[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-126229-en":3,"doc-seo-126229-105":31,"detail-sidebar-cat-0-en-105":92},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":21,"is_downloadable":21,"audit_status":21,"page_count":22,"language":23,"language_code":24,"site_id":25,"html_lang":24,"table_of_contents":26,"faqs":27,"seo_title":28,"seo_description":14,"update_tm":29,"read_time":30},126229,2336475104042,"Skyler","https://ap-avatar.wpscdn.com/avatar/22000c4c32af1715be0?x-image-process=image/resize,m_fixed,w_180,h_180&k=1786537525561427321",8,"Research & Report","Road condition classification from CCTV images using machine learning - Master’s thesis","Understanding and categorizing road conditions is crucial for driver safety and road maintenance. This thesis investigates practical image-based approaches for classifying road conditions using CCTV footage. Two tasks are studied: separating snowy from non-snowy conditions and distinguishing snowy, wet, and dry cases. Multiple machine learning methods are evaluated across several CCTV stations, including texture-based features, CNNs, vision transformers, road segmentation via an image segmentation model, and topological features from persistent homology.","Road condition classification from CCTV images using machine learning  \nMaster’s Thesis in Engineering Mathematics and Computational Science  \nLudvig Askbom  \nDEPARTMENT OF MATHEMATICAL SCIENCES  \nCHALMERS UNIVERSITY OF TECHNOLOGY Gothenburg, Sweden 2023  \n[www.chalmers.se](www.chalmers.se)  \nMaster’s thesis 2023  \nRoad condition classification from CCTV images  \nusing machine learning  \nLudvig Askbom  \nDepartment of Mathematical Sciences Chalmers University of Technology Gothenburg, Sweden 2023  \nRoad condition classification from CCTV images using machine learning Ludvig Askbom  \n© Ludvig Askbom, 2023 .  \nSupervisor: Daniel Persson, Department of Mathematical Sciences  \nExaminer: Daniel Persson, Department of Mathematical Sciences  \nSupervisors at Klimator AB: Gustaf Gulliksson and Jesper Landmér Pedersen  \nMaster’s Thesis 2023  \nDepartment of Mathematical Sciences Chalmers University of Technology SE-412 96 Gothenburg  \nTypeset in LATEX  \nPrinted by Chalmers Reproservice Gothenburg, Sweden 2023  \nRoad condition classification from CCTV images using machine learning  \nLudvig Askbom  \nDepartment of Mathematical Sciences Chalmers University of Technology  \nAbstract  \nUnderstanding and categorizing road conditions is crucial for driver safety and road maintenance. This research explores practical approaches to classify road conditions using images from CCTV stations. Two classification challenges are addressed: distinguishing between snowy and non-snowy conditions and between snowy, wet, and dry conditions.  \nThe thesis evaluates various machine learning methods for road condition classification on multiple CCTV stations, including established and novel approaches. Established methods involve feature extraction through texture analysis and finetuning convolutional neural networks and vision transformers. Novel contributions include training an image segmentation model for road segmentation and utilizing persistent homology for feature extraction. Notably, this thesis sets itself apart by separating data into training and test sets based on CCTV stations. This is important to evaluate the methods’ and models’ abilities to generalize to new CCTV stations.  \nThe best-performing model, a fine-tuned vision transformer, achieved accuracies of 87.9% and 75.3% for classifying snow/no snow and snow/wet/dry, respectively. These results underscore the complexity of the classification problem and highlight the effectiveness of deep learning models for large-scale road condition classification based on images.  \nKeywords: road condition, classification, machine learning, deep learning, feature extraction, vision transformer  \nAcknowledgements  \nI want to thank my family; without your love and support, I would not have made it through the toughest days. I want to thank my friends; without your friendship, I would have dropped out of Chalmers a long time ago. I want to thank my supervisors for helping make this thesis all that it could be. And finally, I want to thank Klimator AB for giving me the opportunity to investigate an interesting and meaningful problem.  \nLudvig Askbom, Gothenburg, November 2023  \nContents  \n1 Introduction 1  \n2 Machine Learning 5  \n2. 1 Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5  \n2.2 Conventional Classifiers . . . . . . . . . . . . . . . . . . . . . . . . . 6  \n2.2. 1 Naive Bayes Classifiers . . . . . . . . . . . . . . . . . . . . . . 6  \n2.2.2 Support Vector Machines . . . . . . . . . . . . . . . . . . . . . 7  \n2.2.3 Random Forests . . . . . . . . . . . . . . . . . . . . . . . . . . 8  \n2.3 Deep Learning for Classification Tasks ................. 10  \n2.3.1 Fully Connected Neural Networks ................ 11  \n2.3.2 Convolutional Neural Networks . . . . . . . . . . . . . . . . . 14  \n2.3.3 Vision Transformers . . . . . . . . . . . . . . . . . . . . . . . 16  \n2.3.4 Image Segmentation . . . . . . . . . . . . . . . . . . . . . . . 