Method for identifying sleeping behavior of operator on duty in inspection field

A technology of on-duty personnel and recognition methods, applied in character and pattern recognition, instruments, computer components, etc., can solve the problems of insufficient variety and inability to deal with the sleeping behavior of many people, and achieve the effect of high recognition accuracy

Pending Publication Date: 2021-08-17
CETC BIGDATA RES INST CO LTD
View PDF4 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] (1) There are not enough identifiable sleeping positions such as prone sleep and back sleep
[0006] (2) It can only recognize the sleeping behavior of a single person in the video data, but cannot deal with the sleeping behavior of multiple people

Method used

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
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Method for identifying sleeping behavior of operator on duty in inspection field

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0027] Such as figure 1 As shown, a method for identifying sleep behavior of personnel on duty in the inspection field includes the following steps:

[0028] ① Data collection: collect the video surveillance data of the working hours of the personnel on duty in the procuratorial field;

[0029] ②Data labeling and processing: Obtain the frame information of the frame rate and size interval from the collected video surveillance data, use the LabelImg image labeling tool to complete the YOLO format labeling of the human body target of the on-duty personnel, and divide the data set into a training set and a verification set;

[0030] ③ Model training and verification: Based on the labeled data in step ② and the YOLOv5 target detection pre-training model, the human target detection model of police officers on duty is obtained;

[0031] ④Model loading and information acquisition: load the human body target detection model, target tracking model and face recognition model of police ...

Embodiment 2

[0042] A method for identifying sleep behavior of a person on duty in a procuratorial field, comprising the following steps:

[0043] S1: Data collection, collecting video surveillance data of the working hours of the personnel on duty in the procuratorial field;

[0044] S2: Data labeling and processing, from the collected video surveillance data to obtain the frame information of the frame rate and size interval, use the LabelImg image labeling tool to complete the YOLO format labeling of the human body target of the on-duty personnel, and divide the data set into a training set and a certain proportion validation set;

[0045] S3: Model training and verification. Based on the above-mentioned marked data, model fine-tuning is performed on the basis of the YOLOv5 target detection pre-training model to obtain a human target detection model for police officers on duty;

[0046] S4: Model loading and information acquisition, loading police officers on duty human target detectio...

Embodiment 3

[0056] Such as figure 1 As shown, a method for identifying sleep behavior of personnel on duty in the inspection field includes the following steps:

[0057] Execute step S1 to obtain off-line video monitoring data of the on-duty personnel in the procuratorial field. These data include different on-duty officers, different monitoring angles, and on-duty conditions at different time periods. The data format is MP4.

[0058] Execute step S2 to obtain frame information from the collected video surveillance data at intervals of 25 fps. Use the LabelImg image annotation tool to complete the YOLO format annotation of the human body target of the duty officer, and label it as "officer". Divide the data set into training set and verification set according to the ratio of 8:2;

[0059] Execute step S3, carry out model training and verification, load the basic pre-training model YOLOv5m.pt of YOLOv5, and perform fine-tuning training on the model for the marked on-duty officer data, an...

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

PUM

No PUM Login to view more

Abstract

The invention provides a method for identifying sleeping behaviors of operators on duty in the inspection field. The method comprises the following steps: (1) collecting video monitoring data of the operators on duty in the inspection field; (2) intercepting frame data from the collected video data according to the size of a frame rate; (3) performing fine tuning training and verification on the acquired data based on a YOLOv5 pre-training model to obtain a human body detection model of the operator on duty; (4) loading an on-duty personnel detection model after fine adjustment based on YOLOv5, a DeepSort target tracking model and the like, and respectively obtaining human body information and face information; (5) based on the multi-model identification result, customizing a sleep behavior research and judgment rule, and identifying a sleep behavior; and (6) when the existence of the sleep behavior is determined, performing sleep behavior early warning. According to the method provided by the invention, specific sleeping individuals can be identified under the condition of multiple persons, various sleeping postures can be processed, the model accuracy is high, and the sleeping behavior identification effect is good.

Description

technical field [0001] The invention relates to a sleeping behavior recognition method for duty personnel in the procuratorial field, which belongs to the fields of computer vision and image processing. Background technique [0002] In the procuratorial field, prison security is one of the most important tasks of administrative units. Once abnormal emergencies such as escape, prison violence, and suicide occur, the relevant responsible persons will be held criminally accountable. Therefore, it is of great significance to improve the informatization level of prison remote law enforcement supervision in the procuratorial field by using computer vision processing related technologies to collect, mine and analyze prison video surveillance information and realize the automation and intelligence of prison law enforcement business. [0003] Sleeping behavior is one of the common behaviors of police officers on duty during law enforcement in prisons. Especially at night, it is diffi...

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

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/00G06K9/62
CPCG06V40/16G06V40/20G06V20/52G06V2201/07G06F18/214G06V10/82
Inventor 闫盈盈丁剑飞范振军罗庆杨秀坤刘汪洋
Owner CETC BIGDATA RES INST CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap