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

Static gesture recognition method based on improved VGGNet network and PCA

A gesture recognition and static technology, applied in character and pattern recognition, neural learning methods, biological neural network models, etc., can solve the problems of unsatisfactory recognition, achieve the effect of shortening calculation time, improving accuracy and efficiency, and improving efficiency

Active Publication Date: 2020-09-22
GUILIN UNIV OF ELECTRONIC TECH
View PDF9 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

At present, human beings have entered the era of big data, and traditional gesture recognition methods are often unsatisfactory in the face of massive data and various external conditions.

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
  • Static gesture recognition method based on improved VGGNet network and PCA
  • Static gesture recognition method based on improved VGGNet network and PCA
  • Static gesture recognition method based on improved VGGNet network and PCA

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0048] A kind of static gesture recognition method based on improving VGGNet network and PCA of the present invention is:

[0049] 1. Set up the Kinect camera 1m-2m in front of the person;

[0050] 2. Start the camera, set the scanning interval to scan 10 times per second, that is, within one second, the camera acquires 10 images of human hand information;

[0051] 3. Train the gesture image model. Improve the traditional VGGNet network and introduce a hash layer to improve the efficiency of gesture recognition under the premise of ensuring accuracy. The specific process is as follows:

[0052] (1) Input the original image I(x,y);

[0053](2) Estimate the noise of each position and remove it. Assuming that the image I seen by the human eye is the product of the image illumination component L and the reflectivity component R, the specific expression is shown in formula 1:

[0054] I(x,y)=R(x,y) L(x,y) (1)

[0055] (3) Separate the three color channel space components and ...

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 static gesture recognition method based on an improved VGGNet network. According to the method, a VGGNet network is improved, different branches are used for learning label information, meanwhile, a hash layer is introduced into the network to project features of a gesture image to a Hamming space, and the gesture processing efficiency is improved by calculating the Hamming distance. An improved VGGNet network is combined with a traditional principal component analysis method to carry out gesture recognition; firstly, an original data set and an enhanced data set areused as input data to train a VGGNet network, the high-dimensional features of the gesture images are obtained, then dimension reduction is carried out on the features of the high-dimensional images through the PCA, the length of feature vectors is reduced, the calculation time of similarity measurement between the images is shortened, and therefore the gesture recognition precision and efficiencyare improved.

Description

technical field [0001] The invention relates to a static gesture recognition method, in particular to a static gesture recognition method based on improved VGGNet network and PCA. Background technique [0002] In the 1990s, in order to solve the problems of manual labeling, researchers began to turn their attention to feature extraction to the image content itself, so a series of content-based image recognition methods were proposed and widely used. In order to accurately describe the characteristics of the image, most of the early CBIR technologies use global visual features to describe the image. This feature description method is relatively simple, and users can perform image retrieval conveniently and efficiently. However, since this method extracts the low-level visual features of the image, when encountering the interference of external factors, such as light intensity, occlusion, deformation and other harsh conditions, the effective features of the image cannot be ac...

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): G06K9/00G06K9/62G06N3/04G06N3/08G06F16/51G06F16/55G06F16/583
CPCG06N3/08G06F16/51G06F16/55G06F16/583G06V40/107G06V40/28G06N3/047G06N3/045G06F18/2135G06F18/2411G06F18/2415
Inventor 谢武贾清玉刘满意强保华崔梦银瞿元昊
Owner GUILIN UNIV OF ELECTRONIC TECH