PV电池表面缺陷AI评估模型训练数据

本数据是光学图像数据集，用于PV电池表面缺陷AI评估模型的训练。包含1000个来自20块不同PV电池模块的300x300像素8位灰度光学扫描图像。全部使用扫描仪获取,所有图像没有镜头畸变的影响。每个样本都对应有一个含浮点缺陷概率和类型(单晶或多晶)的标注,以量化样品质量和所属的PV类型。

This is an optical image dataset for training AI-powered evaluation models for surface defects of photovoltaic (PV) cells. It contains 1000 8-bit grayscale optical scanned images with a resolution of 300×300 pixels, collected from 20 distinct PV modules. All images are acquired using scanners and are free from lens distortion. Each sample is paired with annotations including floating-point defect probability and PV type (monocrystalline or polycrystalline), which are used to quantify the sample quality and its corresponding PV category.

搜集汇总

数据集介绍

特点

PV电池表面缺陷AI评估模型训练数据是一个光学图像数据集，包含1000个300x300像素8位灰度光学扫描图像，用于训练PV电池表面缺陷检测的AI模型。数据集适用于工业检测场景，但禁用涉及公共安全的领域。