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

Test paper correction method based on image feature matching and alignment

A technology of image features and matching pairs, applied in the field of image processing, can solve the problems of not being able to find similar questions, not caring about the location of answers, and inaccurate cutting of questions, so as to achieve accurate corrections and avoid inaccurate positioning.

Pending Publication Date: 2021-07-09
HANGZHOU YUNTI TECH CO LTD
View PDF5 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] However, the online test paper correction method based on OCR recognition does not care whether the answer position is correct, and sometimes similar questions cannot be found due to random photos
And it is prone to the problem of inaccurate topic cutting based on model prediction

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
  • Test paper correction method based on image feature matching and alignment
  • Test paper correction method based on image feature matching and alignment
  • Test paper correction method based on image feature matching and alignment

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0028] A test paper correction method based on image feature matching and alignment, such as figure 1 As shown, the method includes the following steps S1 to S6:

[0029] S1, acquiring the original test paper image to be identified and the target test paper image;

[0030] Wherein, the original test paper image is collected by a camera, and the target test paper image is a test paper template PDF page.

[0031] S2, using an image extraction algorithm to extract the image features of the original test paper image and the target test paper image;

[0032] Preferably, the image extraction algorithm adopts the SIFT algorithm. The SIFT algorithm is invariant to image rotation and scale changes, has strong adaptability to changes in three-dimensional viewing angles and illumination changes, and is capable of performing extractions on massive amounts of information. Extract features quickly and accurately.

[0033] Among them, the specific feature extraction process is as follows:...

Embodiment 2

[0041] Taking one page of a test paper as an example, the specific process of the test paper correction method based on image feature matching and alignment is illustrated.

[0042] Step 1: Capture the original test paper image through the camera, and use the test paper template PDF page for the target test paper image; see figure 2 and image 3 .

[0043] Step 2: using the SIFT algorithm to extract the image features of the original test paper image and the target test paper image; the SIFT algorithm has invariance to image rotation and scale changes, and has strong adaptability to three-dimensional viewing angle changes and illumination changes, and It can quickly and accurately extract features from massive amounts of information.

[0044] Step 3: Set a certain threshold for the number of matching points, use the brute force matching algorithm to match the extracted image features, if it is lower than the threshold, output two images that do not correspond, otherwise out...

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 test paper correction method based on image feature matching and alignment. The method comprises the following steps: S1, obtaining an original test paper image to be identified and a target test paper image; S2, extracting image features of the original test paper image and the target test paper image by using an image extraction algorithm; S3, setting a certain threshold value for the number of matching points, and matching the extracted image features by using a matching algorithm; S4, obtaining a converted homography matrix by using an iterative algorithm through the matched image features, and outputting an aligned image by transforming the shot image according to the homography matrix; S5, digging out the added part of the image through the filled fixed pixel value; and S6, positioning on the original test paper image by using the regional position of each answer in the target test paper image. According to the method, correction is accurate, and the problem that similar questions cannot be searched is avoided.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a test paper correction method based on image feature matching and alignment. Background technique [0002] With the emergence of the epidemic in 2020, online education has developed rapidly. Many companies have begun to develop software such as online AI homework correction to improve the online education chain and provide convenience for users. This way of correcting test papers online has greatly improved work efficiency. [0003] In the existing technology, the common test paper correction is directly based on the image taken by the user to detect the questions, and then perform OCR recognition for each question, including the recognition of the questions and the detection and recognition of the handwritten answers, and then search for the most similar in the question bank. questions and match the correct or incorrect answers. [0004] However, the online test pape...

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): G06K9/00G06K9/46G06K9/62
CPCG06V30/414G06V30/418G06V10/462G06V10/757G06V30/10
Inventor 肖刚杨超周伟陈立
Owner HANGZHOU YUNTI TECH CO LTD