{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,26]],"date-time":"2026-02-26T16:18:04Z","timestamp":1772122684586,"version":"3.50.1"},"posted":{"date-parts":[[2026]]},"group-title":"SSRN","reference-count":38,"publisher":"Elsevier BV","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"abstract":"<jats:p>Accurate measurement of railway surface defects is essential for infrastructure condition monitoring and maintenance planning. Conventional inspection methods rely on manual visual assessment or specialized sensing systems, which may suffer from subjectivity, limited scalability, or high operational cost. Image-based measurement systems provide a scalable alternative by enabling quantitative and repeatable assessment of defect extent directly from visual data. This study proposes FMSegNet, a vision-language based image measurement framework for quantitative assessment of railway surface defects using pixel-level segmentation. The framework integrates a domain-adapted multimodal transformer encoder with a synthetic multi-scale feature aggregation mechanism to accurately delineate defect regions across multiple spatial scales. The resulting segmentation masks function as quantitative measurements of defect spatial extent, enabling objective estimation of defect area, distribution, and severity indicators. Unlike benchmark-only model comparison studies, the proposed framework is designed as a measurement methodology for infrastructure condition assessment, providing reproducible and physically interpretable defect quantification. A composite Focal\u2013Dice loss function is employed to improve robustness under severe class imbalance conditions typical in defect measurement scenarios. Experimental evaluation on a real railway defect dataset demonstrates reliable measurement performance, achieving an Intersection over Union of 0.7153 and Dice coefficient of 0.7385, outperforming established segmentation methods. The results confirm that the proposed framework enables accurate, repeatable, and scalable image-based measurement of railway surface defects, supporting its applicability in infrastructure monitoring and engineering measurement applications.<\/jats:p>","DOI":"10.2139\/ssrn.6308025","type":"posted-content","created":{"date-parts":[[2026,2,26]],"date-time":"2026-02-26T15:42:37Z","timestamp":1772120557000},"source":"Crossref","is-referenced-by-count":0,"title":["A Vision-Language Based Image Measurement Framework for Quantitative Assessment of Railway Surface Defects"],"prefix":"10.2139","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-3802-6268","authenticated-orcid":true,"given":"Saeed","family":"MohammadKhah","sequence":"first","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6824-5221","authenticated-orcid":true,"given":"Joaquim","family":"Tinoco","sequence":"additional","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3758-3102","authenticated-orcid":true,"given":"Jes\u00fas","family":"Balado","sequence":"additional","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1536-2149","authenticated-orcid":true,"given":"Jos\u00e9 C.","family":"Matos","sequence":"additional","affiliation":[]}],"member":"78","reference":[{"key":"ref1","doi-asserted-by":"crossref","DOI":"10.1016\/j.compmedimag.2021.102026","article-title":"Unified focal loss: Generalising dice and cross entropy-based losses to handle class imbalanced medical image segmentation","volume":"95","author":"M Yeung","year":"2022","journal-title":"Computerized Medical Imaging and Graphics"},{"key":"ref2","article-title":"Mixed supervision for surface-defect detection: From weakly to fully supervised learning","volume":"129","author":"J Bo\ufffdi?","year":"2021","journal-title":"Computers in Industry"},{"issue":"12","key":"ref3","doi-asserted-by":"crossref","DOI":"10.3390\/s17122738","article-title":"On-tree mango fruit size estimation using RGB-D images","volume":"17","author":"Z Wang","year":"2017","journal-title":"Sensors"},{"issue":"1","key":"ref4","doi-asserted-by":"crossref","DOI":"10.1038\/s41598-024-84949-1","article-title":"Multi-scale feature fusion of deep convolutional neural networks on cancerous tumor detection and classification using biomedical images","volume":"15","author":"U Prakash","year":"2025","journal-title":"Scientific Reports"},{"issue":"1","key":"ref5","article-title":"Multi-scale sparse convolution and point convolution adaptive fusion point cloud semantic segmentation method","volume":"15","author":"Y Bi","year":"2025","journal-title":"Scientific Reports"},{"issue":"1","key":"ref6","article-title":"A novel lightweight multi-scale feature fusion segmentation algorithm for real-time cervical lesion screening","volume":"15","author":"J Yang","year":"2025","journal-title":"Scientific Reports"},{"key":"ref7","article-title":"Sca-cnn: Spatial and channel-wise attention in convolutional networks for image captioning","author":"L Chen","year":"2017","journal-title":"Proceedings of the IEEE conference on computer vision and pattern recognition"},{"key":"ref8","author":"A Dosovitskiy","year":"2020","journal-title":"An image is worth 16x16 words: Transformers for image recognition at scale"},{"key":"ref9","doi-asserted-by":"crossref","DOI":"10.1016\/j.engappai.2023.106669","article-title":"Semantic segmentation using vision transformers: A survey","volume":"126","author":"H Thisanke","year":"2023","journal-title":"Engineering Applications of Artificial Intelligence"},{"key":"ref10","article-title":"Your vit is secretly an image segmentation model","author":"T Kerssies","year":"2025","journal-title":"Proceedings of the Computer Vision and Pattern Recognition Conference"},{"key":"ref11","article-title":"Swin transformer: Hierarchical vision transformer using shifted windows","author":"Z Liu","year":"2021","journal-title":"Proceedings of the IEEE\/CVF international conference on computer vision"},{"key":"ref12","first-page":"12077","article-title":"SegFormer: Simple and efficient design for semantic segmentation with transformers","volume":"34","author":"E Xie","year":"2021","journal-title":"Advances in neural information processing systems"},{"key":"ref13","article-title":"Flava: A foundational language and vision alignment model","author":"A Singh","year":"2022","journal-title":"Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition"},{"key":"ref14","author":"T Meng","year":"2022","journal-title":"Revisiting multi-scale feature fusion for semantic segmentation"},{"issue":"1","key":"ref15","article-title":"The combined focal loss and dice loss function improves the segmentation of beta-sheets in medium-resolution cryo-electron-microscopy density maps","volume":"4","author":"Y Mu","year":"2024","journal-title":"Bioinformatics Advances"},{"key":"ref16","volume":"15","author":"M A Khan","year":"2023","journal-title":"Image processing techniques for concrete crack detection: A scientometrics literature review. Remote sensing"},{"key":"ref17","article-title":"Peridynamic-driven feature-enhanced Vision Transformer for predicting defects and heterogeneous materials locations: Applications of deep learning in inverse problems","volume":"151","author":"Y Li","year":"2025","journal-title":"Engineering Applications of Artificial Intelligence"},{"key":"ref18","article-title":"Rail defect detection with real time image processing technique","author":"C Ta?timur","year":"2016","journal-title":"IEEE"},{"key":"ref19","doi-asserted-by":"crossref","DOI":"10.1016\/j.rineng.2025.103972","article-title":"Steel surface defect detection and segmentation using deep neural networks","volume":"25","author":"S Ashrafi","year":"2025","journal-title":"Results in Engineering"},{"key":"ref20","doi-asserted-by":"crossref","DOI":"10.1016\/j.bspc.2024.105943","article-title":"MSFF: An automated multi-scale feature fusion deep learning model for spine fracture segmentation using MRI","volume":"91","author":"M U Saeed","year":"2024","journal-title":"Biomedical Signal Processing and Control"},{"key":"ref21","first-page":"1","article-title":"RSPrompter: Learning to prompt for remote sensing instance segmentation based on visual foundation model","volume":"62","author":"K Chen","year":"2024","journal-title":"IEEE Transactions on Geoscience and Remote Sensing"},{"key":"ref22","article-title":"Masked-attention mask transformer for universal image segmentation","author":"B