WO2018103226A1 - 一种基于图片生成笔迹格式的文件的方法及装置 - Google Patents

一种基于图片生成笔迹格式的文件的方法及装置 Download PDF

Info

Publication number
WO2018103226A1
WO2018103226A1 PCT/CN2017/077154 CN2017077154W WO2018103226A1 WO 2018103226 A1 WO2018103226 A1 WO 2018103226A1 CN 2017077154 W CN2017077154 W CN 2017077154W WO 2018103226 A1 WO2018103226 A1 WO 2018103226A1
Authority
WO
WIPO (PCT)
Prior art keywords
handwriting
color value
picture
points
pixel
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2017/077154
Other languages
English (en)
French (fr)
Inventor
李纯冬
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guangzhou Shizhen Electronics Co Ltd
Guangzhou Shiyuan Electronics Thecnology Co Ltd
Original Assignee
Guangzhou Shizhen Electronics Co Ltd
Guangzhou Shiyuan Electronics Thecnology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Guangzhou Shizhen Electronics Co Ltd, Guangzhou Shiyuan Electronics Thecnology Co Ltd filed Critical Guangzhou Shizhen Electronics Co Ltd
Publication of WO2018103226A1 publication Critical patent/WO2018103226A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10Character recognition
    • G06V30/32Digital ink
    • G06V30/333Preprocessing; Feature extraction
    • G06V30/347Sampling; Contour coding; Stroke extraction

Definitions

  • the present invention relates to the field of picture processing, and in particular, to a method and apparatus for generating a file based on a picture in a handwriting format.
  • the written handwriting is generally drawn on the background.
  • Document formats such as pictures and PDFs.
  • the advantage is that when the file is opened again, the file or the handwriting that has been written can be edited twice.
  • the disadvantage is that the file compatibility of the handwriting format is poor, and generally only a specific writing is performed.
  • the software can be opened, and it is not convenient to view the written content in an environment where the writing software is not installed.
  • the writing software cannot perform the second operation on the original handwriting, and can only edit the image to make the file
  • the content before and after is inconsistent, and repeated storage as a picture will also result in a decrease in the clarity of the file, and the user experience is not good.
  • an object of the present invention is to provide a method and apparatus for generating a file based on a picture in a handwriting format, which can generate a file in a handwriting format according to a picture, thereby facilitating a user to perform a second operation on the handwriting.
  • the present invention provides a method for generating a file based on a picture in a handwriting format, comprising:
  • a file in a handwriting format is generated according to the background image and the set of handwriting points.
  • the obtaining, according to the color value of the pixel points of the handwriting identified by the composition, obtaining at least one handwriting color value is specifically:
  • the obtaining, according to the color value of the pixel points of the handwriting identified by the composition, obtaining at least one handwriting color value is specifically:
  • a color value in which the ratio of the number of corresponding pixel points to the total number of recognized pixel points exceeds a predetermined threshold is marked as a handwriting color value.
  • the generating a background image according to the set of handwriting points and the picture comprises:
  • the color value of the queried pixel is replaced with the background color value to generate a background image.
  • the generating a background image according to the set of handwriting points and the picture comprises:
  • a background image is generated by replacing the color value of the queried pixel point with the replacement color value.
  • the generating a replacement color value according to a color value of an adjacent pixel adjacent to the queried pixel point specifically:
  • the color values of the remaining adjacent pixels are blended to generate a replacement color value.
  • the generating a replacement color value according to a color value of a pixel point adjacent to the queried pixel point specifically:
  • the replacement color value is determined based on the number of occurrences of the remaining color values of adjacent pixel points.
  • the present invention also provides an apparatus for generating a file based on a picture in a handwriting format, comprising:
  • a handwriting recognition unit for identifying a picture to be processed and obtaining the recognized handwriting
  • a handwriting color value generating unit configured to obtain at least one handwriting color value according to a color value of a pixel point of the handwriting that constitutes the recognition
  • a handwriting point set generating unit configured to acquire a pixel point in the picture that has the same color value as the handwriting color value, and generate a handwriting point set
  • a background image generating unit configured to generate a background image based on the set of handwriting points and the image
  • a handwriting file generating unit configured to generate a file in a handwriting format according to the background image and the set of handwriting points.
  • the handwriting color value generating unit specifically includes:
  • a first statistic module configured to acquire color values of all pixel points constituting the recognized handwriting, and count the number of pixel points corresponding to each color value
  • the first marking module is configured to mark the color value corresponding to the most pixel point as a handwriting color value.
  • the background image generating unit specifically includes:
  • a background color value obtaining module configured to acquire a background color value of the image
  • a first query module configured to query a pixel in the picture that is located in the set of handwriting points
  • a first color value replacement module configured to replace the color value of the queried pixel point with the background color value to generate a background image.
  • the method and device for generating a file based on a picture in a handwriting format provided by the present invention, by identifying a handwriting in the picture, and obtaining a handwriting color value according to the recognized handwriting, and then using the handwriting color value to obtain all the handwriting in the picture. Pixels, generate a set of handwriting points, and restore the background image according to the set of handwriting points, and finally generate a file of the handwriting format according to the set of handwriting points and the background image, thereby converting the image into a file of the handwriting format.
  • the content can be viewed directly by opening the original image, and an environment or terminal (such as an electronic whiteboard) having a file that can open the handwriting format is installed.
  • the second hand editing can be performed by using the handwriting in the file format converted into a picture, which improves the compatibility of the file and facilitates the operation and experience of the user.
  • FIG. 1 is a schematic flowchart diagram of a method for generating a file based on a picture in a handwriting format according to an embodiment of the present invention.
  • FIG. 2 is a schematic diagram of a picture to be processed according to an embodiment of the present invention.
  • Fig. 3 is a schematic view showing the handwriting recognition of the picture shown in Fig. 2.
  • FIG. 4 is a schematic diagram of acquiring adjacent pixel points adjacent to the queried pixel.
  • FIG. 5 is another schematic diagram of acquiring adjacent pixel points adjacent to the queried pixel.
  • FIG. 6 is a schematic structural diagram of an apparatus for generating a file based on a picture in a handwriting format according to an embodiment of the present invention.
  • the present invention provides a method for generating a file based on a picture in a handwriting format, which includes at least the following steps:
  • the picture to be processed may be saved by using a writing software on the electronic whiteboard to save the background image with the handwriting as a picture format, or may include any handwriting information or handwriting information.
  • the picture is not specifically limited in the present invention.
  • the handwriting or the handwriting generally includes a plurality of horizontal, vertical, circular, and the like elements, so that most of the handwriting in the picture can be identified using the Hough transform-based line and circle detection algorithm.
  • Hough transform is one of the classical methods in image transformation. It is mainly used to separate geometric shapes with certain identical features, such as lines, circles, etc., from images.
  • other well-known graphic separation techniques may also be used to identify handwriting from the picture, which is not specifically limited in the present invention.
  • the user may also receive modification of the recognition result, such as adding missing handwriting, or removing the erroneously recognized portion. If the area to be replenished or removed is small, it can be modified by using Linyi. If the area is large, it can be modified by circle selection to make the recognized handwriting more accurate.
  • step S102 includes:
  • S1021 Acquire color values of pixel points constituting the recognized handwriting, and count the number of pixel points corresponding to each color value.
  • the pixel points of the non-handwriting (such as the pixel of the background adjacent to the handwriting or other interfering pixel points) are inevitably introduced in the process of picture recognition or image separation, the color of the part of the pixel is inevitable.
  • the value is not the same as the color value of the handwriting, so the color value of the real handwriting needs to be obtained from the recognized handwriting.
  • the introduced non-writing pixel points are inevitably small, the pixel points corresponding to each color value can be counted. The number of colors and the color value corresponding to the pixel with the largest number is set as the hand color value of the picture.
  • black is set to the The handwriting color value of the image.
  • the present invention is applicable to the case where the color value of the handwriting in the picture is relatively simple, in particular, all the handwriting in the picture is drawn or written using the same color value.
  • step S102 includes:
  • S1023 Acquire color values of all the pixel points constituting the recognized handwriting, and count the number of pixel points corresponding to each color value.
  • the picture may include handwriting of different colors at the same time, in this case, the number of pixels corresponding to each color value is obtained, and the ratio of the number to the total number of identified pixel points exceeds a predetermined value.
  • the color values of the threshold are all marked as handwriting color values.
  • the threshold is 20%
  • the number of all pixels of the recognized handwriting is 1000
  • the number of pixels having a color value of black is 600
  • the number of pixels having a color value of red is For 300
  • the number of pixels of other color values is 100
  • both black and red are marked as handwriting color values.
  • the threshold can be obtained according to the recognition accuracy of the adopted image separation algorithm.
  • a larger or nearly all handwriting can be extracted from the picture according to the handwriting color value.
  • you can read the color of each pixel of the image row by row by taking the top left corner of the image as the origin O(0,0), the width of the image as the x-axis boundary, and the height of the image as the y-axis boundary.
  • the color value of the pixel point P1(x l , y l ) is the same as the handwriting color value (which may be one or any of a plurality of handwriting color values)
  • the point P1 (x l , y l ) is added . ) to the set of handwriting points until the full picture is read.
  • the document in the handwriting format includes a background image and a handwriting point, so in order to obtain a file in the handwriting format, it is also necessary to reconstruct or restore the background image. specifically:
  • step S104 specifically includes:
  • the color of the background image is single, such as a white background image or a light green background image.
  • it is only necessary to obtain the background color value of the background image then traverse the image, query the pixel points located in the set of the handwriting points, and then replace the color value of the queried pixel point with the background color value. Reconstruct the background image.
  • step S104 specifically includes:
  • the background image may include a background pattern or be composed of a plurality of colors.
  • S105 Generate a file in a handwriting format according to the background image and the set of handwriting points.
  • the background image is first drawn, and then the pixel points in the handwriting point set are drawn on the background image by the corresponding handwriting color value.
  • you can save the file in the handwriting format (such as the file in ENB format).
  • the user can open the file in the handwriting format through a specific writing software and directly edit the handwriting.
  • the method for generating a document based on a picture in a handwriting format obtains a handwriting color value according to the recognized handwriting by identifying a handwriting in the image, and then obtaining the image by using a handwriting color value.
  • the pixels of all the handwritings are generated, the set of handwriting points is generated, and the background image is restored according to the set of handwriting points, and finally the file of the handwriting format is generated according to the handwriting point set and the background image, thereby converting the image into a file of the handwriting format.
  • the content can be viewed directly by opening the original image, and an environment or terminal (such as an electronic whiteboard) having a file that can open the handwriting format is installed.
  • the second hand editing can be performed by using the handwriting in the file format converted into a picture, which improves the compatibility of the file and facilitates the operation and experience of the user.
  • step S1045 specifically includes:
  • the color values of four adjacent pixel points adjacent to the pixel point may be acquired, or 8 in a square area centered on the pixel point may be acquired.
  • the color values of adjacent pixels (as shown in Figure 5).
  • the pixels located in the set of the handwriting points in the adjacent pixel points are removed, so as to prevent the pixels from interfering with the recovery of the background image. effect.
  • the color values of the remaining adjacent pixels are blended to generate a replacement color value. Since the color of the pattern is generally continuous transition, the original color value of the position occupied by the pixel point can be obtained by mixing the color values of the adjacent pixel points, wherein the mixing can be a direct pixel point.
  • the RGB value is mixed, or the RGB value can be converted into a hue and then mixed. In general, converting RGB to hue and then mixing has a better effect.
  • step S1045 specifically includes:
  • S10454 Acquire color values of adjacent pixel points adjacent to the queried pixel point.
  • the color values of four adjacent pixel points adjacent to the pixel point may be acquired, and the colors of eight adjacent pixel points in the square area centered on the pixel point may also be acquired. value.
  • the pixels located in the set of the handwriting points in the adjacent pixel points are removed, so as to avoid interference of the adjacent pixel points. Restore the effect of the background image.
  • S10456 Determine a replacement color value according to the number of occurrences of the color values of the remaining adjacent pixel points.
  • the upper P up (x l , y l ) and the lower P down (x) of the point P l are read.
  • the replacement color value takes the same color value as the three values
  • the replacement color value takes two color values of the same value
  • the color value of the replacement color value is a mixed color value of two color values
  • C u , C d , C l , and C r are different, it means that it is an isolated handwriting point, which can be removed as a noise point, removed in the handwriting point set, and replaced with a color value of four values. Mix color values.
  • an embodiment of the present invention further provides an apparatus 100 for generating a file based on a picture in a handwriting format, including:
  • the handwriting recognition unit 10 is configured to recognize the picture to be processed, and obtain the recognized handwriting.
  • the handwriting color value generating unit 20 is configured to obtain at least one handwriting color value according to the color value of the pixel points of the handwriting that constitute the recognition.
  • the handwriting point set generating unit 30 is configured to acquire pixel points in the picture that have the same color value as the handwriting color value, and generate a handwriting point set.
  • the background image generating unit 40 is configured to generate a background image based on the set of handwriting points and the image.
  • the handwriting file generating unit 50 is configured to generate a file in a handwriting format according to the background image and the set of handwriting points.
  • the handwriting color value generating unit 20 specifically includes:
  • a first statistic module 21 configured to acquire color values of all pixel points constituting the recognized handwriting, and count the number of pixel points corresponding to each color value;
  • the first marking module 22 is configured to mark the color value corresponding to the most pixel point as a handwriting color value.
  • the handwriting color value generating unit 20 specifically includes:
  • the second statistic module 23 is configured to acquire color values of pixel points constituting the recognized handwriting, and count the number of pixel points corresponding to each color value.
  • the second marking module 24 is configured to mark, as the handwriting color value, a color value that exceeds a predetermined threshold value between the number of corresponding pixel points and the total number of identified pixel points.
  • the background image generating unit 40 specifically includes:
  • a background color value obtaining module 41 configured to acquire a background color value of the image
  • the first query module 42 is configured to query pixel points in the picture that are located in the set of handwriting points;
  • the first color value replacement module 43 is configured to replace the color value of the queried pixel point with the background color value to generate a background image.
  • the background image generating unit 40 specifically includes:
  • the second query module 44 is configured to query pixel points in the picture that are located in the set of handwriting points.
  • the replacement color value generating module 45 is configured to generate a replacement color value according to a color value of at least 4 adjacent pixel points adjacent to the queried pixel point.
  • the second color value replacement module 46 is configured to replace the color value of the queried pixel point with the replacement color value to generate a background image.
  • the replacement color value generating module 45 specifically includes:
  • the first adjacent color value obtaining sub-module 451 is configured to acquire color values of at least four adjacent pixel points adjacent to the queried pixel point.
  • the first removal sub-module 452 is configured to remove adjacent pixel points located in the set of handwriting points.
  • the color value mixing sub-module 453 is configured to mix the color values of the remaining adjacent pixel points to generate a replacement color value.
  • the replacement color value generating module 45 specifically includes:
  • the second adjacent color value obtaining sub-module 454 is configured to acquire color values of at least four adjacent pixel points adjacent to the queried pixel point.
  • a second removal sub-module 455, configured to remove adjacent pixel points located in the set of handwriting points
  • the color value setting sub-module 456 is configured to determine a replacement color value according to the number of occurrences of the color values of the remaining adjacent pixel points.
  • the apparatus 100 for generating a document based on a picture in a handwriting format recognizes a handwriting in the picture by the handwriting recognition unit 10, and obtains a handwriting color value according to the recognized handwriting by the handwriting color value generating unit 20, and then The handwriting point set generating unit 30 obtains the pixel points of all the handwritings in the picture by using the handwriting color value, generates a handwriting point set, and then restores the background image according to the handwriting point set by the background image generating unit 40, and finally the handwriting file generating unit 50 according to The handwriting point set and the background image generate a file in a handwriting format, thereby converting the picture into a file in a handwriting format.
  • the content can be viewed directly by opening the original image, and an environment or terminal (such as an electronic whiteboard) having a file that can open the handwriting format is installed.
  • the second hand editing can be performed by using the handwriting in the file format converted into a picture, which improves the compatibility of the file and facilitates the operation and experience of the user.
  • the storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Processing Or Creating Images (AREA)
  • Image Processing (AREA)
  • User Interface Of Digital Computer (AREA)

Abstract

一种基于图片生成笔迹格式的文件的方法,包括:对待处理的图片进行识别,获得识别到的笔迹(S101);根据组成识别到的所述笔迹的像素点的颜色值获得至少一个笔迹颜色值(S102);获取所述图片中与所述笔迹颜色值具有相同颜色值的像素点,生成笔迹点集合(S103);基于所述笔迹点集合及所述图片生成背景图(S104);根据所述背景图及所述笔迹点集合生成笔迹格式的文件(S105)。一种基于图片生成笔迹格式的文件的装置,可将图片转换为可编辑的笔迹格式的文件,方便用户对笔迹进行编辑。

Description

一种基于图片生成笔迹格式的文件的方法及装置 技术领域
本发明涉及图片处理领域,尤其涉及一种基于图片生成笔迹格式的文件的方法及装置。
背景技术
利用电子白板进行书写时,一般会将书写的笔迹绘制在背景上,在书写完成后,可以选择将笔迹与背景保存为特定的笔迹格式的文件(如ENB格式的文件)或直接保存为通用的图片、PDF等文档格式。
对于保存为笔迹格式的文件,其好处在于,当再次打开该文件时,可对文件或已经书写的笔迹进行二次编辑,缺点是这种笔迹格式的文件兼容性较差,一般只有特定的书写软件才能够打开,不方便在没有安装书写软件的环境下查看书写内容。而如果直接保存为通用的图片、PDF等文档格式,虽然方便了在各种环境下的查看和传输,但是书写软件无法对原有的笔迹进行二次操作,只能够对图片进行编辑,使得文件的前后内容不一致,且反复保存为图片还会造成文件清晰度的下降,用户体验不佳。
发明内容
针对上述问题,本发明的目的在于提供一种基于图片生成笔迹格式的文件的方法及装置,可根据图片生成笔迹格式的文件,方便用户对笔迹进行二次操作。
本发明提供了一种基于图片生成笔迹格式的文件的方法,包括:
对待处理的图片进行识别,获得识别到的笔迹;
根据组成识别到的所述笔迹的像素点的颜色值获得至少一个笔迹颜色值;
获取所述图片中与所述笔迹颜色值具有相同颜色值的像素点,生成笔迹点集合;
基于所述笔迹点集合及所述图片生成背景图;
根据所述背景图及所述笔迹点集合生成笔迹格式的文件。
优选地,所述根据组成识别到的所述笔迹的像素点的颜色值获得至少一个笔迹颜色值具体为:
获取组成识别到的所述笔迹的所有像素点的颜色值,并统计与每个颜色值对应的像素点的个数;
将与个数最多的像素点对应的颜色值设置为笔迹颜色值。
优选地,所述根据组成识别到的所述笔迹的像素点的颜色值获得至少一个笔迹颜色值具体为:
获取组成识别到的所述笔迹的所有像素点的颜色值,并统计与每个颜色值对应的像素点的个数;
将对应的像素点的个数与识别到的像素点总数的比值超过预定阈值的颜色值标记为笔迹颜色值。
优选地,所述根据所述笔迹点集合及所述图片生成背景图,具体包括:
获取所述图片的背景颜色值;
查询所述图片中位于所述笔迹点集合内的像素点;
将查询到的所述像素点的颜色值替换为所述背景颜色值,生成背景图。
优选地,所述根据所述笔迹点集合及所述图片生成背景图,具体包括:
查询所述图片中位于所述笔迹点集合内的像素点;
根据与查询到的所述像素点相邻的相邻像素点的颜色值生成替换颜色值;
将查询到的所述像素点的颜色值替换为所述替换颜色值,生成背景图。
优选地,所述根据与查询到的所述像素点相邻的相邻像素点的颜色值生成替换颜色值,具体为:
获取与查询到的所述像素点相邻的相邻像素点的颜色值;
去除位于所述笔迹点集合内的相邻像素点;
对余下的相邻像素点的颜色值进行混合,生成替换颜色值。
优选地,所述根据与查询到的所述像素点相邻的像素点的颜色值生成替换颜色值,具体为:
获取与查询到的所述像素点相邻的相邻像素点的颜色值;
去除位于所述笔迹点集合内的相邻像素点;
根据余下的相邻像素点的颜色值的出现次数确定替换颜色值。
本发明还提供了一种基于图片生成笔迹格式的文件的装置,包括:
笔迹识别单元,用于对待处理的图片进行识别,获得识别到的笔迹;
笔迹颜色值生成单元,用于根据组成识别到的所述笔迹的像素点的颜色值获得至少一个笔迹颜色值;
笔迹点集合生成单元,用于获取所述图片中与所述笔迹颜色值具有相同颜色值的像素点,生成笔迹点集合;
背景图生成单元,用于基于所述笔迹点集合及所述图片生成背景图;
笔迹文件生成单元,用于根据所述背景图及所述笔迹点集合生成笔迹格式的文件。
优选地,所述笔迹颜色值生成单元具体包括:
第一统计模块,用于获取组成识别到的所述笔迹的所有像素点的颜色值,并统计与每个颜色值对应的像素点的个数;
第一标记模块,用于将与个数最多的像素点对应的颜色值标记为笔迹颜色值。
优选地,所述背景图生成单元具体包括:
背景颜色值获取模块,用于获取所述图片的背景颜色值;
第一查询模块,用于查询所述图片中位于所述笔迹点集合内的像素点;
第一颜色值替换模块,用于将查询到的所述像素点的颜色值替换为所述背景颜色值,生成背景图。
本发明提供的基于图片生成笔迹格式的文件的方法及装置,通过识别所述图片中的笔迹,并根据识别到的笔迹获得笔迹颜色值,再利用笔迹颜色值获得所述图片中的所有笔迹的像素点,生成笔迹点集合,并根据笔迹点集合恢复出背景图,最后根据笔迹点集合和背景图生成笔迹格式的文件,从而将图片转换为笔迹格式的文件。一方面,对于没有安装有可以打开笔迹格式的文件的环境或终端,可以直接通过打开原有的图片进行内容的查看,而对于安装有可以打开笔迹格式的文件的环境或终端(如电子白板),可以用对图片转换成的笔迹格式的文件中的笔迹进行二次编辑,提高了文件的兼容性,方便了用户的操作和体验。
附图说明
为了更清楚地说明本发明的技术方案,下面将对实施方式中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图仅仅是本发明的一些实施方式,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图获得其他的附图。
图1是本发明实施例提供的基于图片生成笔迹格式的文件的方法的流程示意图。
图2是本发明实施例提供的待处理的图片的示意图。
图3是对图2所示的图片进行笔迹识别后的示意图。
图4是获取与查询到的像素点相邻的相邻像素点的一种示意图。
图5是获取与查询到的像素点相邻的相邻像素点的另一种一种示意图。
图6是本发明实施例提供的基于图片生成笔迹格式的文件的装置的结构示意图。
具体实施方式
下面将结合本发明实施例中的附图,对本发明实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例仅仅是本发明一部分实施例,而不是 全部的实施例。基于本发明中的实施例,本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例,都属于本发明保护的范围。
请参阅图1,本发明提供了一种基于图片生成笔迹格式的文件的方法,其至少包括如下步骤:
S101,对待处理的图片进行识别,获得识别到的笔迹。
如图2所示,在本发明实施例中,所述待处理的图片可由电子白板上的书写软件将绘制有笔迹的背景图保存为图片格式得到,也可以是任意包含有字迹信息或笔迹信息的图片,本发明不做具体限定。
在本发明实施例中,考虑到字迹或笔迹一般包含有比较多的横、竖、圆等元素,因此可使用基于霍夫变换的直线、圆检测算法识别出图片中的大部分笔迹。其中,霍夫变换是图像变换中的经典手段之一,主要用来从图像中分离出具有某种相同特征的几何形状,如,直线,圆等。当然,应当理解的是,在本发明实施例中,也可采用其他的公知的图文分离技术从所述图片中识别得到笔迹,本发明不做具体限定。
需要说明的是,由于笔迹还常常包含有一些不规则的几何形状,比如非圆点、弯钩等,这些几何形状一般较难直接通过图文分离技术完全分离出来,因此,应当理解是的,识别出的笔迹仅是图片中的部分笔迹而非全部笔迹(如图3中颜色较黑的几何形状为未被识别出的笔迹)。
此外,还需要说明的是,在本发明实施例中,在识别完成后,还可接收用户对识别结果的修改,如补充遗漏的笔迹,或者去除错误识别的部分等。如果需要补充或者去除的区域较小,可以使用临摹的方式修改,如果区域较大,可以使用圈选的方式修改,以使识别得到的笔迹更加准确。
S102,根据组成识别到的所述笔迹的像素点的颜色值获得至少一个笔迹颜色值。
具体地,在一个实现方式中,步骤S102包括:
S1021,获取组成识别到的所述笔迹的像素点的颜色值,并统计与每个颜色值对应的像素点的个数。
S1022,将与个数最多的像素点对应的颜色值设置为笔迹颜色值。
在本发明实施例中,由于在图片识别或者图文分离过程中,不可避免引入非笔迹的像素点(如与笔迹相邻的背景的像素点或其他干扰像素点),这部分像素点的颜色值与笔迹的颜色值不一样,因此需要从识别到的笔迹中获取真正的笔迹的颜色值,由于引入的非笔迹的像素点必然占少数,因此,可统计与每个颜色值对应的像素点的个数,并将与个数最多的像素点对应的颜色值设置为所述图片的笔迹颜色值。
例如,假设识别到组成笔迹的所有像素点的数目为1000个,其中,颜色值为黑色的像素点的数目为900个,其他颜色的像素点的数目为100个,则将黑色设置为所述图片的笔迹颜色值。
需要说明的是,本发明适用于图片中的笔迹的颜色值较单一的情况,特别是图片中的所有笔迹均使用同一种颜色值绘制或书写。
在一个实现方式中,步骤S102包括:
S1023,获取组成识别到的所述笔迹的所有像素点的颜色值,并统计与每个颜色值对应的像素点的个数。
S1024,将对应的像素点的个数与识别到的像素点总数的比值超过预定阈值的颜色值标记为笔迹颜色值。
在一些情况中,所述图片中可能同时包括不同颜色的笔迹,则此时,获取与每个颜色值对应的像素点的个数,并将个数与识别到的像素点总数的比值超过预定阈值的颜色值均标记为笔迹颜色值。
例如,假设所述阈值为20%,且识别到的笔迹的所有像素点的数目为1000个,其中,颜色值为黑色的像素点的数目为600个,颜色值为红色的像素点的数目为300个,其他颜色值的像素点的数目为100个,则将黑色和红色均标记为笔迹颜色值。
需要说明的是,由于识别到的笔迹中不可避免的包含有非笔迹的像素点(如背景像素点)的颜色值,因此需要合理设置阈值来滤除这些干扰,一方面,如 果阈值设置得过小,则可能出现将这些非笔迹的像素点的颜色值也标记为笔迹颜色值,而如果阈值设置得比较大,一些占比比较小的笔迹可能会被滤除,影响最终生成的文件的效果。具体的阈值可根据采用的图文分离算法的识别准确率统计得到。
S103,获取所述图片中与所述笔迹颜色值具有相同颜色值的像素点,生成笔迹点集合。
在本发明实施例中,在获得所述图片的笔迹颜色值后,即可根据所述笔迹颜色值从所述图片中提取出更大部分或接近全部的笔迹。在提取时,可以以图片的左上角为原点O(0,0)、图片的宽度为x轴边界、以图片的高度为y轴边界,逐行逐列读取图片的每个像素点的颜色值,如果像素点P1(xl,yl)的颜色值与笔迹颜色值(可以是一个,也可以是多个笔迹颜色值中的任何一个)一样,则添加点P1(xl,yl)到笔迹点集合中,直到读取完整张图片。
需要说明的是,在所述笔迹点集合中,除了记录每个像素点的位置坐标外,还同时记录其对应的颜色值。
S104,基于所述笔迹点集合及所述图片生成背景图。
在本发明实施例中,笔迹格式的文件包括背景图和笔迹点,因此为了获得笔迹格式的文件,还需要重构或恢复出背景图。具体地:
在一个实施例中,步骤S104具体包括:
S1041,获取所述图片的背景颜色值。
S1042,查询所述图片中位于所述笔迹点集合内的像素点。
S1043,将查询到的所述像素点的颜色值替换为所述背景颜色值,生成背景图。
对于一些情况,背景图的颜色是单一的,比如白色的背景图或浅绿色的背景图。则此时,只需要获取背景图的背景颜色值,然后遍历图片,查询位于所述笔迹点集合内的像素点,再将查询到的所述像素点的颜色值替换为所述背景颜色值,就重构得到背景图。
在一个实施例中,步骤S104具体包括:
S1044,查询所述图片中位于所述笔迹点集合内的像素点。
S1045,根据与查询到的所述像素点相邻的至少4个相邻像素点的颜色值生成替换颜色值。
S1046,将查询到的所述像素点的颜色值替换为所述替换颜色值,生成背景图。
在一些情况下,背景图可能包括有背景图案或者由多种颜色构成,则此时,在恢复背景图时,需要查询所述图片中位于笔迹点集合内的像素点,然后根据与查询到的像素点相邻的像素点的颜色值生成替换颜色值,并将查询到的所述像素点的颜色值设置为所述替换颜色值,得到背景图。由于查询到的像素点的替换颜色值是由与其相邻的像素点的颜色值决定的,因而可尽量保证背景图的重构效果。
S105,根据所述背景图及所述笔迹点集合生成笔迹格式的文件。
在本发明实施例中,在得到笔迹点集合及所述背景图后,只要先绘制所述背景图,再在背景图上将笔迹点集合中的像素点以对应的所述笔迹颜色值绘制出来,就可以保存得到笔迹格式的文件(如ENB格式的文件)。用户可通过特定的书写软件打开所述笔迹格式的文件,并直接对笔迹进行编辑。
综上所述,本发明实施例提供的基于图片生成笔迹格式的文件的方法,通过识别所述图片中的笔迹,并根据识别到的笔迹获得笔迹颜色值,再利用笔迹颜色值获得所述图片中的所有笔迹的像素点,生成笔迹点集合,并根据笔迹点集合恢复出背景图,最后根据笔迹点集合和背景图生成笔迹格式的文件,从而将图片转换为笔迹格式的文件。一方面,对于没有安装有可以打开笔迹格式的文件的环境或终端,可以直接通过打开原有的图片进行内容的查看,而对于安装有可以打开笔迹格式的文件的环境或终端(如电子白板),可以用对图片转换成的笔迹格式的文件中的笔迹进行二次编辑,提高了文件的兼容性,方便了用户的操作和体验。
为了便于对本发明实施例的了解,下面对本发明的一些优选实施例做更进一步的描述。
第一个优选实施例:
优选地,步骤S1045具体包括:
S10451,获取与查询到的所述像素点相邻的相邻像素点的颜色值。
在本优选实施例中,可获取与所述像素点相邻的4个相邻像素点的颜色值(如图4所示),也可以获取与所述像素点为中心的正方形区域内的8个相邻像素点的颜色值(如图5所示)。
S10452,去除位于所述笔迹点集合内的相邻像素点。
在本优选实施例中,由于所述笔迹点集合内的像素点必然不属于背景,因而去除相邻像素点中的位于所述笔迹点集合内的像素点,以免这些像素点干扰恢复背景图的效果。
S10453,对余下的相邻像素点的颜色值进行混合,生成替换颜色值。
在本优选实施例中,将余下的相邻像素点的颜色值进行混合,生成替换颜色值。由于图案的颜色一般是连续过渡的,因而可通过对相邻像素点的颜色值进行混合来获取被查询到所述像素点占据的位置的原来的颜色值,其中,混合可以是直接用像素点的RGB值混合,也可以是将RGB值转换为色相后再混合,一般来说,将RGB转换为色相后再混合有比较好的效果。
第二个优选实施例:
优选地,步骤S1045具体包括:
S10454,获取与查询到的所述像素点相邻的相邻像素点的颜色值。
在本优选实施例中,可获取与所述像素点相邻的4个相邻像素点的颜色值,也可以获取与所述像素点为中心的正方形区域内的8个相邻像素点的颜色值。
S10455,去除位于所述笔迹点集合内的相邻像素点。
在本优选实施例中,由于笔迹点集合内的像素点必然不属于背景,因而去除相邻像素点中的位于所述笔迹点集合内的像素点,以免这些相邻像素点干扰 恢复背景图的效果。
S10456,根据余下的相邻像素点的颜色值的出现次数确定替换颜色值。
例如,如果像素点Pl(xl,yl)为位于笔迹点集合内的像素点,则读取该点Pl的上Pup(xl,yl-1)、下Pdown(xl,yl+1)、左Pleft(xl-1,yl)、右Pright(xl+1,yl)四个方向相邻像素点Cu,Cd,Cl,Cr,且假设这四个相邻像素点均不位于笔迹点集合内,则进行判断:
如果Cu,Cd,Cl,Cr的颜色值都一样,则替换颜色值取该颜色值;
如果Cu,Cd,Cl,Cr的颜色值有3个一样,则替换颜色值取与3个值一样的颜色值;
如果Cu,Cd,Cl,Cr的颜色值有2个一样,另外2个不一样,则替换颜色值取2个值一样的颜色值;
如果Cu,Cd,Cl,Cr的颜色值有2个一样,另外2个也一样,则替换颜色值取2个颜色值的混合颜色值;
如果Cu,Cd,Cl,Cr的颜色值都不一样,则说明是孤立的笔迹点,可以当成噪点去掉,将其在笔迹点集合中移除,替换颜色值取四个值的混合颜色值。
请一并参阅图6,本发明实施例还提供了一种基于图片生成笔迹格式的文件的装置100,包括:
笔迹识别单元10,用于对待处理的图片进行识别,获得识别到的笔迹。
笔迹颜色值生成单元20,用于根据组成识别到的所述笔迹的像素点的颜色值获得至少一个笔迹颜色值。
笔迹点集合生成单元30,用于获取所述图片中与所述笔迹颜色值具有相同颜色值的像素点,生成笔迹点集合。
背景图生成单元40,用于基于所述笔迹点集合及所述图片生成背景图。
笔迹文件生成单元50,用于根据所述背景图及所述笔迹点集合生成笔迹格式的文件。
优选地,在一个实现方式中,所述笔迹颜色值生成单元20具体包括:
第一统计模块21,用于获取组成识别到的所述笔迹的所有像素点的颜色值,并统计与每个颜色值对应的像素点的个数;
第一标记模块22,用于将与个数最多的像素点对应的颜色值标记为笔迹颜色值。
优选地,在一个实现方式中,所述笔迹颜色值生成单元20具体包括:
第二统计模块23,用于获取组成识别到的所述笔迹的像素点的颜色值,并统计与每个颜色值对应的像素点的个数。
第二标记模块24,用于将对应的像素点的个数与识别到的像素点总数的比值超过预定阈值的颜色值标记为笔迹颜色值。
优选地,在一个实现方式中,所述背景图生成单元40具体包括:
背景颜色值获取模块41,用于获取所述图片的背景颜色值;
第一查询模块42,用于查询所述图片中位于所述笔迹点集合内的像素点;
第一颜色值替换模块43,用于将查询到的所述像素点的颜色值替换为所述背景颜色值,生成背景图。
优选地,在一个实现方式中,所述背景图生成单元40具体包括:
第二查询模块44,用于查询所述图片中位于所述笔迹点集合内的像素点。
替换颜色值生成模块45,用于根据与查询到的所述像素点相邻的至少4个相邻像素点的颜色值生成替换颜色值。
第二颜色值替换模块46,用于将查询到的所述像素点的颜色值替换为所述替换颜色值,生成背景图。
优选地,所述替换颜色值生成模块45具体包括:
第一相邻颜色值获取子模块451,用于获取与查询到的所述像素点相邻的至少4个相邻像素点的颜色值。
第一去除子模块452,用于去除位于所述笔迹点集合内的相邻像素点。
颜色值混合子模块453,用于对余下的相邻像素点的颜色值进行混合,生成替换颜色值。
优选地,所述替换颜色值生成模块45具体包括:
第二相邻颜色值获取子模块454,用于获取与查询到的所述像素点相邻的至少4个相邻像素点的颜色值。
第二去除子模块455,用于去除位于所述笔迹点集合内的相邻像素点;
颜色值设置子模块456,用于根据余下的相邻像素点的颜色值的出现次数确定替换颜色值。
本发明实施例提供的基于图片生成笔迹格式的文件的装置100,通过笔迹识别单元10识别所述图片中的笔迹,并由笔迹颜色值生成单元20根据识别到的笔迹获得笔迹颜色值,再由笔迹点集合生成单元30利用笔迹颜色值获得所述图片中的所有笔迹的像素点,生成笔迹点集合,接着由背景图生成单元40根据笔迹点集合恢复出背景图,最后笔迹文件生成单元50根据笔迹点集合和背景图生成笔迹格式的文件,从而将图片转换为笔迹格式的文件。一方面,对于没有安装有可以打开笔迹格式的文件的环境或终端,可以直接通过打开原有的图片进行内容的查看,而对于安装有可以打开笔迹格式的文件的环境或终端(如电子白板),可以用对图片转换成的笔迹格式的文件中的笔迹进行二次编辑,提高了文件的兼容性,方便了用户的操作和体验。
以上所揭露的仅为本发明一种较佳实施例而已,当然不能以此来限定本发明之权利范围,本领域普通技术人员可以理解实现上述实施例的全部或部分流程,并依本发明权利要求所作的等同变化,仍属于发明所涵盖的范围。
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程,是可以通过计算机程序来指令相关的硬件来完成,所述的程序可存储于一计算机可读取存储介质中,该程序在执行时,可包括如上述各方法的实施例的流程。其中,所述的存储介质可为磁碟、光盘、只读存储记忆体(Read-Only Memory,ROM)或随机存储记忆体(Random Access Memory,RAM)等。

Claims (10)

  1. 一种基于图片生成笔迹格式的文件的方法,其特征在于,包括:
    对待处理的图片进行识别,获得识别到的笔迹;
    根据组成识别到的所述笔迹的像素点的颜色值获得至少一个笔迹颜色值;
    获取所述图片中与所述笔迹颜色值具有相同颜色值的像素点,生成笔迹点集合;
    基于所述笔迹点集合及所述图片生成背景图;
    根据所述背景图及所述笔迹点集合生成笔迹格式的文件。
  2. 根据权利要求1所述的基于图片生成笔迹格式的文件的方法,其特征在于,所述根据组成识别到的所述笔迹的像素点的颜色值获得至少一个笔迹颜色值具体为:
    获取组成识别到的所述笔迹的所有像素点的颜色值,并统计与每个颜色值对应的像素点的个数;
    将与个数最多的像素点对应的颜色值设置为笔迹颜色值。
  3. 根据权利要求1所述的基于图片生成笔迹格式的文件的方法,其特征在于,所述根据组成识别到的所述笔迹的像素点的颜色值获得至少一个笔迹颜色值具体为:
    获取组成识别到的所述笔迹的所有像素点的颜色值,并统计与每个颜色值对应的像素点的个数;
    将对应的像素点的个数与识别到的像素点总数的比值超过预定阈值的颜色值标记为笔迹颜色值。
  4. 根据权利要求1所述的基于图片生成笔迹格式的文件的方法,其特征在于,所述根据所述笔迹点集合及所述图片生成背景图,具体包括:
    获取所述图片的背景颜色值;
    查询所述图片中位于所述笔迹点集合内的像素点;
    将查询到的所述像素点的颜色值替换为所述背景颜色值,生成背景图。
  5. 根据权利要求1所述的基于图片生成笔迹格式的文件的方法,其特征在于,所述根据所述笔迹点集合及所述图片生成背景图,具体包括:
    查询所述图片中位于所述笔迹点集合内的像素点;
    根据与查询到的所述像素点相邻的相邻像素点的颜色值生成替换颜色值;
    将查询到的所述像素点的颜色值替换为所述替换颜色值,生成背景图。
  6. 根据权利要求5所述的基于图片生成笔迹格式的文件的方法,其特征在于,所述根据与查询到的所述像素点相邻的相邻像素点的颜色值生成替换颜色值,具体为:
    获取与查询到的所述像素点相邻的相邻像素点的颜色值;
    去除位于所述笔迹点集合内的相邻像素点;
    对余下的相邻像素点的颜色值进行混合,生成替换颜色值。
  7. 根据权利要求5所述的基于图片生成笔迹格式的文件的方法,其特征在于,所述根据与查询到的所述像素点相邻的像素点的颜色值生成替换颜色值,具体为:
    获取与查询到的所述像素点相邻的相邻像素点的颜色值;
    去除位于所述笔迹点集合内的相邻像素点;
    根据余下的相邻像素点的颜色值的出现次数确定替换颜色值。
  8. 一种基于图片生成笔迹格式的文件的装置,其特征在于,包括:
    笔迹识别单元,用于对待处理的图片进行识别,获得识别到的笔迹;
    笔迹颜色值生成单元,用于根据组成识别到的所述笔迹的像素点的颜色值获得至少一个笔迹颜色值;
    笔迹点集合生成单元,用于获取所述图片中与所述笔迹颜色值具有相同颜色值的像素点,生成笔迹点集合;
    背景图生成单元,用于基于所述笔迹点集合及所述图片生成背景图;
    笔迹文件生成单元,用于根据所述背景图及所述笔迹点集合生成笔迹格式的文件。
  9. 根据权利要求8所述的基于图片生成笔迹格式的文件的装置,其特征在于,所述笔迹颜色值生成单元具体包括:
    第一统计模块,用于获取组成识别到的所述笔迹的所有像素点的颜色值,并统计与每个颜色值对应的像素点的个数;
    第一标记模块,用于将与个数最多的像素点对应的颜色值标记为笔迹颜色值。
  10. 根据权利要求8所述的基于图片生成笔迹格式的文件的装置,其特征在于,所述背景图生成单元具体包括:
    背景颜色值获取模块,用于获取所述图片的背景颜色值;
    第一查询模块,用于查询所述图片中位于所述笔迹点集合内的像素点;
    第一颜色值替换模块,用于将查询到的所述像素点的颜色值替换为所述背景颜色值,生成背景图。
PCT/CN2017/077154 2016-12-07 2017-03-17 一种基于图片生成笔迹格式的文件的方法及装置 Ceased WO2018103226A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201611114430.1A CN106599818B (zh) 2016-12-07 2016-12-07 一种基于图片生成笔迹格式的文件的方法及装置
CN201611114430.1 2016-12-07

Publications (1)

Publication Number Publication Date
WO2018103226A1 true WO2018103226A1 (zh) 2018-06-14

Family

ID=58596299

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/077154 Ceased WO2018103226A1 (zh) 2016-12-07 2017-03-17 一种基于图片生成笔迹格式的文件的方法及装置

Country Status (2)

Country Link
CN (1) CN106599818B (zh)
WO (1) WO2018103226A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111160265A (zh) * 2019-12-30 2020-05-15 Oppo(重庆)智能科技有限公司 文件转换方法、装置、存储介质及电子设备
CN112256146A (zh) * 2020-09-09 2021-01-22 广州视源电子科技股份有限公司 一种白板书写笔迹的测试方法、装置、设备和存储介质

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108874254B (zh) * 2018-06-12 2019-04-09 掌阅科技股份有限公司 电子书阅读器手写铅笔笔迹的绘制方法及电子书阅读器
CN109062491A (zh) * 2018-07-23 2018-12-21 广州视源电子科技股份有限公司 交互智能设备的笔迹处理方法和装置
CN109215098B (zh) * 2018-08-20 2023-04-11 广州视源电子科技股份有限公司 笔迹擦除方法和装置
CN110174978A (zh) * 2019-05-13 2019-08-27 广州视源电子科技股份有限公司 数据处理方法、装置、智能交互平板及存储介质
CN113688838B (zh) * 2021-10-25 2022-03-22 江西软云科技股份有限公司 红色笔迹提取方法、系统、可读存储介质及计算机设备

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2004038685A1 (en) * 2002-10-24 2004-05-06 Immediad Aps System for presenting and controlling information on a display device
CN101593324A (zh) * 2009-06-17 2009-12-02 浙江师范大学 基于可信计算应用技术的网络多级审批方法及系统
CN102289322A (zh) * 2011-08-25 2011-12-21 盛乐信息技术(上海)有限公司 手写笔迹处理方法及系统
CN102541347A (zh) * 2011-11-15 2012-07-04 上海大学 手写笔体自动识别系统和方法
CN102968789A (zh) * 2012-10-25 2013-03-13 清华大学 基于数字图像处理的出土简帛字形图像的提取方法
CN105205488A (zh) * 2015-07-13 2015-12-30 西北工业大学 基于Harris角点和笔画宽度的文字区域检测方法

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6909803B2 (en) * 2000-12-22 2005-06-21 Canon Kabushiki Kaisha Text color detection for copier image processing
JP4857173B2 (ja) * 2007-04-25 2012-01-18 日立オムロンターミナルソリューションズ株式会社 画像処理装置、画像処理方法及び画像処理プログラム
JP5089713B2 (ja) * 2010-01-18 2012-12-05 シャープ株式会社 画像圧縮装置、圧縮画像出力装置、画像圧縮方法、コンピュータプログラム及び記録媒体
CN102136064A (zh) * 2011-03-24 2011-07-27 成都四方信息技术有限公司 一种图像文字识别系统
CN103544722B (zh) * 2013-10-25 2016-08-17 深圳市掌网立体时代视讯技术有限公司 一种数字书画的笔迹修饰方法及装置

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2004038685A1 (en) * 2002-10-24 2004-05-06 Immediad Aps System for presenting and controlling information on a display device
CN101593324A (zh) * 2009-06-17 2009-12-02 浙江师范大学 基于可信计算应用技术的网络多级审批方法及系统
CN102289322A (zh) * 2011-08-25 2011-12-21 盛乐信息技术(上海)有限公司 手写笔迹处理方法及系统
CN102541347A (zh) * 2011-11-15 2012-07-04 上海大学 手写笔体自动识别系统和方法
CN102968789A (zh) * 2012-10-25 2013-03-13 清华大学 基于数字图像处理的出土简帛字形图像的提取方法
CN105205488A (zh) * 2015-07-13 2015-12-30 西北工业大学 基于Harris角点和笔画宽度的文字区域检测方法

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111160265A (zh) * 2019-12-30 2020-05-15 Oppo(重庆)智能科技有限公司 文件转换方法、装置、存储介质及电子设备
CN112256146A (zh) * 2020-09-09 2021-01-22 广州视源电子科技股份有限公司 一种白板书写笔迹的测试方法、装置、设备和存储介质
CN112256146B (zh) * 2020-09-09 2024-05-03 广州视源电子科技股份有限公司 一种白板书写笔迹的测试方法、装置、设备和存储介质

Also Published As

Publication number Publication date
CN106599818B (zh) 2020-10-27
CN106599818A (zh) 2017-04-26

Similar Documents

Publication Publication Date Title
CN106599818B (zh) 一种基于图片生成笔迹格式的文件的方法及装置
CN109670500B (zh) 一种文字区域获取方法、装置、存储介质及终端设备
TWI536277B (zh) