Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

A method for estimating three-dimensional human body postures and hand information

A technology of human body posture and hands, applied in the field of image processing, can solve the problems of incompatibility between accuracy and cost

Active Publication Date: 2019-04-16
ANHUI UNIVERSITY
View PDF2 Cites 24 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0012] The above method cannot balance the accuracy and cost, and cannot be effectively promoted in the consumer market under the premise of ensuring high precision

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
  • A method for estimating three-dimensional human body postures and hand information
  • A method for estimating three-dimensional human body postures and hand information
  • A method for estimating three-dimensional human body postures and hand information

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0064] The structural features and technical advantages of the present invention will now be described in detail in conjunction with the accompanying drawings.

[0065] A method for estimating three-dimensional human body posture and hand information is carried out by a computer in the following steps:

[0066] Step 1: Obtain the color map and the depth map through the depth-of-field color camera, and transmit them to the computer; the depth-of-field color camera can be replaced by an RGB camera and a depth camera / infrared camera;

[0067] Step 2: Hybrid joint prediction by computer;

[0068] Step 3: Fit the human body model to the joint points by the computer, and constrain the head direction to estimate the human body posture and body shape;

[0069] Step 4: The computer uses the point cloud data to optimize the details of the fitting results;

[0070] Step 5: The computer optimizes the fitting process and outputs the results.

[0071] Further, the depth-of-field color ca...

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 relates to a method for estimating three-dimensional human body postures and hand information, in particular to a novel unique method for automatically and accurately estimating the postures of a human body and a hand and the actual figure of the human body at the same time through single-frame data of a color depth camera. Firstly, joint prediction of a depth camera and deep learning are combined with a method for predicting joints from a color map, and meanwhile steering of the head of a human body is obtained to solve the problem that the head posture cannot be matched in theprior art. Meanwhile, the gesture of the hand is restored in the human body estimation process, so that the human body result is more vivid and specific, the 3D gesture of the hand is predicted from the color image by using a deep learning method, and then the fitting of the human body gesture and the latest parameterized human body model SMPL + H is completed by using a 3D human body joint fitting energy model. And finally, the obtained human body surface point cloud is registered with the fitted model to improve the accuracy of human body shape estimation. Therefore, an accurate three-dimensional human body model containing the real hand posture is estimated.

Description

technical field [0001] The invention belongs to the technical field of image processing, in particular to a method for estimating three-dimensional human posture and hand information. Background technique [0002] We are committed to accurately estimating the naked shape and pose of a clothed person, including the accurate state of the hand, using single-frame data from the depth-of-field color camera / Kinect. Compared with some complex systems, we only use a consumer-grade depth camera, and we can really estimate the human body in any pose using only one frame of data. In this work, we propose a creative hybrid joint prediction method, which combines human joints predicted from depth data by depth cameras with 3D joints predicted from color images by deep learning methods to obtain more accurate and robust joint positions. , and at the same time, the actual turning of the head of the human body can be obtained from the camera. At the same time, we can estimate the posture ...

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): G06T7/246G06T7/292G06T17/00
CPCG06T7/251G06T7/292G06T17/00G06T2207/30196G06T2207/10012Y02T10/40
Inventor 方贤勇杨继魁汪粼波李薛剑
Owner ANHUI UNIVERSITY
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products