Smoking recognition method, device and equipment for logistics warehouse and storage medium

An identification method and warehouse technology, applied in the field of logistics warehouse management, can solve problems such as fire expansion and spread, difficulty in evacuating people, and goods blocking evacuation channels, etc., and achieve the effects of improving output efficiency, facilitating call and training, and reducing image processing time

Pending Publication Date: 2021-01-29
SHANGHAI DONGPU INFORMATION TECH CO LTD
View PDF0 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

A large number of goods with various categories meet and flow in each section, making the entire warehouse building space form a state of high fire density. Once a fire occurs, it will easily lead to the expansion and spread of the fire, causing heavy property losses and casualties
[0004] Logistics warehouses have continuous operations such as warehousing, warehousing, loading and unloading, and distribution processing, which will relatively result in a small number of evacuation exits and long distances from evacuation channels; some logistics warehouses are insufficient in area, and evacuation channels are easily blocked by goods entering and leaving the warehouse. Some logistics warehouses are not equipped with fire emergency lights and light evacuation signs, and it is very difficult for people to evacuate and escape in a space environment with low visibility
When a fire breaks out, it is very difficult for firefighters to enter the storage area, and it is difficult to fight the fire, which may easily cause casualties

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
  • Smoking recognition method, device and equipment for logistics warehouse and storage medium
  • Smoking recognition method, device and equipment for logistics warehouse and storage medium
  • Smoking recognition method, device and equipment for logistics warehouse and storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0052] Please see figure 1 , the present embodiment provides a smoking recognition method for a logistics warehouse, including:

[0053] Step S1: Obtain the historical images of smoking behavior, mark the historical images with targets, and establish an image data set; the targets include cigarette butts, fingers with cigarettes and smoke;

[0054] Step S2: Create a YOLOv4 model, add hole convolution to the BackBone network in the YOLOv4 model, and obtain an improved YOLOv4 model;

[0055] Step S3: Input the image data set in step S1 into the improved YOLOv4 model for training to obtain a smoking recognition model;

[0056] Step S4: Input the image of the logistics warehouse into the smoking recognition model, and judge whether there is smoking behavior in the image.

[0057] Specifically, in step S1, smoking behaviors in different scenarios are collected, such as gesture smoking, cigarette in mouth, cigarette holder, red dot of cigarette butt, smoke, etc. The images of smo...

Embodiment 2

[0084] The present invention also provides a smoking identification device for logistics warehouses, see Figure 5 , the device consists of:

[0085] The data set creation module is used to obtain historical images of smoking behavior, carry out target labeling to historical images, and establish an image data set; the target includes cigarette butts, fingers with cigarettes and smoke;

[0086] The model creation module is used to create the YOLOv4 model, and the hole convolution is added to the BackBone network in the YOLOv4 model to obtain an improved YOLOv4 model;

[0087] The model training module is used to input the image data set in the data set creation module into the improved YOLOv4 model for training to obtain a smoking recognition model;

[0088] The smoking recognition module is used for inputting the image of the logistics warehouse into the smoking recognition model, and judging whether there is smoking behavior in the image.

[0089] Wherein, the data set cre...

Embodiment 3

[0092] The above-mentioned embodiment 2 describes in detail the smoking recognition device for logistics warehouses of the present invention from the perspective of modular functional entities. The smoking recognition device for logistics warehouses of the present invention is described in detail below from the perspective of hardware processing.

[0093] Please see Image 6 The smoking recognition device 500 for logistics warehouses may have relatively large differences due to different configurations or performances, and may include one or more processors (central processing units, CPU) 510 (for example, one or more processors) and Storage 520, one or more storage media 530 (such as one or more mass storage devices) for storing application programs 533 or data 532 . Wherein, the memory 520 and the storage medium 530 may be temporary storage or persistent storage. The program stored in the storage medium 530 may include one or more modules (not shown in the figure), and each...

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 discloses a smoking recognition method, device and equipment for a logistics warehouse and a storage medium, and aims to solve the problems that the space of the logistics warehouse in the industry is large, many inflammable materials exist, no partition exists, and a fire is easily caused by smoking; and the method includes establishing an image data set by obtaining historical images of smoking behaviors and performing target labeling on the historical images; creating and optimizing a YOLOv4 model, and inputting the image data set into the optimized YOLOv4 model for training to obtain a smoking recognition model; and inputting the image of the logistics warehouse into the smoking recognition model in real time, and judging whether a smoking behavior exists in the image ornot. Smoking behaviors of workers in a monitoring area are automatically recognized, fire disasters caused by smoking are prevented, and pre-warning, normal state detection and standard management areachieved.

Description

technical field [0001] The invention belongs to the technical field of logistics warehouse management, and in particular relates to a smoking identification method, device, equipment and storage medium for logistics warehouses. Background technique [0002] The goods stored in the logistics warehouse are mainly home appliances, daily necessities, etc. Although the goods flow quickly and stay in the logistics warehouse for a short time, due to the needs of circulation and distribution, a large number of wooden or plastic pallets, packing boxes, etc. are used in the logistics warehouse. Combustion is easy to occur, and there are a large number of electrical equipment in the logistics warehouse. Once leakage or failure occurs, the fire hazard will also increase. [0003] The internal setting of the logistics warehouse requires large space, barrier-free, and integrated, so the building area is relatively large. Due to the need for convenient operation, there are few partition wa...

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): G06K9/00G06N3/04G06N3/08
CPCG06N3/08G06V20/52G06N3/045
Inventor 李斯赵齐辉
Owner SHANGHAI DONGPU INFORMATION TECH 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