Cheng","year":"2022","journal-title":"Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition"},{"issue":"4","key":"ref23","doi-asserted-by":"crossref","first-page":"1126","DOI":"10.1007\/s11263-023-01894-8","article-title":"Segvit v2: Exploring efficient and continual semantic segmentation with plain vision transformers","volume":"132","author":"B Zhang","year":"2024","journal-title":"International Journal of Computer Vision"},{"issue":"10","key":"ref24","doi-asserted-by":"crossref","first-page":"14004","DOI":"10.1109\/TITS.2024.3397509","article-title":"Superpixel-guided multi-type rail segmentation via contextual information aggregation","volume":"25","author":"X Ni","year":"2024","journal-title":"IEEE Transactions on Intelligent Transportation Systems"},{"key":"ref25","article-title":"Image segmentation using text and image prompts","author":"T L\ufffdddecke","year":"2022","journal-title":"Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition"},{"issue":"7","key":"ref26","doi-asserted-by":"crossref","first-page":"6427","DOI":"10.1109\/TITS.2024.3412170","article-title":"Railway intrusion detection based on machine vision: A survey, challenges, and perspectives","volume":"25","author":"Z Cao","year":"2024","journal-title":"IEEE Transactions on Intelligent Transportation Systems"},{"key":"ref27","doi-asserted-by":"crossref","DOI":"10.1016\/j.engappai.2023.105965","article-title":"MSRConvNet: classification of railway track defects using multiscale residual convolutional neural network","volume":"121","author":"H Acikgoz","year":"2023","journal-title":"Engineering Applications of Artificial Intelligence"},{"key":"ref28","article-title":"Uncertainty inspired domain adaptation network for rail surface defect segmentation","volume":"135","author":"Y Yan","year":"2024","journal-title":"Engineering Applications of Artificial Intelligence"},{"key":"ref29","article-title":"Rail surface defect detection using a transformer-based network","volume":"38","author":"F Guo","year":"2024","journal-title":"Journal of Industrial Information Integration"},{"key":"ref30","doi-asserted-by":"crossref","DOI":"10.1016\/j.aei.2025.104067","article-title":"AI-enabled defect detection in industrial products: A comprehensive survey, key insights and future research challenges","volume":"69","author":"L Nahar","year":"2026","journal-title":"Advanced Engineering Informatics"},{"key":"ref31","doi-asserted-by":"crossref","first-page":"43370","DOI":"10.1109\/ACCESS.2023.3271748","article-title":"Deep learning for automatic vision-based recognition of industrial surface defects: A survey","volume":"11","author":"M Prunella","year":"2023","journal-title":"IEEE Access"},{"key":"ref32","article-title":"Focal loss for dense object detection","author":"T.-Y Lin","year":"2017","journal-title":"Proceedings"},{"key":"ref33","doi-asserted-by":"crossref","DOI":"10.1109\/3DV.2016.79","article-title":"V-net: Fully convolutional neural networks for volumetric medical image segmentation","author":"F Milletari","year":"2016","journal-title":"2016 fourth international conference on 3D vision"},{"key":"ref34","article-title":"Loss odyssey in medical image segmentation","volume":"71","author":"J Ma","year":"2021","journal-title":"Medical image analysis"},{"key":"ref35","article-title":"Language-aware vision transformer for referring segmentation","author":"Z Yang","year":"2024","journal-title":"IEEE Transactions on Pattern Analysis and Machine Intelligence"},{"key":"ref36","article-title":"Mma: Multi-modal adapter for vision-language models","author":"L Yang","year":"2024","journal-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition"},{"key":"ref37","author":"O V Hees"},{"key":"ref38","article-title":"prEN 17397-1:2019 -Railway applications -Rail defects","author":"E C F S Cen)","year":"2019","journal-title":"Part 1: Rail defect management"}],"container-title":[],"original-title":[],"deposited":{"date-parts":[[2026,2,26]],"date-time":"2026-02-26T15:45:20Z","timestamp":1772120720000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.ssrn.com\/abstract=6308025"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026]]},"references-count":38,"URL":"https:\/\/doi.org\/10.2139\/ssrn.6308025","relation":{},"subject":[],"published":{"date-parts":[[2026]]},"subtype":"preprint"}}