Vision-Based Lane Following and Traffic Sign Recognition for Resource-Constrained Autonomous Vehicles

TL;DR

This paper introduces a lightweight vision-based perception framework for resource-limited autonomous vehicles, combining efficient lane detection and traffic sign recognition with real-time performance and high accuracy.

Contribution

The paper develops a computationally efficient perception system using lightweight CNNs and novel lane tracking methods suitable for embedded autonomous vehicles.

Findings

Achieves real-time lane tracking with 3.16% maximum offset RMSE.

EfficientNet-B0 attains 98.77% offline accuracy and 90% real-time accuracy.

MobileNetV2 offers faster inference with slightly lower accuracy.

Abstract

Autonomous vehicles (AVs) rely on real-time perception systems to understand road environments and ensure safe navigation. However, implementing reliable perception algorithms on resource-constrained embedded platforms remains challenging due to limited computational resources. This paper presents a lightweight vision-based framework that integrates lane detection, lane tracking, and traffic sign recognition for embedded autonomous vehicles. A computationally efficient threshold-based lane segmentation method combined with perspective transformation and histogram-based curvature estimation is used for robust lane tracking under varying illumination conditions. A rule-based steering controller generates steering commands to maintain stable vehicle navigation. For traffic sign recognition, two lightweight convolutional neural networks (CNNs), EfficientNet-B0 and MobileNetV2, are evaluated using a custom dataset captured from the vehicle's onboard camera. Experimental results show that the system achieves real-time performance while maintaining accurate lane tracking with only 3.16% maximum offset RMSE. EfficientNet-B0 achieves a high offline classification accuracy of 98.77% on the test dataset, while achieving 90% accuracy during real-time on-device deployment, outperforming MobileNetV2 in both settings. MobileNetV2, however, offers slightly faster inference and lower computational cost. These results highlight the effectiveness of lightweight vision-based perception pipelines for resource-constrained autonomous driving applications.