一种基于人工智能的高速公路恶劣天气识别方法
一种基于人工智能的高速公路恶劣天气识别方法
本发明公开了一种基于人工智能的高速公路恶劣天气识别方法，包括如下步骤：S1、形成视频数据；S2、针对天气状态对视频数据进行分类，分别划分为晴天、阴天、雨天、雾天和雪天五种类别，分别存储；S3、形成提取图片，对应类别进行存储；S4、对形成分析模型；S5、形成关键点像素；S6、形成对比数据库；S7、通过摄像机对高速公路环境进行实时监控，设定比对周期，对视频状态和对比数据库进行对比分析；S8、符合一个类别的天气种类后，输出对比结果；S9、确定识别的恶劣天气种类，进行预警提示，通过识别结果结合温湿度变化进行最终判断，避免误判而干涉，利于高效识别和预警，保证高速公路管理和使用的安全性，方便推广使用。
The invention discloses a highway severe weather identification method based on artificial intelligence. The method comprises the following steps: S1, forming video data; s2, classifying the video data according to weather states, dividing the video data into five categories of sunny days, cloudy days, rainy days, foggy days and snowy days, and storing the five categories respectively; s3, forming extracted pictures, and storing corresponding categories; s4, forming an analysis model; s5, forming a key point pixel; s6, forming a comparison database; s7, monitoring the highway environment in real time through a camera, setting a comparison period, and comparing and analyzing the video state and the comparison database; s8, outputting a comparison result after the weather types accord with the weather types of one category; and S9, the type of the identified severe weather is determined, early warning prompt is performed, and final judgment is performed by combining an identification result with temperature and humidity changes, so that interference caused by misjudgment is avoided, high-efficiency identification and early warning are facilitated, the safety of highway management and use is ensured, and popularization and use are facilitated.
Document information
一种基于人工智能的高速公路恶劣天气识别方法
2024-07-19
Patent
Electronic Resource
Chinese