Abstract:Mobile Mapping System (MMS) simultaneously collects the Lidar points and video log images in a scenario with the laser profiler and digital camera. Besides the textural details of video log images, it also captures the 3D geometric shape of point cloud. It is widely used to survey the street view and roadside transportation infrastructure, such as traffic sign, guardrail, etc., in many transportation agencies. Although many literature on traffic sign detection are available, they only focus on either Lidar or imagery data of traffic sign. Based on the well-calibrated extrinsic parameters of MMS, 3D Lidar points are, the first time, incorporated into 2D video log images to enhance the detection of traffic sign both physically and visually. Based on the local elevation, the 3D pavement area is first located. Within a certain distance and height of the pavement, points of the overhead and roadside traffic signs can be obtained according to the setup specification of traffic signs in different transportation agencies. The 3D candidate planes of traffic signs are then fitted using the RANSAC plane-fitting of those points. By projecting the candidate planes onto the image, Regions of Interest (ROIs) of traffic signs are found physically with the geometric constraints between laser profiling and camera imaging. The Random forest learning of the visual color and shape features of traffic signs is adopted to validate the sign ROIs from the video log images. The sequential occurrence of a traffic sign among consecutive video log images are defined by the geometric constraint of the imaging geometry and GPS movement. Candidate ROIs are predicted in this temporal context to double-check the salient traffic sign among video log images. The proposed algorithm is tested on a diverse set of scenarios on the interstate highway G-4 near Beijing, China under varying lighting conditions and occlusions. Experimental results show the proposed algorithm enhances the rate of detecting traffic signs with the incorporation of the 3D planar constraint of their Lidar points. It is promising for the robust and large-scale survey of most transportation infrastructure with the application of MMS.

Driving by the Rules: A Benchmark for Integrating Traffic Sign Regulations into Vectorized HD Map

Integrating Visual and Range Data for Road Detection.

A Traffic Sign Discovery Driven System for Traffic Rule Updating.

Road Object Detection for HD Map: Full-element Survey, Analysis and Perspectives

OpenLane-V2: A Topology Reasoning Benchmark for Unified 3D HD Mapping

LiDAR-Video Driving Dataset: Learning Driving Policies Effectively

TRAFFIC CONTROL RECOGNITION WITH AN ATTENTION MECHANISM USING SPEED-PROFILE AND SATELLITE IMAGERY DATA

The Mapillary Traffic Sign Dataset for Detection and Classification on a Global Scale

MapVision: CVPR 2024 Autonomous Grand Challenge Mapless Driving Tech Report

Preliminary Establishment and Analysis of Automated Driving Benchmark in China

No driver, No Regulation? --Online Legal Driving Behavior Monitoring for Self-driving Vehicles

Traffic Sign Interpretation in Real Road Scene

Which Traffic Light Should You Look At? Automatically Associating Traffic Lights with Roads in High-Definition Map (industrial Paper).

Learning from Maps: Visual Common Sense for Autonomous Driving

Lidar-Incorporated Traffic Sign Detection from Video Log Images of Mobile Mapping System

LIDAR and vision-based real-time traffic sign detection and recognition algorithm for intelligent vehicle

Fine-grained Traffic Sign Detection and Matching Algorithm

Driving with Regulation: Interpretable Decision-Making for Autonomous Vehicles with Retrieval-Augmented Reasoning via LLM

Driving with Prior Maps: Unified Vector Prior Encoding for Autonomous Vehicle Mapping

Vision-based mapping of lane semantics and topology for intelligent vehicles

Understanding Drivers' Visual and Comprehension Loads in Traffic Violation Hotspots Leveraging Crowd-Based Driving Simulation