2026-10-10T03:22:00Z https://data.mendeley.com/oai

oai:data.mendeley.com/zs55dv63kv.1 2026-09-17T12:10:28Z

NOGAY, HIDIR SELCUK Microscopic Image Dataset for CAT6 Cable Outer-Sheath Surface Condition Classification Mendeley Data This dataset contains 2,400 original microscopic images acquired from the outer sheath of an approximately 10-m-long CAT6 Ethernet cable under controlled laboratory conditions. The cable had been in use for an extended period and exhibited different surface conditions along its length. Images were captured using a USB digital microscope with a 1920 × 1080 pixel image sensor under consistent illumination, working distance, and viewing conditions. The dataset is organized into three balanced surface-condition classes: Intact, Bending_Damage, and Severe_Damage, with 800 images in each class. Intact images show no visible surface damage; Bending_Damage images exhibit localized texture distortion or surface waviness without visible material rupture; and Severe_Damage images contain visible cracks, material rupture, or deep surface discontinuities. All 2,400 images are original microscope acquisitions. No images were generated through data augmentation, extracted from video frames, or produced by cropping larger source images. The images were acquired across multiple physical regions of the same cable; however, physical region identifiers were not recorded during acquisition. The dataset is intended as a task-specific resource for supervised microscopic classification of CAT6 cable outer-sheath surface conditions and for comparative evaluation of image-classification methods under controlled acquisition conditions. Because the dataset originates from a single cable specimen, it should not be interpreted as representing cross-cable or real-world generalization. Future extensions may incorporate multiple cable specimens, materials, manufacturers, acquisition conditions, and region-level metadata. The original 2,400 images are provided in the Raw_Dataset.zip archive, organized into three class-specific folders: Intact, Bending_Damage, and Severe_Damage. Citation: If you use this dataset in your research, please cite the associated research article. The full citation and DOI of the article will be added to this dataset record upon publication. Engineering Artificial Intelligence Computer Vision Image Processing Electrical Component Deep Learning Dataset doi:10.17632/zs55dv63kv.1 oai:data.mendeley.com/zs55dv63kv.1 Creative Commons Attribution 4.0 International http://creativecommons.org/licenses/by/4.0 https://data.mendeley.com/datasets/zs55dv63kv 2026-09-17T12:10:28Z