The invention discloses a vehicle base full-automatic operation area intrusion alarm system based on machine vision. According to the technical scheme, the alarm conditions of the full-automatic operation area operation time period (access permission time period) and the non-operation time period (access prohibition time period) are distinguished. In the allowed entering period, through appearance recognition based on a machine vision analysis module, the labor protection wearing condition and personnel identity information of the worker can be analyzed, the working behavior of the worker in the vehicle base is further standardized, and the safety problem caused by the fact that the worker does not wear or mistakenly wears labor protection appliances is effectively reduced. In addition, through behavior recognition of the machine vision analysis module, the non-standard behavior of the worker entering the full-automatic operation area can be prevented, and therefore the situation that the worker is injured and the like is avoided.
本申请公开了基于机器视觉的车辆基地全自动运行区入侵报警系统。本技术方案中,区分了全自动运行区作业时段(允许进入时段)和非作业时段(禁止进入时段)的报警情况。在允许进入时段中,通过基于机器视觉分析模块的外观识别,能够分析工作人员的劳保穿戴情况以及人员身份信息,进一步规范了工作人员在车辆基地中的工作行为,有效减少因不穿戴或错误穿戴劳保用品而引起的安全问题。此外通过机器视觉分析模块的行为识别,可以预防工作人员在进入全自动运行区的不规范行为,从而避免产生人员受伤等情况的发生。
Vehicle base full-automatic operation area intrusion alarm system based on machine vision
基于机器视觉的车辆基地全自动运行区入侵报警系统
07.03.2023
Patent
Elektronische Ressource
Chinesisch
IPC: | B61L Leiten des Eisenbahnverkehrs , GUIDING RAILWAY TRAFFIC |
Europäisches Patentamt | 2019
|Europäisches Patentamt | 2021
|Model-based vision for automatic alarm interpretation
IEEE | 1991
|