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

A Document Image Quality Evaluation Method

A document image and quality evaluation technology, applied in the field of image processing, can solve problems such as inability to evaluate document image quality and incomplete correlation of document image quality.

Active Publication Date: 2022-07-22
NANCHANG HANGKONG UNIVERSITY
View PDF10 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, OCR results sometimes do not fully correlate with the quality of the document image, resulting in failure to provide a proper quality rating for the document image

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
  • A Document Image Quality Evaluation Method
  • A Document Image Quality Evaluation Method
  • A Document Image Quality Evaluation Method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0032] like figure 1 As shown in the method flow chart, the specific implementation process of the embodiment is as follows:

[0033] (1) Generate a document image quality comparison network model through document image training in the document image dataset.

[0034] The document image quality label in the used document image dataset is the average subjective score (MeanOpinionScore, MOS) in the range of [0, 10]. When the MOS value is closer to 0, the image quality is worse, and vice versa, the closer it is to 10, the image quality. the better. The document image consists of two types of document images X with different image quality labels in the document image dataset. 1 , X 2 composition.

[0035]The document image quality comparison network model consists of two branches, the two branches are ResNet50 networks with shared weights and the same structure, the input of a single branch is the document image X and its quality label y, and the output of a single branch f(X ...

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 document image quality evaluation method. A plurality of image blocks are cut out from the document image and grouped into the same small batch, and are respectively input into the document image quality evaluation network model for training. After the training of the network model is completed, a plurality of image blocks are intercepted from the image of the document to be evaluated and input to the network model for evaluating the image quality of the document to obtain the predicted quality scalar of the image of the document to be evaluated. Specifically, it includes the following steps: generating a document image quality comparison network model through document image training in the document image dataset; using the document image quality comparison network model to fine-tune the document image quality evaluation network model; using the document image dataset training to generate document image quality Evaluate the network model; use the document image quality evaluation network model to evaluate the document image to be evaluated, and the output of the network model is the predicted quality scalar of the document image. This method can obtain the objective document image quality evaluation results that conform to the visual characteristics of human eyes.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to a document image quality evaluation method. Background technique [0002] As an important source of visual information, images convey rich information to people. Storing documents as images captured by mobile devices is one of the main ways of information preservation today. However, in the process from document image acquisition to display, there are many inevitable factors that will cause document image distortion, such as out-of-focus, motion blur, noise, compression and so on. Document image distortion will directly lead to the loss of document image information. At this time, using an effective document image quality evaluation method can provide people with a scalar that reflects the document image quality under a certain standard, and then help people to take effective measures from the acquisition to the display of the image to ensure the image qual...

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): G06T7/00G06N3/04G06N3/08
CPCG06T7/0002G06N3/084G06T2207/20081G06T2207/30168G06T2207/30176G06N3/045
Inventor 谌青云杨词慧李正浩李克伟陈财森
Owner NANCHANG HANGKONG UNIVERSITY