Dataset Information

Towards Accurate Ground Plane Normal Estimation from Ego-Motion.

ABSTRACT: In this paper, we introduce a novel approach for ground plane normal estimation of wheeled vehicles. In practice, the ground plane is dynamically changed due to braking and unstable road surface. As a result, the vehicle pose, especially the pitch angle, is oscillating from subtle to obvious. Thus, estimating ground plane normal is meaningful since it can be encoded to improve the robustness of various autonomous driving tasks (e.g., 3D object detection, road surface reconstruction, and trajectory planning). Our proposed method only uses odometry as input and estimates accurate ground plane normal vectors in real time. Particularly, it fully utilizes the underlying connection between the ego pose odometry (ego-motion) and its nearby ground plane. Built on that, an Invariant Extended Kalman Filter (IEKF) is designed to estimate the normal vector in the sensor's coordinate. Thus, our proposed method is simple yet efficient and supports both camera- and inertial-based odometry algorithms. Its usability and the marked improvement of robustness are validated through multiple experiments on public datasets. For instance, we achieve state-of-the-art accuracy on KITTI dataset with the estimated vector error of 0.39°.

SUBMITTER: Zhang J

PROVIDER: S-EPMC9741436 | biostudies-literature | 2022 Dec

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Towards Accurate Ground Plane Normal Estimation from Ego-Motion.

Zhang Jiaxin J Sui Wei W Zhang Qian Q Chen Tao T Yang Cong C

Sensors (Basel, Switzerland) 20221201 23

In this paper, we introduce a novel approach for ground plane normal estimation of wheeled vehicles. In practice, the ground plane is dynamically changed due to braking and unstable road surface. As a result, the vehicle pose, especially the pitch angle, is oscillating from subtle to obvious. Thus, estimating ground plane normal is meaningful since it can be encoded to improve the robustness of various autonomous driving tasks (e.g., 3D object detection, road surface reconstruction, and trajecto ...[more]

PMID: 36502078

Similar Datasets

Project description:ObjectiveTo use image registration techniques to study the clavicular rotation of the shoulders in the sagittal plane.MethodsFrom 28 April 2019 to 20 May 2019, 13 healthy adults (7 males and 6 females) with no history of shoulder trauma surgery or chronic pain were recruited. Patients' ages ranged from 22 to 42 years, with a mean age of 26.5 years. Three-dimensional composite images of the sternum-clavicle-humerus were taken using CT images of upper limb movement in the sagittal plane in the 13 healthy adults. Four different postures were registered: (i) anatomical supine position; (ii) elbow joints lifted anteriorly in the supine position; (iii) posterosuperior hyperextension of the elbow joints in the prone position; and (iv) posteroinferior hyperextension of the elbow joints in the prone position. Image data from the humerus and clavicle in three of the postures were processed to calculate Euler angles for movements in the sagittal plane. SPSS 19 was used to perform statistical analyses.ResultsThere was no significant difference in the angles of change in the clavicle and humerus between the dominant and non-dominant sides under different movement patterns. For upper limb movements in the sagittal plane, the clavicle displayed different Euler angles in different postures. The rotation angle from the anatomical to the horizontal position was the smallest angle, with an average value of 7.1°, whereas the rotation angle from horizontal to posterosuperior hyperextension was the largest, with an average value of 37.2°. When the upper limb moved from anterior protraction to a posterosuperior extension, the intrinsic rotation angle of the clavicle reached its maximum, with an average value of 27.9°; when moved from the anatomical to the horizontal position, 9.1% of the sagittal rotation was executed by the clavicle. During rotation from the horizontal position to posterosuperior hyperextension and from the anatomical to posterior extension, the clavicle showed relatively higher weights at 29.5% and 37.0%, respectively.ConclusionOur results showed that dominance was not a consideration when studying clavicular rotation. Image registration is an effective method that can be used to study upper limb scapular movements. Through comparing and analyzing the data, two postures had relatively large changes in the rotation angle. This can help improve indicators of clavicular rotational function during physical examinations and postoperative functional evaluations.

Dataset Information

Towards Accurate Ground Plane Normal Estimation from Ego-Motion.

Publications

Towards Accurate Ground Plane Normal Estimation from Ego-Motion.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets