Fault-Level Grading of Photovoltaic Cells Employing Lightweight Deep Learning Models

Table 1

Details of the dataset used in the proposed work.

Dataset detail
Number of images	2624 images
Image size	300 × 300
Image type	Gray scale
Bits/image	8-bit
Number of solar modules	44 modules (18 monocrystalline + 26 polycrystalline)
Target classes	4 (defect probability: 0, 0.33, 0.66, 1)