Thermal images dataset, Transformer, 1 phase dry type
Description
--Contact email: najafi.mohamad.sh@gmail.com m.najafi@nit.ac.ir --Babol Noshirvani University of Technology This version(3) contains 45 samples of Ground Truth images--determined by experts--for segmentation in order to reach the region of interest(ROI) in transformer. --Applying this dataset is allowed just by considering the creator's citation for any academic or other purposes. ---This is thermal images (IRT) dataset in the context of condition monitoring of electrical equipment--Transformers. All artificially generated defects are internal faults, so they neither depend on external pieces nor failure in initial setup components. Regarding the transformer, 8 different cases of short circuit failures in common core winding were examined. The thermal image acquisition is done on the workbench by a Dali-tech T4/T8 infrared thermal camera in an Electrical Machines Laboratory at the environment temperature of 23°. The number of images is 255 in 9 conditions- one of them is healthy/normal condition. You can find the other specification in the attached PDF files or the related published paper.(DOI: 10.1109/ICSPIS51611.2020.9349599)
Files
Steps to reproduce
The related Paper: M. Najafi, Y. Baleghi, S. A. Gholamian and S. Mehdi Mirimani, "Fault Diagnosis of Electrical Equipment through Thermal Imaging and Interpretable Machine Learning Applied on a Newly-introduced Dataset," 2020 6th Iranian Conference on Signal Processing and Intelligent Systems (ICSPIS), Mashhad, Iran, 2020, pp. 1-7, doi: 10.1109/ICSPIS51611.2020.9349599. ---Contact email: najafi.mohamad.sh@gmail.com m.najafi@nit.ac.ir