Improved SSD dual-network examination room examinee position rapid detection method

A detection method and dual-network technology, which is applied in the fields of identification, positioning and population statistics of examinees, can solve problems such as low accuracy and slow speed, and achieve the effects of improved detection rate, good robustness, and high detection accuracy

Active Publication Date: 2019-10-25
SHAANXI NORMAL UNIV
View PDF3 Cites 10 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The technical problem to be solved by the present invention is to overcome the disadvantages of slow speed and low accuracy of the above-mentioned prior art, and provide a robust, high detection precision, Fast, accurate and improved SSD dual-network rapid detection method for the position of candidates in the examination room

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
  • Improved SSD dual-network examination room examinee position rapid detection method
  • Improved SSD dual-network examination room examinee position rapid detection method
  • Improved SSD dual-network examination room examinee position rapid detection method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0052] The images in this embodiment come from the examination room examinee data set composed of single-frame images of the examination room monitoring video under the standardized examination room environment obtained by means of crawlers on the Internet. In this embodiment, 700 images in the data set are used as the training set, and 180 images As a test set, and the training set does not overlap with the test set.

[0053] exist Figure 1~4 Among them, the rapid detection method of the examinee's position in the examination room of the improved SSD double network of the present embodiment is made up of following steps:

[0054] (1) Image preprocessing

[0055] Select 700 training sample images and 180 test sample images from the image data set, and use bilinear interpolation to normalize the selected images to 300×300 by pixel size.

[0056] (2) Build a dynamic threshold SSD network

[0057] (a) The SSD network structure is used as the initial structure of the dynamic t...

Embodiment 2

[0087] The images in this embodiment come from the examination room examinee data set composed of single-frame images of the examination room monitoring video under the standardized examination room environment obtained by means of crawlers on the Internet. In this embodiment, 700 images in the data set are used as the training set, and 180 images As a test set, and the training set does not overlap with the test set.

[0088] The fast detection method of the examinee's position in the examination room of the improved SSD double network of the present embodiment is made up of following steps:

[0089] (1) Image preprocessing

[0090] Select 700 training sample images and 180 test sample images from the image data set, and use bilinear interpolation to normalize the selected images to 250×250 according to the pixel size.

[0091] (2) Build a dynamic threshold SSD network

[0092] (a) The SSD network structure is used as the initial structure of the dynamic threshold SSD network...

Embodiment 3

[0113] The images in this embodiment come from the examination room examinee data set composed of single-frame images of the examination room monitoring video under the standardized examination room environment obtained by means of crawlers on the Internet. In this embodiment, 700 images in the data set are used as the training set, and 180 images As a test set, and the training set does not overlap with the test set.

[0114] The fast detection method of the examinee's position in the examination room of the improved SSD double network of the present embodiment is made up of following steps:

[0115] (1) Image preprocessing

[0116] Select 700 training sample images and 180 test sample images from the image data set, and use bilinear interpolation to normalize the selected images to 500×500 according to the pixel size.

[0117] (2) Build a dynamic threshold SSD network

[0118] (a) The SSD network structure is used as the initial structure of the dynamic threshold SSD netwo...

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 an improved SSD dual-network examination room examinee position rapid detection method. The method comprises the steps of image preprocessing, construction of a dynamic threshold SSD network, construction of an up-sampling SSD network, training of improved SSD dual networks and testing of a test sample image. On the basis of an SSD network structure, according to the characteristics of examinee data sets in an examination room, a method for dynamically adjusting an intersection-parallel ratio threshold value is constructed, the positive sample size of small targets in the SSD network is increased, an up-sampling layer is added into the SSD network, the image characteristics of the small targets are enhanced, and the image target detection rate is improved in combination with the thought of a parallel dual-network structure. Compared with the prior art, the method has the advantages of good robustness, high accuracy and the like, and is suitable for detecting examinees and invigilating teachers in an examination scene.

Description

technical field [0001] The invention belongs to the technical field of image processing and target detection, and in particular relates to the identification, positioning and population counting of examinees in a single-frame image obtained from a monitoring video of a standardized examination room. Background technique [0002] Examination is an important link in teaching activities, the main way to measure and judge the academic level of students, and an important means for the assessment and selection of various talents in our country. Standardized examination rooms have played an important role in various important national examinations in recent years. If you only rely on manual real-time observation of surveillance video for identification and judgment, it is inevitable that there will be misjudgments or omissions due to fatigue caused by long-term, uninterrupted continuous work. Therefore, advanced computer vision technology is integrated into standardized examination...

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): G06K9/00G06K9/62G06T7/70G06N3/04G06N3/08G06Q50/20
CPCG06T7/70G06N3/084G06Q50/205G06V20/52G06N3/045G06F18/214
Inventor 马苗陶丽丽裴炤高子昂
Owner SHAANXI NORMAL UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap