Eureka AIR delivers breakthrough ideas for toughest innovation challenges, trusted by R&D personnel around the world.

Neural network system and image crowd counting method based on neural network system

A neural network and crowd technology, applied in the field of computer vision, can solve problems such as mismatching, no multi-scale image feature processing, etc., to achieve the effect of improving accuracy

Active Publication Date: 2021-05-14
广东众聚人工智能科技有限公司
View PDF6 Cites 7 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The related crowd counting method based on deep learning, when using the attention mechanism, often divides the crowd picture into several blocks of different crowd density levels, and assigns different weights. The disadvantage of this is that it is different from the real pixel-based density. Figure does not match
In addition, many models do not refine the multi-scale image features extracted by the convolution of different convolution kernel sizes

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
  • Neural network system and image crowd counting method based on neural network system
  • Neural network system and image crowd counting method based on neural network system
  • Neural network system and image crowd counting method based on neural network system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0063] This embodiment proposes a neural network system, which is based on a pixel-level multi-scale attention mechanism, and is used to predict the crowd density of a crowd image to be predicted. Integrating the crowd density can realize crowd counting in the crowd image to be predicted. The system includes: a shared encoder, a density feature prediction branch, a pixel-level multi-scale attention branch and a fusion module.

[0064] The shared encoder is used to obtain the image of the crowd to be predicted, and extract the multi-scale fusion information of the image of the crowd to be predicted X’ .

[0065] The density feature prediction branch is connected with the shared encoder, which will X’ as input for the X’ Obtain the image of the crowd to be predicted S density feature map, where, S is an integer greater than or equal to 1.

[0066] The pixel-level multi-scale attention branch is connected with the shared encoder, which will X’ as input for the X’ Obtain t...

Embodiment 2

[0108] This embodiment provides a method for counting people in an image based on a neural network system. The method is based on the neural network system described in Embodiment 1, and is used to realize crowd counting of crowd images. Figure 6 It is a flow chart of an image crowd counting method based on a neural network system provided by an embodiment of the present invention. Such as Figure 6 As shown, the method includes steps S10-S40.

[0109] S10: Obtain a plurality of training crowd images; perform density labeling on each training crowd image to generate a label density map of each training crowd image; integrate the label density map to obtain each training crowd image the total number of people.

[0110] S20: Construct any neural network system as described in Embodiment 1.

[0111] S30: Input each training group image into the neural network system in turn to obtain a density prediction map of each training group image; use the label density map of each tra...

Embodiment 3

[0143] Figure 8 It is a schematic structural diagram of a computer device provided by an embodiment of the present invention. Such as Figure 8 As shown, the device includes a processor 810 and a memory 820 . The number of processors 810 may be one or more, Figure 8 A processor 810 is taken as an example.

[0144] The memory 820, as a computer-readable storage medium, can be used to store software programs, computer-executable programs and modules, such as program instructions / modules of the image crowd counting method based on the neural network system in the embodiment of the present invention. The processor 810 executes the software programs, instructions and modules stored in the memory 820 to implement the above-mentioned image crowd counting method based on the neural network system.

[0145] The memory 820 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system and an application program req...

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 neural network system and an image crowd counting method based on the neural network system. The neural network system is used for predicting the crowd density of a to-be-predicted crowd image, and comprises: a shared encoder, which is used for extracting multi-scale fusion information X' of the to-be-predicted crowd image; a density feature prediction branch, which is connected with the shared encoder and is used for acquiring S density feature maps of the to-be-predicted crowd image; a pixel-level multi-scale attention branch, which is connected with the shared encoder and is used for acquiring S attention masks of the to-be-predicted crowd image; and a fusion module, which is connected with the density feature prediction branch and the pixel-level multi-scale attention branch, and is used for fusing the S density feature maps and the S attention masks. According to the method, the crowd density information of the pixel level is considered, and the multi-scale information is fused, so that the crowd counting precision is improved.

Description

technical field [0001] The embodiments of the present invention relate to the technical field of computer vision, and in particular to a neural network system and an image crowd counting method based on the neural network system. Background technique [0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art. [0003] With the development of society, urban population density continues to grow rapidly, and there are more and more large-scale population gathering places in cities and towns. Timely and efficient monitoring and management of crowd density can effectively prevent accidents such as crowding and stampede. Therefore, the crowd counting task has also received extensive attention from all walks of life in recent years. The image crowd counting method can be stably deployed and applied in various scenarios, such as traffic monitoring systems, security robots, shopping mall ...

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): G06K9/00G06K9/62G06N3/04G06N3/08
CPCG06N3/084G06V20/53G06N3/048G06N3/044G06N3/045G06F18/253G06F18/214
Inventor 张凯贺甜甜丁冬睿逯天斌
Owner 广东众聚人工智能科技有限公司
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
Eureka Blog
Learn More
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com