配电房AI图像识别数据集

配电房AI图像识别数据集主要面向适用于边缘侧人工智能快速推理与决策技术研究，本数据集用于异常监测，除了包含互联网爬取的数据外，还使用以下三种方式进行数据采集：采集下雨天地面积水或者人工倒水图像数据；采集装有浑浊泥水和清澈水的白色长方体容器来模拟水浸场景；在室内和室外通过燃烧纸张的方式来模拟火焰场景，数据量79.9MB。

This AI-powered image recognition dataset for power distribution rooms is primarily targeted at research on edge-side artificial intelligence fast inference and decision-making technologies. It is designed for anomaly monitoring. In addition to data crawled from the Internet, data was collected via the following three methods: 1. Collecting image data of ground puddles during rainy days or images obtained by manually pouring water; 2. Collecting images of white rectangular containers filled with turbid muddy water and clear water to simulate water immersion scenarios; 3. Simulating fire scenarios by burning paper both indoors and outdoors, and collecting corresponding image data. The total data volume of this dataset is 79.9 MB.

搜集汇总

背景与挑战

背景概述

该数据集是一个面向配电房异常监测的AI图像识别数据集，主要用于边缘侧人工智能快速推理与决策技术研究。数据通过互联网爬取和模拟场景（如水浸和火焰）采集，总数据量为79.9MB，包含205个文件，由威胜信息技术股份有限公司创建，属于国家重点研发计划项目。