Children visual impairment screening system and method based on deep learning algorithm
What is Al technical title?
Al technical title is built by PatSnap Al team. It summarizes the technical point description of the patent document.
A vision impairment, deep learning technology, applied in the field of pediatric vision impairment screening system, can solve problems such as inapplicability and lack of video quality control modules
Inactive Publication Date: 2021-01-01
苏州体素信息科技有限公司
View PDF6 Cites 0 Cited by
Summary
Abstract
Description
Claims
Application Information
AI Technical Summary
This helps you quickly interpret patents by identifying the three key elements:
Problems solved by technology
Method used
Benefits of technology
Problems solved by technology
However, this research is mainly carried out in medical scenarios. The deep neural network architecture and technical route adopted are different from this patent. It is
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
Click on the blue label to locate the original text in one second.
Reading with bidirectional positioning of images and text.
Smart Image
Examples
Experimental program
Comparison scheme
Effect test
Embodiment 1
[0111] According to a method for screening children with visual impairment based on deep learning algorithm provided by the present invention, comprising:
[0112] Step M1: Obtain a video of a child's upper body;
[0113] Step M2: Analyze the video in real time, use the quality control algorithm to extract high-quality segments in the video that meet the preset requirements, and use the key point information of the face to cut out high-quality facial segments and eye segments;
[0114] Step M3: using face segment information and eye segment information to train the face classification network and eye classification network respectively until the loss function converges, and obtain the trained face classification network and eye classification network;
[0115] Step M4: Use the trained face classification network and eye classification network to encode each face array and eye array to obtain the corresponding probability of suffering from eye diseases;
[0116] Step M5: Integ...
Embodiment 2
[0208] Embodiment 2 is a modification of embodiment 1
[0209] Step 1: Take a video of the child's upper body with the smartphone's front-facing camera.
[0210] Step 2: Analyze the video in real time, and use the quality control algorithm to extract high-quality clips in the video.
[0211] Step 3: For each extracted high-quality segment, use the video classification network to give the label of whether the child is healthy or not and the corresponding probability.
[0212] The quality control algorithm proposed in the step 2 includes the following steps:
[0213] Step 2.1: Extract the input video frame by frame to form a series of video frames.
[0214] Step 2.2: Input the extracted video frame into the face key point extraction network, and the quality control network gets the key point information of the face
[0215] Step 2.3: Use the face key point information obtained in step 2.2 to perform two-dimensional transformation to calculate the yaw angle of the face orienta...
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
Login to view more
Abstract
The invention provides a children visual impairment screening method and system based on a deep learning algorithm. The children visual impairment screening method comprises the steps: acquiring a video of the upper body of a child; analyzing the video in real time, extracting high-quality segments meeting preset requirements in the video, and cutting out high-quality face segments and eye segments; respectively training a face classification network and an eye classification network by utilizing the face segment information and the eye segment information until the loss function converges toobtain the trained face classification network and eye classification network; using the trained face classification network and eye classification network to encode each face array and each eye arrayto obtain a corresponding probability of suffering from eye diseases; integrating the obtained corresponding probabilities of suffering from the eye diseases, and finally outputting a yin-yang judgment result of judging whether the child suffers from the visual impairment or not. According to the invention, the problem of visual impairment screening by using the upper body video of the child shotby using the camera of the smart phone is solved.
Description
technical field [0001] The present invention relates to a rapid vision screening method for human eyesight, in particular to a screening system and method for visual impairment in children based on a deep learning algorithm. Background technique [0002] At present, there are many existing technologies that can perform key point detection and expression recognition on human faces with high accuracy. However, there is currently no patent or product that can specifically screen children for vision health conditions. Even if the child's expression can be recognized, it is difficult to deduce from the expression whether the child has eye abnormalities, such as strabismus, nystagmus, etc. [0003] Patent document CN106169073A (application number: 201610539445.6) discloses an expression recognition method and system, which can improve the recognition accuracy of facial expressions. The method comprises: obtaining a facial expression image as a training sample and a test sample, ...
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
Application Date:The date an application was filed.
Publication Date:The date a patent or application was officially published.
First Publication Date:The earliest publication date of a patent with the same application number.
Issue Date:Publication date of the patent grant document.
PCT Entry Date:The Entry date of PCT National Phase.
Estimated Expiry Date:The statutory expiry date of a patent right according to the Patent Law, and it is the longest term of protection that the patent right can achieve without the termination of the patent right due to other reasons(Term extension factor has been taken into account ).
Invalid Date:Actual expiry date is based on effective date or publication date of legal transaction data of invalid patent.
Login to view more 
Login to view more