Escalator pedestrian pose target detection method and system based on Faster R-CNN

A target detection and elevator system technology, applied in the direction of instruments, biological neural network models, character and pattern recognition, etc., can solve problems such as the inability to predict safety accidents, achieve real-time prevention of elevator accidents, improve safety, and ensure detection accuracy Effect
CN112163477AActive Publication Date: 2021-01-01厦门市特种设备检验检测院 +1

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
厦门市特种设备检验检测院
Publication Date
2021-01-01

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
  • Figure 3
    Figure 3
Patent Text Reader

Abstract

The invention discloses an escalator pedestrian pose target detection method and system based on a Faster R-CNN. The escalator pedestrian pose target detection method comprises the following steps that: S1, a VGG16 convolutional neural network extracts a feature map of escalator pedestrian poses; S2, the feature map is input into an RPN layer network; S3, the RPN layer network judges the categoryand the position of a detection object in the feature map; S4, the RPN layer network sends the results to an escalator system; and S5, the escalator system obtains the pedestrian pose information according to the received category and position result information, emergency braking is conducted on the escalator system when dangerous poses exist in the related positions, and the escalator system operates normally when the dangerous poses do not exist in the related positions. According to the escalator pedestrian pose target detection method and the system, the escalator monitoring video streamdata is collected, and the processed pedestrian pose information is sent to the escalator system after the escalator monitoring video stream data is processed by using the FastRCNN network, so that the escalator system is emergently braked to achieve the purpose of preventing escalator accidents in real time when dangerous poses exist in related positions, and the overall safety is improved.
Need to check novelty before this filing date? Find Prior Art

Description

technical field

[0001] The present invention is applied to the field of escalator safety, and is specifically a method and system for detecting pedestrian pose targets of escalators based on Faster R-CNN. Background technique

[0002] The escalator (hereinafter referred to as the escalator) is an electromechanical device that starts frequently and operates under a long-term load. It is a special equipment that involves life safety and is relatively dangerous. Escalators are widely used in subways, shopping malls, hospitals and other public places with a huge flow of people. They have complex characteristics such as frequent operation and changing load dynamics, which are easy to trigger failures and cause safety accidents. The government, media and citizens pay close attention to the safe operation of escalators. If escalator passengers can take corresponding protective measures in time when they fall accidentally, the injury of the accident can be effectively reduced. How...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More