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

A text image line segmentation method and system based on signal processing

A text image and signal processing technology, applied in the field of image processing, can solve the problems of high computational complexity and poor analysis of handwritten text images, and achieve the effect of reducing line segmentation errors, reducing influence, and improving accuracy

Active Publication Date: 2020-07-14
GUANGZHOU SHIYUAN ELECTRONICS CO LTD
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] In order to solve the above-mentioned technical problems of poor analysis effect of handwritten text image or high computational complexity, the embodiment of the present invention proposes a text image line segmentation method and system based on signal processing

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 text image line segmentation method and system based on signal processing
  • A text image line segmentation method and system based on signal processing
  • A text image line segmentation method and system based on signal processing

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0064] Embodiment 1 of the present invention proposes a text image line segmentation method based on signal processing, such as figure 1 As shown, the method includes:

[0065] Step S101: cutting the text image along a direction perpendicular to the writing direction of the text line in the text image, and dividing it into multiple sub-text images;

[0066] Wherein, the number of sub-text images can be selected according to factors such as the size of the text image and usage scenarios. The number of sub-text images can be adjusted in combination with subsequent segmentation effects, and the accuracy of segmentation can be improved through continuous adjustment.

[0067] The writing direction of the text line in the text image is generally from left to right, or from top to bottom. When the writing direction of the text line in the text image is from left to right, the cutting direction is the longitudinal direction of the text image; when the writing direction of the text l...

Embodiment 2

[0113] In this embodiment, on the basis of the text image line segmentation method based on signal processing proposed in Embodiment 1, step S109 is added, as Figure 6 shown.

[0114] Step S109: According to the dividing line determined in step S108, segment the sub-image of each text line, count the median value of the line height of the sub-images of all text lines that are segmented, and compare the line height of the sub-image of each text line with Compared with the median value of the statistics, if the line height of the sub-image of a certain text line is much larger or much smaller than the median value of the statistics, it indicates that the sub-image of the text line is an abnormal image, or that the sub-image of the text line is a blank line , the line of text is removed.

[0115] The text image line segmentation method based on signal processing proposed in Embodiment 2 of the present invention judges whether the segmented text line is a blank line according to...

Embodiment 3

[0122] In this embodiment, on the basis of the text image line segmentation method based on signal processing in Embodiment 1, step S110 is added, as Figure 9 shown.

[0123] Step S110: According to the dividing line determined in step S108, segment the sub-image of each text line, sum each row of pixels of the sub-image of each text line, obtain a summation sequence, count the summation sequence greater than Set the number of elements of the threshold. If the number is less than the set threshold, it indicates that the sub-image of the text line does not contain text and is a blank line, and the text line is eliminated.

[0124] The text image line segmentation method based on signal processing proposed in Embodiment 3 of the present invention judges whether the segmented text line is a blank line according to whether the image contains text, which can further improve the blank line removal effect.

[0125] In this embodiment, on the basis of the text image line segmentatio...

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 present invention relates to a text image line segmentation method and system based on signal processing, wherein the method includes: cutting the text image into a plurality of sub-text images; for each sub-text image, obtaining the row pixel summation value of each pixel row, Composing the row pixel summation sequence of the sub-text image; performing low-pass filtering and band-pass filtering on the row pixel summation sequence of each sub-text image; calculating the local maximum value of the row pixel summation sequence of each sub-text image and filtering ; Connect the pixels corresponding to the local maximum values ​​of the adjacent sub-text images along the writing direction of the text line in the text image in turn after filtering to obtain the text line direction line of the entire text image; calculate the direction of two adjacent text lines Text image gradient between lines, which determines the dividing line between two adjacent text lines. The text image line segmentation method and system based on signal processing proposed by the embodiments of the present invention can reduce the influence of noise on local maximum values ​​and improve the accuracy of segmentation.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to a text image line segmentation method and system based on signal processing. Background technique [0002] A text image refers to an image containing text. Segmenting a text image into text lines is the basic task of text image analysis projects such as word retrieval, keyword search, and text alignment. The text contained in the text image may be a machine-printed text, or a text naturally handwritten by a writer on a writing surface. [0003] Line segmentation algorithms are frequently used techniques in text image analysis. The line segmentation algorithm is to segment multiple sub-images from a text image, and each sub-image contains only one line of text. When analyzing a text image, the sub-image containing a single line of text obtained through the line segmentation algorithm is easier to process than the text image, and it is easy to obtain better a...

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): G06K9/00G06K9/34
CPCG06V30/40G06V10/267G06F2218/00
Inventor 李德健
Owner GUANGZHOU SHIYUAN ELECTRONICS CO LTD