does not apply to the outdoor traffic scene in direct sunlight for autonomous vehicles. ![]() VINS-RGBD is proposed to fuse RGB-D and IMU data for small ground rescue robots, which overcomes the problem of scale unobservability mentioned in. To address the above problems, there are some studies to integrate VIO algorithms with additional sources of information. Therefore, the trade-off between efficiency and accuracy in practical applications should be considered in order to choose an appropriate off-the-shelf VIO algorithm. Delmerico and Scaramuzza performed a thorough evaluation of open source VIO, suggesting that EKF-based algorithms display better efficient performance, whereas optimization-based algorithms provide higher accuracy. By far, VIO algorithms mainly consist of EKF-based methods, including MSF, ROVIO, MSCKF, Stereo-MSCKF, S-MSCKF, R-VIO and optimization-based methods including the Open Keyframe-based Visual-Inertial SLAM (OKVIS), VINS-Mono, PL-VIO and Basalt. For vehicle positioning, Ramezani and Khoshelham propose a stereo Multi-State Constraint Kalman Filter (Stereo-MSCKF) using MSCKF, which shows a promising potential to provide positioning with sub-meter accuracy in short periods of GNSS signal loss. While VO algorithms undergo a lack of robustness when confronted with motion blur, low texture scenes, illumination changes, sharp turns, partial or full occlusions and the influence of dynamic objects, visual-inertial odometry (VIO) algorithms can significantly improve the performance by fusing low-cost IMUs, which attract much attention in the field of micro air vehicles (MAVs). A detailed review of the VO methods can be found in the literature. On the other hand, visual odometry (VO), which adopts visual sensors, is gaining significant interest by the robotics community owing to the abundant perceptive information, small size and low cost. ![]() The positioning technology based on the Light Detection and Ranging (LIDAR) odometry has the advantages of algorithm maturity, high precision and real-time performance, but its application in vehicles is limited due to the high cost of LIDAR. However, the performance degrades quickly in GPS-denied environments. The fusion of the Global Navigation Satellite System (GNSS) and the Inertial Navigation System (INS) can provide high-accuracy global positioning. Therefore, it is necessary to fuse data from other sensors to improve positioning accuracy. The accumulated errors will result in a drift in position estimation gradually over time. However, it is sensitive to systematic and non-systematic errors. It exhibits the characteristic of high positioning accuracy in short-term operation. It uses the information from the proprioceptive sensors, including the drive motor encoder sensor and steering wheel angle sensor, to estimate the position relative to a starting location with the method of dead-reckoning, based on the kinematics model. ![]() Wheel odometry is a fundamental method in most autonomous vehicles, owing to low cost and high reliability. Providing a real-time accurate vehicle position is one of the key technologies to enhance the active safety of advanced driver assistance systems (ADASs) and achieve autonomous driving. This paper accompanies the source code for the robotics community. Both qualitative and quantitative experimental results evaluated under real-world datasets from an Ackermann steering vehicle lead to the following demonstration: ACK-MSCKF can significantly improve the pose estimation accuracy of S-MSCKF under the special motions of autonomous vehicles, and keep accurate and robust pose estimation available under different vehicle driving cycles and environmental conditions. This way, additional constraints from the Ackermann measurements are exploited to improve the pose estimation accuracy. In contrast with S-MSCKF, in which the inertial measurement unit (IMU) propagates the vehicle motion and then the propagation is corrected by stereo visual measurements, we successively update the propagation with Ackermann error state measurements and visual measurements after the process model and state augmentation. To address this problem, a tightly-coupled Ackermann visual-inertial odometry (ACK-MSCKF) is proposed to fuse Ackermann error state measurements and the Stereo Multi-State Constraint Kalman Filter (S-MSCKF) with a tightly-coupled filter-based mechanism. Visual-Inertial Odometry (VIO) is subjected to additional unobservable directions under the special motions of ground vehicles, resulting in larger pose estimation errors. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |