云梯系统平衡车梯控告警分析数据

依据对平衡车入梯行为的监测，及时在轿厢内提醒乘客安全乘梯，有效防止平衡车在电梯内引发安全事故；设备抓图便于在发生突发状况时，快速了解现场状况，定位电梯及平衡车所在位置，对电梯内居民的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，方便物业对平衡车入梯行为进行日常监管，合理地分配人力和物力资源，提前采取措施，预防平衡车违规进入电梯导致的安全事故，保障居民安全，提升居民对物业服务的满意度和信任度；根据一段时间内告警的数据规模、现场图片、时间等，公开报道一段时间内平衡车入梯告警数据，可以提高公众对平衡车入梯危险行为的重视程度。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=3，表示目标结果为平衡车，系统会向平台发送平衡车入梯告警，告警时间为时间戳，更加精准。

This system monitors the behavior of electric balance bikes entering elevator cars, and issues timely safety reminders to passengers inside the elevator cabin to prevent safety accidents caused by balance bikes in elevators. Captured images enable rapid on-site situation assessment and location tracking of both the elevator and the balance bike during emergencies, effectively safeguarding the safety of residents in the elevator. Regular analysis of alarm data and captured on-site images facilitates property management's daily supervision of balance bike entry into elevators, enabling rational allocation of human and material resources, proactive prevention of safety accidents caused by unauthorized balance bike access to elevators, ensuring resident safety, and improving residents' satisfaction and trust in property management services. Furthermore, publishing periodic alarm data including its scale, on-site images and timestamps can enhance public awareness of the risks associated with balance bikes entering elevators.

1. Detect operating elevators using a real-time elevator door state detection algorithm based on sensors.
2. Enable the in-elevator camera image capture service with a capture rate of 1 FPS, and upload the captured images to Huawei Cloud Object Storage Service (OBS). The cloud-based algorithm server pulls the images from Huawei Cloud OBS, and conducts image detection via a target detection algorithm trained on deep neural networks, outputting relevant image attributes.
3. The output image attributes include: target category, target bounding box, and detection confidence score.
4. Identify the detected target category: if the category value equals 3, the target is confirmed as a balance bike, and the system will send a balance bike entry alert to the platform with the alert time marked by a timestamp, ensuring higher detection accuracy.

搜集汇总

数据集介绍

特点

该数据集包含云梯系统对平衡车入梯行为的告警数据，涵盖告警类型、时间、位置及设备抓图等信息，用于安全监测和物业管理。