Pixel-Level Pavement Distress Assessment Using Instance Segmentation

Hugging Face Daily Papers Papers
pavement-assessment crack-detection instance-segmentation mask-rcnn computer-vision deep-learning

Summary

This paper presents a vision-based pavement distress analysis system using Mask R-CNN instance segmentation, achieving high precision and recall for crack detection and quantification on a custom dataset.

Automated pavement distress assessment requires more than image-level classification or coarse bounding box detection, demanding precise localization of thin, branching, and irregular cracks to achieve the geometric precision necessary for maintenance-relevant quantification. This paper presents a vision-based pavement distress analysis system based on Mask R-CNN instance segmentation and evaluates it on UWGB-StreetCrack, a custom field-collected roadway image dataset acquired with a vehicle-mounted smartphone and manually annotated with polygon labels for longitudinal cracks, transverse cracks, alligator cracks, and potholes. Five Detectron2-based Mask R-CNN backbone variants were considered under a consistent fine-tuning protocol. The best-performing model, Mask R-CNN with a ResNet-101 FPN backbone, achieved 84.23% precision, 90.04% recall, and an F1 score of 87.04% under the project-specific bounding-box matching protocol. The same model produced an aggregate predicted crack-area fraction of 2.164%, closely matching the 2.170% ground-truth crack-area fraction. To contextualize the segmentation system against a detector-oriented alternative, a CSPDarknet53-based YOLO detector was also adapted and retrained on the dataset, reaching 27.5% precision and 20.7% recall on the validation protocol. The results show that instance segmentation is a practical direction for field pavement imagery and aggregate crack-area estimation, while also exposing open challenges in annotation consistency, class imbalance, confounder rejection, and mask-level benchmarking.
Original Article
View Cached Full Text

Cached at: 05/26/26, 06:42 AM

Paper page - Pixel-Level Pavement Distress Assessment Using Instance Segmentation

Source: https://huggingface.co/papers/2605.26095

Abstract

A vision-based pavement distress analysis system using Mask R-CNN instance segmentation demonstrates superior performance for crack detection and quantification compared to object detection approaches, achieving high precision and recall metrics on a custom field-collected dataset.

Automated pavement distress assessment requires more than image-level classification or coarse bounding box detection, demanding precise localization of thin, branching, and irregular cracks to achieve the geometric precision necessary for maintenance-relevant quantification. This paper presents a vision-based pavement distress analysis system based onMask R-CNNinstance segmentationand evaluates it on UWGB-StreetCrack, a custom field-collected roadway image dataset acquired with a vehicle-mounted smartphone and manually annotated withpolygon labelsfor longitudinal cracks, transverse cracks, alligator cracks, and potholes. FiveDetectron2-basedMask R-CNNbackbone variants were considered under a consistent fine-tuning protocol. The best-performing model,Mask R-CNNwith aResNet-101 FPNbackbone, achieved 84.23% precision, 90.04% recall, and anF1 scoreof 87.04% under the project-specificbounding-box matching protocol. The same model produced an aggregate predictedcrack-area fractionof 2.164%, closely matching the 2.170% ground-truthcrack-area fraction. To contextualize the segmentation system against a detector-oriented alternative, aCSPDarknet53-basedYOLOdetector was also adapted and retrained on the dataset, reaching 27.5% precision and 20.7% recall on the validation protocol. The results show thatinstance segmentationis a practical direction for field pavement imagery and aggregate crack-area estimation, while also exposing open challenges in annotation consistency, class imbalance, confounder rejection, and mask-level benchmarking.

View arXiv pageView PDFAdd to collection

Get this paper in your agent:

hf papers read 2605\.26095

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2605.26095 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2605.26095 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2605.26095 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles