Human body fall detection method and device

A detection method and detection device technology, applied in image data processing, image enhancement, instruments, etc., can solve the problems of decreased detection rate of human body, low detection rate and accuracy of human fall events, and difficult acquisition

Inactive Publication Date: 2018-09-04
ZHEJIANG DAHUA TECH CO LTD
View PDF11 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] The present invention provides a method and device for detecting human falls, which are used to solve the problems in the prior art in the top-mounted scene, which are difficult to collect and the detection rate of human body drops, resulting in the detection rate and accuracy of human fall events. Low rate of technical issues

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
  • Human body fall detection method and device
  • Human body fall detection method and device
  • Human body fall detection method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0076] In the actual scene, when the human body falls, the speed of the human body itself will undergo a large change, that is, the human body will suddenly accelerate. In addition, in the top-mounted scene of the video equipment, when the human body falls, the area of ​​the area corresponding to the head or shoulder in the image will also change accordingly, that is, in the video image collected under the top-mounted scene, The area of ​​the area corresponding to the head or shoulders will be reduced due to a fall of the human body. Therefore, in the top-mounted scene of the video equipment, by combining the area of ​​the area corresponding to the head or shoulder in the captured video and the movement speed of the human body, it can be judged whether the human body has fallen.

[0077] Therefore, according to an aspect of the embodiments of the present invention, a method for detecting a human body fall is provided, which is applied to a top-mounted scene of an image acquisi...

Embodiment 2

[0093] Specifically, in the embodiment of the present invention, a method for detecting a human fall is disclosed, and the corresponding sequence flow chart of the method is as follows figure 2 As shown, that is, in the top-mounted scene of the image acquisition device, after acquiring the image of the current frame, perform the following steps:

[0094] Step S201: Obtain the first area of ​​the target to be detected determined by the body recognition part in the current frame image;

[0095] Step S202: Obtain the current speed of the target to be detected;

[0096] Step S203: Judging whether the current speed of the target to be detected is greater than the speed range of the normal movement of the target to be detected, if the judgment result is yes, execute step S205, and if the judgment result is no, execute step S204;

[0097] Step S204: Continue to track the target to be detected;

[0098] Step S205: Judging whether the area of ​​the first area is smaller than the cor...

Embodiment 3

[0104] Based on the above embodiments, the embodiment of the present invention specifically describes the process of determining the first region of the object to be detected in the current frame image determined by the body recognition part in step S101.

[0105] Wherein, step S101 specifically includes:

[0106] Extract the region corresponding to at least one human body recognition part of the target to be detected from the current frame image through the feature extraction layer of CNN;

[0107] Mapping the extracted area corresponding to at least one human body recognition part to the current frame image through the feature mapping layer of CNN to generate a rectangular target frame;

[0108] The rectangular target frame is determined as the first area where the target to be detected is determined by the body recognition part.

[0109] CNN is a feed-forward neural network, artificial neurons can respond to surrounding units, and can process large images. Generally, the ...

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 human body fall detection method and device. The method is applied in a top mounting scene and comprises the following steps of: obtaining a current frame of image; determining a first area, determined by at least one preset human body recognition part, of a to-be-detected target in the current frame of image according to the human body recognition part; determining a current speed of the to-be-detected target according to the first area in the current frame of image and a second area corresponding to a to-be-detected target in a first image which is located in frontof the current frame and spaced from the current frame by a preset number of frames; judging whether the current speed is greater than a stored normal movement speed range of the to-be-detected target; if the current speed is greater than the normal movement speed range, judging whether an area of the first area is smaller than a stored corresponding area threshold value when the to-be-detected target moves normally or not; and if the area of the first area is smaller than the area threshold value, judging that the to-be-detected target falls. According to the method and device, movement speeds of to-be-detected targets and areas, determined by human body recognition parts, of areas are detected, so that the problems that the acquisition difficulty is high, the human body detection rate islow and the human body fall event detection rate and correctness are relatively low are solved.

Description

technical field [0001] The invention relates to the technical field of video monitoring, in particular to a method and device for detecting human falls. Background technique [0002] According to statistics, falls have become one of the most common and serious safety problems in the daily life of the elderly. With the decline of camera manufacturing costs and the popularity of the Internet, many users monitor the elderly living alone in real time through the use of network cameras. When the elderly fall, an early warning can be issued. [0003] At present, the commonly used methods for judging human falls include: judging based on human body detection of falling behavior, that is, first performing human body detection, and then judging whether the human body has a falling behavior according to the ratio of the length, width, and height of the human body; and human body based on model training. To judge the fall behavior, that is, based on different fall postures, extract t...

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): G06K9/00G06N3/04G06T7/00G06T7/254G06T7/269G06T7/62
CPCG06T7/0002G06T7/254G06T7/269G06T7/62G06T2207/30196G06T2207/20084G06V20/44G06V20/41G06N3/045
Inventor 孙莉龚磊
Owner ZHEJIANG DAHUA 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