2.5D Vehicle Odometry Estimation (2111.08398v1)

Abstract: It is well understood that in ADAS applications, a good estimate of the pose of the vehicle is required. This paper proposes a metaphorically named 2.5D odometry, whereby the planar odometry derived from the yaw rate sensor and four wheel speed sensors is augmented by a linear model of suspension. While the core of the planar odometry is a yaw rate model that is already understood in the literature, we augment this by fitting a quadratic to the incoming signals, enabling interpolation, extrapolation, and a finer integration of the vehicle position. We show, by experimental results with a DGPS/IMU reference, that this model provides highly accurate odometry estimates, compared with existing methods. Utilising sensors that return the change in height of vehicle reference points with changing suspension configurations, we define a planar model of the vehicle suspension, thus augmenting the odometry model. We present an experimental framework and evaluations criteria by which the goodness of the odometry is evaluated and compared with existing methods. This odometry model has been designed to support low-speed surround-view camera systems that are well-known. Thus, we present some application results that show a performance boost for viewing and computer vision applications using the proposed odometry

• The paper proposes a novel 2.5D odometry method that augments yaw rate-based planar models with a suspension model capturing vehicle height variations.
• Experimental validation using DGPS and IMU ground truth demonstrates that the method achieves significantly higher odometry accuracy than traditional techniques.
• The approach enhances low-speed surround-view camera systems and computer vision applications, boosting ADAS performance through improved vehicle dynamics estimation.

The paper "2.5D Vehicle Odometry Estimation" focuses on enhancing the accuracy of vehicle odometry estimation, which is crucial for Advanced Driver Assistance Systems (ADAS). The authors propose a novel technique termed "2.5D odometry," which augments traditional planar odometry with additional data derived from vehicle suspension models. Here’s a detailed overview:

Core Concept and Methodology

1. Planar Odometry Basis:
   • The foundational component of the proposed 2.5D odometry is a yaw rate-based planar odometry model. This model utilizes data from wheel speed sensors and yaw rate sensors, which is a technique well-established in the literature.
2. Augmentation through Suspension Modeling:
   • The innovation lies in augmenting this planar model with a suspension model that accounts for variations in vehicle height at different reference points. The suspension model is linear and is integrated into the odometry estimation process.
   • The augmented model involves fitting a quadratic polynomial to incoming sensor signals, which facilitates interpolation and extrapolation. This enhancement allows for finer integration of vehicle position data.

Experimental Validation

1. Reference System:
   • The accuracy of the proposed model was validated using a differential GPS (DGPS) and Inertial Measurement Unit (IMU) as reference systems. These provide ground truth data against which the odometry estimates can be compared.
2. Experimental Framework:
   • The paper details an experimental framework designed to evaluate the goodness of the odometry model. Key criteria and metrics for performance evaluation are outlined, although specific details are not provided in the summary.
3. Performance Comparison:
   • The results demonstrate that the 2.5D odometry model delivers highly accurate estimates compared to existing methods. This improvement is attributed to the integration of suspension data, which provides a more comprehensive representation of vehicle dynamics.

Application and Practical Utility

1. Low-speed Surround-view Camera Systems:
   • The model is particularly beneficial for low-speed maneuvering scenarios, where surround-view camera systems are often used. Accurate odometry is crucial for these systems to function effectively.
2. Computer Vision Applications:
   • The paper also presents application results showing a performance boost in computer vision tasks when utilizing the proposed odometry model. This showcases the practical benefits and potential for integration into existing ADAS technologies.

In summary, the "2.5D Vehicle Odometry Estimation" paper proposes an innovative method that enhances traditional odometry by integrating vehicle suspension data, resulting in more accurate estimations. Its practical implications are demonstrated in low-speed surround-view camera systems, highlighting its potential utility in enhancing ADAS functionalities. The experimental validation confirms the model’s superior performance compared to existing methods, establishing it as a valuable contribution to vehicle odometry research.