A Unified Detection Pipeline for Robust Object Detection in Fisheye-Based Traffic Surveillance

2025-10-22

AI Summary

Fisheye images suffer from severe radial distortion and non-uniform spatial resolution, causing significant performance degradation of standard object detectors—particularly in peripheral image regions—thus limiting their applicability in traffic monitoring. To address this, we propose a unified distortion-aware detection framework comprising three core components: (1) distortion-aware preprocessing via geometric rectification, (2) multi-scale feature alignment to ensure consistent representation across distorted regions, and (3) adaptive post-processing for refined localization and classification. Furthermore, we introduce a novel ensemble mechanism that synergistically integrates multiple state-of-the-art detectors, enhancing robustness against distortion-induced artifacts. The framework substantially improves detection accuracy for small and multi-scale objects near image boundaries. Evaluated on the 2025 AI City Challenge Track 4 (62 participating teams), our method achieves an F1-score of 0.6366, ranking eighth overall—demonstrating both effectiveness and practical viability for real-world fisheye-based traffic surveillance.

Technology Category

• Computer Vision: Object Detection & Categorization
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Search and Optimization: Distributed Search

Application Category

• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Responsible Web: Measurement, analysis, and circumvention of Web censorship

Abstract

Fisheye cameras offer an efficient solution for wide-area traffic surveillance by capturing large fields of view from a single vantage point. However, the strong radial distortion and nonuniform resolution inherent in fisheye imagery introduce substantial challenges for standard object detectors, particularly near image boundaries where object appearance is severely degraded. In this work, we present a detection framework designed to operate robustly under these conditions. Our approach employs a simple yet effective pre and post processing pipeline that enhances detection consistency across the image, especially in regions affected by severe distortion. We train several state-of-the-art detection models on the fisheye traffic imagery and combine their outputs through an ensemble strategy to improve overall detection accuracy. Our method achieves an F1 score of0.6366 on the 2025 AI City Challenge Track 4, placing 8thoverall out of 62 teams. These results demonstrate the effectiveness of our framework in addressing issues inherent to fisheye imagery.

Problem

• Addressing object detection challenges in fisheye camera imagery
• Handling radial distortion and nonuniform resolution in fisheye images
• Improving detection accuracy near distorted image boundaries

Innovation

• Pre and post processing pipeline for fisheye distortion
• Ensemble strategy combining multiple detection models
• Training detectors specifically on fisheye traffic imagery