A Robust Pavement Mapping System Based on Normal-Constrained Stereo Visual Odometry

Abstract

Pavement condition is crucial for civil infrastructure maintenance. This task usually requires efficient road damage localization, which can be accomplished by the visual odometry system embedded in unmanned aerial vehicles (UAVs). However, the state-of-the-art visual odometry and mapping methods suffer from large drift under the degeneration of the scene structure. To alleviate this issue, we integrate normal constraints into the visual odometry process, which greatly helps to avoid large drift. By parameterizing the normal vector on the tangential plane, the normal factors are coupled with traditional reprojection factors in the pose optimization procedure. The experimental results demonstrate the effectiveness of the proposed system. The overall absolute trajectory error is improved by approximately 20%, which indicates that the estimated trajectory is much more accurate than that obtained using other state-of-the-art methods.

Pavement Inspection in Transport Infrastructures Using Unmanned Aerial Vehicles (UAVs)

“…Approaches that can be employed to evaluate image data collected using UAVs include the Geographic Information System (GIS) and the Convolutional Neural Network (CNN). CNN relies on an object detection approach, while GIS uses visual odometry for pavement mapping and evaluation [8,19,[28][29][30]. Both methods are used to detect and quantify pavement distress and aid in calculating pavement condition indexes.…”

“…Finally, the use of multiple UAVs is tested to reduce inspection time [48]. With regard to increasing efficiency, accuracy and reliability in UAV operations, including data collection and processing, [29] explored a robust pavement mapping system based on stereo visual odometry with normal constraints to support pavement inspection systems; [44] studied the automated processing of images captured by UAVs using Pix4D mapper 2.2 pro software, resulting in rapid generation of digital surface models and orthomosaics; and [52] proposed the integration of UAVs with onboard sensing and computing, together with a smart laser pointer equipped with built-in sensing and communication capabilities to accomplish a human-supervised system.…”