[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-127284-en":3,"doc-seo-127284-105":30,"detail-sidebar-cat-0-en-105":91},{"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":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},127284,2336475104736,"Quinn","https://ap-avatar.wpscdn.com/avatar/22000c4c5e0e5b17e70?x-image-process=image/resize,m_fixed,w_180,h_180&k=1786591360781797222",8,"Research & Report","Automated Hydrophobicity Assessment of Composite Insulators via Contour Analysis and Machine Learning","Environmental degradation from ultraviolet (UV) radiation, pollution, and moisture exposure reduces the hydrophobicity of composite insulators, undermining high-voltage system reliability. Conventional IEC 62073 spray inspections rely on operator judgement and can vary in field consistency. This study introduces an automated data-driven hydrophobicity classification approach using contour analysis and machine learning. Using 4,500 labelled images, six supervised models were trained and evaluated. Gradient Boosting and Random Forest achieved ~75% accuracy and AUC > 0.90, supporting reliable field-deployable monitoring and potential real-time grid integration.","Automated Hydrophobicity Assessment of Composite Insulators via Contour Analysis and  \nMachine Learning  \nS. Salisu Department of Electrical Engineering and Electronics  \nUniversity of Liverpool Liverpool, United Kingdom[salihu@liverpool.ac.uk](salihu@liverpool.ac.uk)  \nC. Zachariades Department of Electrical Engineering and Electronics  \nUniversity of Liverpool Liverpool, United Kingdom  \n[C.Zachariades@liverpool.ac.uk](C.Zachariades@liverpool.ac.uk)  \nAbstract-Environmental degradation such as ultraviolet (UV) radiation, pollution, and moisture exposure compromises the hydrophobicity of composite insulators, threatening the reliability of high-voltage systems. Traditional inspection methods, such asthe IEC 62073 spray method, are widely used but suffer from operator subjectivity and inconsistency in field applications. To address these limitations, this study proposes an automated, datadriven approach for classifying hydrophobicity states using contour analysis and machine learning (ML). A dataset of 4,500 labelled insulator images was used to train and evaluate six ML models. Among them, Gradient Boosting and Random Forest classifiers achieved approximately 75% accuracy and Area Under the Curve (AUC) scores greater than 0.90. This framework offers reliable, field-deployable hydrophobicity monitoring with potential for real-time grid integration.  \nI. INTRODUCTION  \nComposite insulators have gained popularity over the past few decades compared to glass or ceramic solutions due to their performance advantages especially in adverse weather conditions. This can be mainly attributed to the hydrophobic behaviour of their surface. This hydrophobicity helps prevent continuous water films, reducing leakage currents and the risk of flashovers [1-3] . However, exposure to UV radiation, pollutants, and humidity can degrade this property over time, increasing the likelihood of insulator failure [4-6] . Conventional methods like the spray method per IEC 62073 are prone to operator bias [8-10] .  \nThe widely used spray method, outlined in IEC 62073, offers a simple way to assess hydrophobicity but suffers from operator subjectivity [7] . To address this, researchers have explored alternatives including dynamic discharge testing, signal processing, and machine learning models. While deep learning methods like CNNs show high accuracy, they often require complex equipment or large datasets, limiting practical use [8,9] . This work introduces a lightweight, scalable method that combines image processing statistical features extraction, and machine learning (ML) for automated hydrophobicity classification (AHC) . To determine the best classifier for the AHC model six candidates were evaluated.  \nII. METHODOLOGY  \nA. Dataset  \nThe study utilized a dataset comprising 4,500 manually labelled images, categorized into seven distinct hydrophobicity classes. The images were sourced from a publicly available dataset, ensuring objectivity and minimizing the risk of confirmation bias [7] . Representative samples of the original images are presented in Fig 1.  \nB. Image Processing  \nThe image processing workflow involved several key steps to prepare the data for feature extraction:  \na. Conversion – Simplified image content by removing colour information, making further processing more efficient.  \nb. Gaussian Blurring – Applied to reduce image noise and improve the clarity of droplet edges.  \nc. Binarization – Transformed grayscale images into binary format to clearly separate water droplets from the background.  \nd. Contour Detection – Enabled the identification and extraction of shape-based features, including the number of contours, droplet area, perimeter, and circularity metrics [10–15] . The detected contours surrounding each droplet were overlaid for visual reference, as illustrated in Fig 2.  \nC. Machine Learning  \nTo evaluate the effectiveness of machine learning in classifying hydrophobicity levels, six supervised classification algorithms wer","cbCain0OWlsXvWIW","https://ap.wps.com/l/cbCain0OWlsXvWIW","pdf",740885,1,4,"English","en",105,"# Introduction\n# Methodology\n## Dataset\n## Image Processing\n## Machine Learning\n# Results\n## Accuracy and Confusion Matrix\n## ROC-AUC Analysis","[{\"question\":\"为什么需要自动化的疏水性评估方法？\",\"answer\":\"UV辐射、污染和湿气会随时间降低复合绝缘子的疏水性，从而增加失效风险。传统IEC 62073喷雾方法受人为判断影响，现场一致性较差。\"},{\"question\":\"研究使用了什么数据集与分类设定？\",\"answer\":\"研究使用4,500张人工标注的绝缘子图像，并将其分为7个疏水性类别。训练/测试采用80/20划分以进行评估。\"},{\"question\":\"哪些机器学习模型表现最好？\",\"answer\":\"梯度提升（Gradient Boosting）与随机森林（Random Forest）在分类准确率与ROC-AUC方面表现最佳。两者准确率约为75%和73%，且AUC均超过0.90。\"}]","Automated Hydrophobicity Assessment of Composite Insulators via Contour Analysis and Machine Learning | PDF",1785938098,10,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"automated-hydrophobicity-assessment-of-composite-insulators-via-contour-analysis-and-machine-learning","",{"@graph":36,"@context":85},[37,53,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":21},"https://docshare.wps.com/document/automated-hydrophobicity-assessment-of-composite-insulators-via-contour-analysis-and-machine-learning/127284/",{"url":52,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":41,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-08-22","2026-08-05",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"为什么需要自动化的疏水性评估方法？","Question",{"text":75,"@type":76},"UV辐射、污染和湿气会随时间降低复合绝缘子的疏水性，从而增加失效风险。传统IEC 62073喷雾方法受人为判断影响，现场一致性较差。","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"研究使用了什么数据集与分类设定？",{"text":80,"@type":76},"研究使用4,500张人工标注的绝缘子图像，并将其分为7个疏水性类别。训练/测试采用80/20划分以进行评估。",{"name":82,"@type":73,"acceptedAnswer":83},"哪些机器学习模型表现最好？",{"text":84,"@type":76},"梯度提升（Gradient Boosting）与随机森林（Random Forest）在分类准确率与ROC-AUC方面表现最佳。两者准确率约为75%和73%，且AUC均超过0.90。","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,134],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":21,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":29,"doc_module":4,"doc_module_name":46,"category_name":132,"show_sort_weight":29,"slug":133},"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]