20  \n2.3.5 Fine-tuning . . . . . . . . . . . . . ","cbCaik7s931BUNdS","https://ap.wps.com/l/cbCaik7s931BUNdS","pdf",7855541,2,1,88,"English","en",105,"# 1 Introduction\n# 2 Machine Learning\n## 2.1 Classification\n## 2.2 Conventional Classifiers\n## 2.3 Deep Learning for Classification Tasks\n## 2.4 Model Evaluation and Selection\n# 3 Feature Extraction\n## 3.1 Texture Analysis\n## 3.2 Topological Data Analysis and Persistent Homology\n# 4 Methodology and Datasets\n## 4.1 Conventional Method\n## 4.2 Deep Learning Method\n## 4.3 Evaluation of Performance\n# 5 Results\n## 5.1 Conventional Method\n## 5.2 Deep Learning Method\n# 6 Discussion","[{\"question\":\"What road condition classification problems does the thesis address?\",\"answer\":\"It studies two classification challenges using CCTV images: snow vs. non-snow, and snow vs. wet vs. dry. The goal is to determine the road condition category from visual data.\"},{\"question\":\"Which machine learning approaches are evaluated?\",\"answer\":\"The thesis evaluates conventional models using texture-based feature extraction and deep models including fine-tuned convolutional neural networks and vision transformers. It also includes a road segmentation model and persistent homology for feature extraction.\"},{\"question\":\"How does the thesis improve evaluation for generalization across CCTV stations?\",\"answer\":\"Data are separated into training and test sets based on CCTV stations. This design assesses how well models generalize to new stations rather than only performing within the same camera source.\"}]","Road condition classification from CCTV images using machine learning - Master’s thesis | PDF",1785903949,222,{"code":4,"msg":32,"data":33},"ok",{"site_id":25,"language":24,"slug":34,"title":13,"keywords":35,"description":14,"schema_data":36,"social_meta":87,"head_meta":89,"extra_data":91,"updated_unix":29},"road-condition-classification-from-cctv-images-using-machine-learning-masters-thesis","",{"@graph":37,"@context":86},[38,54,69],{"@type":39,"itemListElement":40},"BreadcrumbList",[41,45,48,51],{"item":42,"name":43,"@type":44,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":46,"name":47,"@type":44,"position":20},"https://docshare.wps.com/document/","Document",{"item":49,"name":12,"@type":44,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":44,"position":53},"https://docshare.wps.com/document/road-condition-classification-from-cctv-images-using-machine-learning-masters-thesis/126229/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":24,"description":14,"dateModified":62,"datePublished":63,"encodingFormat":61,"isAccessibleForFree":64,"interactionStatistic":65},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":42,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-24","2026-08-05",true,{"@type":66,"interactionType":67,"userInteractionCount":20},"InteractionCounter",{"@type":68},"ViewAction",{"@type":70,"mainEntity":71},"FAQPage",[72,78,82],{"name":73,"@type":74,"acceptedAnswer":75},"What road condition classification problems does the thesis address?","Question",{"text":76,"@type":77},"It studies two classification challenges using CCTV images: snow vs. non-snow, and snow vs. wet vs. dry. The goal is to determine the road condition category from visual data.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"Which machine learning approaches are evaluated?",{"text":81,"@type":77},"The thesis evaluates conventional models using texture-based feature extraction and deep models including fine-tuned convolutional neural networks and vision transformers. It also includes a road segmentation model and persistent homology for feature extraction.",{"name":83,"@type":74,"acceptedAnswer":84},"How does the thesis improve evaluation for generalization across CCTV stations?",{"text":85,"@type":77},"Data are separated into training and test sets based on CCTV stations. This design assesses how well models generalize to new stations rather than only performing within the same camera source.","https://schema.org",{"og:url":52,"og:type":88,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":90,"canonical":52},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":93},[94,98,102,106,111,116,121,124,129,132,136],{"id":21,"doc_module":4,"doc_module_name":47,"category_name":95,"show_sort_weight":96,"slug":97},"Story & Novel",90,"story-novel",{"id":20,"doc_module":4,"doc_module_name":47,"category_name":99,"show_sort_weight":100,"slug":101},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":47,"category_name":103,"show_sort_weight":104,"slug":105},"Exam",70,"exam",{"id":107,"doc_module":4,"doc_module_name":47,"category_name":108,"show_sort_weight":109,"slug":110},5,"Comic",60,"comic",{"id":112,"doc_module":4,"doc_module_name":47,"category_name":113,"show_sort_weight":114,"slug":115},6,"Technology",50,"technology",{"id":117,"doc_module":4,"doc_module_name":47,"category_name":118,"show_sort_weight":119,"slug":120},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":47,"category_name":12,"show_sort_weight":122,"slug":123},30,"research-report",{"id":125,"doc_module":4,"doc_module_name":47,"category_name":126,"show_sort_weight":127,"slug":128},9,"Religion & Spirituality",20,"religion-spirituality",{"id":127,"doc_module":4,"doc_module_name":47,"category_name":130,"show_sort_weight":127,"slug":131},"World Cup","world-cup",{"id":133,"doc_module":4,"doc_module_name":47,"category_name":134,"show_sort_weight":133,"slug":135},10,"Lifestyle","lifestyle",{"id":137,"doc_module":4,"doc_module_name":47,"category_name":138,"show_sort_weight":107,"slug":139},19,"General","general"]