Accuracy–Efficiency Trade-Off: Optimizing YOLOv8 for Structural Crack Detection

To address the accuracy–efficiency trade-off faced by deep learning models in structural crack detection, this paper proposes an optimized version of the YOLOv8 model. YOLO (You Only Look Once) is a real-time object detection algorithm known for its high speed and decent accuracy. To improve crack f...

Full description

Saved in:
Bibliographic Details
Main Authors: Jiahui Zhang, Zoia Vladimirovna Beliaeva, Yue Huang
Format: Article
Language:English
Published: MDPI AG 2025-06-01
Series:Sensors
Subjects:
Online Access:https://www.mdpi.com/1424-8220/25/13/3873
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849319907039117312
author Jiahui Zhang
Zoia Vladimirovna Beliaeva
Yue Huang
author_facet Jiahui Zhang
Zoia Vladimirovna Beliaeva
Yue Huang
author_sort Jiahui Zhang
collection DOAJ
description To address the accuracy–efficiency trade-off faced by deep learning models in structural crack detection, this paper proposes an optimized version of the YOLOv8 model. YOLO (You Only Look Once) is a real-time object detection algorithm known for its high speed and decent accuracy. To improve crack feature representation, the backbone is enhanced with the SimAM attention mechanism. A lightweight C3Ghost module reduces parameter count and computation, while a bidirectional multi-scale feature fusion structure replaces the standard neck to enhance efficiency. Experimental results show that the proposed model achieves a mean Average Precision (mAP) of 88.7% at 0.5 IoU and 69.4% for mAP@0.5:0.95, with 12.3% fewer Giga Floating Point Operations (GFlops), and faster inference. These improvements significantly enhance the detection of fine cracks while maintaining real-time performance, making it suitable for engineering scenarios.
format Article
id doaj-art-07e9c7be892344fc9dfdc01110e73155
institution Kabale University
issn 1424-8220
language English
publishDate 2025-06-01
publisher MDPI AG
record_format Article
series Sensors
spelling doaj-art-07e9c7be892344fc9dfdc01110e731552025-08-20T03:50:17ZengMDPI AGSensors1424-82202025-06-012513387310.3390/s25133873Accuracy–Efficiency Trade-Off: Optimizing YOLOv8 for Structural Crack DetectionJiahui Zhang0Zoia Vladimirovna Beliaeva1Yue Huang2Institute of Civil Engineering and Architecture, Ural Federal University, St. Mira19, 620002 Yekaterinburg, RussiaInstitute of Civil Engineering and Architecture, Ural Federal University, St. Mira19, 620002 Yekaterinburg, RussiaInstitute of Civil Engineering and Architecture, Ural Federal University, St. Mira19, 620002 Yekaterinburg, RussiaTo address the accuracy–efficiency trade-off faced by deep learning models in structural crack detection, this paper proposes an optimized version of the YOLOv8 model. YOLO (You Only Look Once) is a real-time object detection algorithm known for its high speed and decent accuracy. To improve crack feature representation, the backbone is enhanced with the SimAM attention mechanism. A lightweight C3Ghost module reduces parameter count and computation, while a bidirectional multi-scale feature fusion structure replaces the standard neck to enhance efficiency. Experimental results show that the proposed model achieves a mean Average Precision (mAP) of 88.7% at 0.5 IoU and 69.4% for mAP@0.5:0.95, with 12.3% fewer Giga Floating Point Operations (GFlops), and faster inference. These improvements significantly enhance the detection of fine cracks while maintaining real-time performance, making it suitable for engineering scenarios.https://www.mdpi.com/1424-8220/25/13/3873YOLOv8crack detectionattention mechanismSimAMC3Ghostfeature pyramid
spellingShingle Jiahui Zhang
Zoia Vladimirovna Beliaeva
Yue Huang
Accuracy–Efficiency Trade-Off: Optimizing YOLOv8 for Structural Crack Detection
Sensors
YOLOv8
crack detection
attention mechanism
SimAM
C3Ghost
feature pyramid
title Accuracy–Efficiency Trade-Off: Optimizing YOLOv8 for Structural Crack Detection
title_full Accuracy–Efficiency Trade-Off: Optimizing YOLOv8 for Structural Crack Detection
title_fullStr Accuracy–Efficiency Trade-Off: Optimizing YOLOv8 for Structural Crack Detection
title_full_unstemmed Accuracy–Efficiency Trade-Off: Optimizing YOLOv8 for Structural Crack Detection
title_short Accuracy–Efficiency Trade-Off: Optimizing YOLOv8 for Structural Crack Detection
title_sort accuracy efficiency trade off optimizing yolov8 for structural crack detection
topic YOLOv8
crack detection
attention mechanism
SimAM
C3Ghost
feature pyramid
url https://www.mdpi.com/1424-8220/25/13/3873
work_keys_str_mv AT jiahuizhang accuracyefficiencytradeoffoptimizingyolov8forstructuralcrackdetection
AT zoiavladimirovnabeliaeva accuracyefficiencytradeoffoptimizingyolov8forstructuralcrackdetection
AT yuehuang accuracyefficiencytradeoffoptimizingyolov8forstructuralcrackdetection