Building Surface Defect and Damage Detection Method based on YOLOv8-act
Detecting surface defects and damage on buildings is crucial for ensuring structural safety and durability. Traditional manual inspection methods are inefficient and highly subjective, making them inadequate for large-scale, high-precision engineering demands. This paper proposes a building surface defect detection method based on YOLOv8-act. By replacing the original SILU activation function with the SELU activation function-which possesses self-normalizing properties-the model's ability to extract features from low-contrast, irregularly shaped defects is enhanced. Experimental results demonstrate that the YOLOv8-act model achieves an mAP50 of 0.641, outperforming other compared models. This approach maintains real-time processing and deployment convenience while improving detection accuracy, providing an effective solution for automated and precise detection of building surface defects.
References
- Choutri K, Lagha M, Meshoul S, et al. Fire detection and geo-localization using uav’s aerial images and yolo-based models[J]. Applied Sciences, 2023, 13(20): 11548.
- Singla R, Sharma S, Sharma S K. Infrared imaging for detection of defects in concrete structures[C]//IOP Conference Series: Materials Science and Engineering. IOP Publishing, 2023, 1289(1): 012064.
- Zhang Z, Zhong G, Ding P, et al. ELS-YOLO: efficient lightweight YOLO for steel surface defect detection[J]. Electronics, 2025, 14(19): 3877.
- Wang H, Zhao J. Research on Defect Detection on Steel Rails Based on Improved YOLO11n Algorithm[J]. Applied Sciences, 2026, 16(2): 842.
- Wei G, Wan F, Zhou W, et al. Bfd-Yolo: A Yolov7-based detection method for building façade defects[J]. Electronics, 2023, 12(17): 3612..
- Gao K, Chen L, Li Z, et al. Automated Identification and Analysis of Cracks and Damage in Historical Buildings Using Advanced YOLO-Based Machine Vision Technology[J]. Buildings, 2025, 15(15): 2675.
- Krizhevsky A, Sutskever I, Hinton G E. Imagenet classification with deep convolutional neural networks[J]. Advances in neural information processing systems, 2012, 25.
- Oquab M, Bottou L, Laptev I, et al. Learning and transferring mid-level image representations using convolutional neural networks[C]//Proceedings of the IEEE conference on computer vision and pattern recognition. 2014: 1717-1724..
- Hussain M. Yolov5, yolov8 and yolov10: The go-to detectors for real-time vision[J]. arXiv preprint arXiv:2407.02988, 2024.
- Hussain M. Yolov1 to v8: Unveiling each variant–a comprehensive review of yolo[J]. IEEE access, 2024, 12: 42816-42833.
- Klambauer G, Unterthiner T, Mayr A, et al. Self-normalizing neural networks[J]. Advances in neural information processing systems, 2017, 30.
- Lin X, Meng Y, Sun L, et al. Building Surface Defect Detection Based on Improved YOLOv8[J]. Buildings, 2025, 15(11): 1865.
- He K, Zhang X, Ren S, et al. Deep residual learning for image recognition[C]//Proceedings of the IEEE conference on computer vision and pattern recognition. 2016: 770-778.