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

Eye positioning method and system

A technology of eye positioning and eye position, applied in the computer field

Active Publication Date: 2017-05-31
北京海鑫科金高科技股份有限公司
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] However, this method involves exhaustive search many times in the positioning process, including exhaustive search of candidate areas to find eye positions and exhaustive search of all candidate eye pairs, which greatly improves the positioning time of the algorithm, and is suitable for applications with high real-time requirements scene, the algorithm has certain limitations

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
  • Eye positioning method and system
  • Eye positioning method and system
  • Eye positioning method and system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0075] Embodiment 1 of the present invention proposes an eye positioning method, see figure 1 , Including the following steps:

[0076] Step 101: Preprocess several images as input samples, and preprocess several pairs of eye position coordinates as target samples.

[0077] Step 102: Use the input sample and the target sample to train a combined classifier.

[0078] Step 103: Preprocess the image to be located and input it into the combined classifier to obtain the eye position coordinates corresponding to the image to be located.

[0079] It can be seen that in the embodiment of the present invention, the relationship between the original image and the position coordinates of the eyes is directly used for eye positioning, and the combined classifier is trained by using several images as input samples and several pairs of eye position coordinates as target samples. The original image to be positioned is input into the trained combined classifier, so that the position coordinates of th...

Embodiment 2

[0081] In the following, embodiment 2 is used to illustrate a preferred implementation process of the present invention in more detail. See figure 2 , image 3 , Figure 4 with Figure 5 , The process includes the following steps:

[0082] Step 201: Preprocess several original images as input samples.

[0083] In this step, before performing face detection on the original image, it is necessary to perform preliminary processing on the original image. The main purpose is to increase the variety of samples, see image 3 Shown. Here, common image processing methods can be used to process the image, including: adding noise, rotating, blurring, etc., in order to simulate fisheye lens imaging, radial distortion processing is also added. Various changes randomly act on the training samples with a certain probability, which not only increases the variation of the samples, but also increases the number of samples.

[0084] Perform face detection on the processed image to obtain several fac...

Embodiment 3

[0142] Embodiment 3 of the present invention also proposes an eye positioning system, see Image 6 , The system includes:

[0143] The preprocessing module 601 is used to preprocess several images as input samples, and preprocess several pairs of eye position coordinates as target samples;

[0144] The combined classifier training module 602 is configured to train a combined classifier by using the input sample and the target sample;

[0145] The positioning module 603 is configured to preprocess the image to be located and input it into the combined classifier to obtain the eye position coordinates corresponding to the image to be located.

[0146] Wherein, the preprocessing module 601 includes an input sample preprocessing subunit 6010 and a target sample preprocessing subunit 6011, wherein,

[0147] The input sample preprocessing subunit 6010 is used to perform face detection on several images to obtain several face detection images;

[0148] Normalize several face detection images t...

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 method and system for positioning eyes. The method includes the steps that a plurality of images are preprocessed and then serve as input samples, and a plurality of eye position coordinates are preprocessed and then serve as target samples; a combined classifier is trained by means of the input samples and the target samples; images to be positioned are preprocessed and then input to the combined classifier, so that eye position coordinates corresponding to the images to be positioned are obtained. According to the method and system, the eyes can be positioned quickly and accurately.

Description

Technical field [0001] The invention relates to the field of computer technology, in particular to an eye positioning method and system. Background technique [0002] Eye positioning technology is mainly used in face recognition, expression recognition and other related fields. Its main goal is to accurately locate the eye position in a complex background. It is the core link between two-dimensional image data and higher-level cognition. [0003] At present, the problem of eye positioning is basically solved based on the learning method. The learning-based method mainly performs eye positioning by extracting features and training the classifier model. The learning-based method can well adapt to the complexity of the external environment. Change is the mainstream method in the field of eye positioning. There are many related studies in eye positioning methods based on learning, and one of the more effective eye positioning methods is achieved in the following ways: [0004] In the 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 Patents(China)
IPC IPC(8): G06K9/66
Inventor 刘晓春赵元兴王贤良
Owner 北京海鑫科金高科技股份有限公司