云梯系统煤气罐入梯告警数据

依据本数据，帮助识别潜在的安全隐患，频繁搬运煤气罐可能意味着存在违规使用或存储煤气的行为，从而及时进行干预和纠正；设备抓图便于在发生突发状况时，快速了解现场状况，定位电梯及乘客所在位置，对居民安全起到实际保障作用；定期对告警数据设备抓图所反映的现场情况进行分析，方便物业加强对重点区域和时段的监管；根据一段时间内告警的数据规模、现场图片、时间等，帮助电梯生产销售企业选择销售合适的电梯类型，研判是否需要开发定制化的电梯服务，如一旦识别煤气罐入梯，就进行语音提示；公开报道一段时间内煤气罐入梯告警数据，可以用于开展燃气安全宣传教育，提高公众的重视程度。1.通过基于传感器的实时电梯门开关状态检测算法检出处于运行状态的电梯； 2.开启电梯内摄像头抓图服务(抓图频率1FPS),并将图片上传到华为云obs。云端算法服务器从华为云obs拉取到图片后，通过基于深度学习神经网络训练的目标检测算法进行图片检测，输出图片相关属性；3.输出的图片属性包含：目标类型category、目标坐标框box、目标结果置信度score； 4.对输出的目标类型进行识别，若目标类型category=10，表示目标结果为煤气罐，系统会向平台发送煤气罐入梯告警，告警时间为时间戳，更加精准。

This dataset is developed to assist in identifying potential safety hazards: frequent transportation of gas cylinders may indicate violations of gas usage or storage regulations, enabling timely intervention and correction. The in-elevator camera image capture service enables rapid acquisition of on-site conditions and localization of elevators and passengers during emergencies, providing practical safeguards for resident safety. Regular analysis of on-site conditions reflected in alarm data and captured images allows property management teams to strengthen supervision over key areas and time periods. Based on metrics including alarm data scale, on-site images, and timestamps over a specified period, elevator manufacturing and sales enterprises can select appropriate elevator models for sales, and evaluate whether to develop customized elevator services—for example, triggering voice prompts upon detection of a gas cylinder entering the elevator. Publicly releasing gas cylinder-in-elevator alarm data over a period can also be utilized for gas safety publicity and education to enhance public awareness of relevant risks.
1. Detect elevators in operation using a sensor-based real-time elevator door switch state detection algorithm;
2. Enable the in-elevator camera image capture service with a capture frequency of 1 FPS, and upload the captured images to Huawei Cloud OBS. After the cloud-based algorithm server retrieves the images from Huawei Cloud OBS, it performs image detection via a target detection algorithm trained on deep learning neural networks, and outputs image-related attributes;
3. The output image attributes include: target category, target bounding box, and target confidence score;
4. Identify the target category from the output results. If the target category equals 10, indicating that the detected target is a gas cylinder, the system will send a gas cylinder-in-elevator alarm to the platform, with the alarm time recorded as a timestamp to ensure more precise monitoring.

搜集汇总

数据集介绍

特点

该数据集记录了电梯内煤气罐入梯的告警信息，包括位置、时间、图片等，用于安全隐患识别和安全管理。数据每日更新，通过深度学习算法检测煤气罐并触发告警。