Unlock instant, AI-driven research and patent intelligence for your innovation.

Joint point detection method and posture recognition method and device

A detection method and joint point technology, applied in the field of image processing, can solve problems such as inability to extract multi-scale information, low joint point detection efficiency, low joint point detection accuracy, etc.

Pending Publication Date: 2021-04-06
MEGVII BEIJINGTECH CO LTD
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Because multi-scale information plays an important role in joint point detection, the existing related node detection models usually use cascaded pyramid framework to extract multi-scale information. Although this kind of joint point model has high detection accuracy, the cascaded pyramid framework is too Huge, resulting in low efficiency of joint point detection
At present, the joint point detection model of the single-pyramid framework is a lightweight model. Compared with the cascaded pyramid framework, the single-pyramid framework is light enough, but the joint point detection model of the single-pyramid framework cannot extract sufficient multi-scale information. Leading to lower accuracy of joint point detection
In summary, the existing joint point detection methods cannot efficiently detect and obtain high-precision joint point detection results

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
  • Joint point detection method and posture recognition method and device
  • Joint point detection method and posture recognition method and device
  • Joint point detection method and posture recognition method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0035] First, refer to figure 1 An example electronic device 100 for implementing a joint point detection method, gesture recognition method and device according to an embodiment of the present invention will be described.

[0036] Such as figure 1 Shown is a schematic structural diagram of an electronic device. The electronic device 100 includes one or more processors 102, one or more storage devices 104, an input device 106, an output device 108, and an image acquisition device 110. These components pass through a bus system 112 and / or other forms of connection mechanisms (not shown). It should be noted that figure 1 The components and structure of the electronic device 100 shown are only exemplary, not limiting, and the electronic device may have figure 1 Some components shown may also have figure 1 Other components and structures are not shown.

[0037] The processor 102 can be implemented in at least one hardware form of a digital signal processor (DSP), a field prog...

Embodiment 2

[0044] see figure 2 A schematic flow chart of a joint point detection method shown, the method mainly includes the following steps S202 to S206:

[0045] Step S202, acquiring an image of a target to be detected.

[0046] Wherein, the target image can be an image containing a target object, the target image can be obtained based on the original image to be detected, the original image contains at least one target object, and the target image can be obtained by intercepting the target object contained in the original image, In some embodiments, the original image can be captured by a device with a shooting function (for example, a smart phone or a camera), and a manual upload channel can also be provided for the user, so as to obtain the original image uploaded by the user. In order to better detect the joint points of each target object in the original image, the embodiment of the present invention can intercept the area where each target object is located from the original i...

specific Embodiment approach

[0052] In practical applications, considering that the original image may contain many target objects, which is not conducive to detecting the joint points of each target object in the original image, the embodiment of the present invention intercepts the area where each target object is located from the original image The target image is obtained, and then the joint points contained in each target image are detected respectively, so as to improve the accuracy of joint point detection to a certain extent. The embodiment of the present invention provides a specific implementation manner for acquiring a target image to be detected, and for details, refer to the following steps (1) to (3):

[0053] (1) Obtain the original image to be detected. Wherein, the original image may include an image collected by an electronic device with a shooting function such as a smart phone or a camera, and may also include an image uploaded by a user or an image downloaded by the electronic device ...

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 joint point detection method and device and a posture recognition method and device, and relates to the technical field of image processing, and the method comprises the steps: obtaining a to-be-detected target image; performing articulation point detection on the target image through an articulation point detection model to obtain a corresponding articulation point thermodynamic diagram in the target image; wherein the joint point detection model comprises a feature extraction network and a feature fusion network; wherein the feature extraction network comprises dense connection blocks and orthogonal attention modules between the dense connection blocks; using the orthogonal attention module for extracting multi-scale features; wherein the feature fusion network comprises second-order fusion modules of multiple stages, and the second-order fusion module of each stage is used for carrying out weighted fusion on multi-scale features and aggregation features output by the second-order fusion module of the previous stage; and determining articulation points in the target image based on the articulation point thermodynamic diagram. According to the invention, a high-precision joint point detection result can be efficiently detected.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to a joint point detection method, gesture recognition method and device. Background technique [0002] Human body pose recognition aims to locate key skeletal points (also known as joint points) in images such as eyes, ears and shoulders. With the research of neural networks and the development of hardware facilities, the detection accuracy of joint points has been greatly improved. . Because multi-scale information plays an important role in joint point detection, the existing related node detection models usually use cascaded pyramid framework to extract multi-scale information. Although this kind of joint point model has high detection accuracy, the cascaded pyramid framework is too Huge, resulting in low efficiency of joint point detection. At present, the joint point detection model of the single-pyramid framework is a lightweight model. Compared with the ca...

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
IPC IPC(8): G06K9/00G06K9/62G06N3/04G06N3/08G06T3/40
CPCG06T3/4038G06N3/08G06V40/20G06N3/048G06N3/045G06F18/253
Inventor 罗正雄王志成蔡元昊
Owner MEGVII BEIJINGTECH CO LTD