Induction Motor-Thermal Images
Description
Dataset comprising 488 high-resolution thermal images of induction motors captured under varying operational conditions: no, medium, and full load. The dataset includes motors with both healthy and faulty bearings, resulting in six distinct operational states: Healthy Full Load (HFL), Healthy Medium Load (HML), Healthy No Load (HNL), Corroded Full Load (CFL), Corroded Medium Load (CML), and Corroded No Load (CNL). Distributed as 112 for HFL, 82 for HML, 90 for HNL, 49 for CFL, 79 for CML, and 76 for CNL. Images were captured using a FLIR C5 thermal camera, offering a thermal resolution of 160 × 120 pixels and a temperature detection range of -20°C to 400°C.
Files
Steps to reproduce
V. V. Kulkarni, M. Kundu, and J. B. Simha, ―State Transition Classification for Predictive Maintenance Using Thermal Images and Deep Learning,‖ Int. J. Intell. Eng. Syst., vol. 18, https://doi.org/10.22266/ijies2025.0630.39 pp. 557–570, 2025.
Institutions
- REVA UniversityKarnataka, Bengaluru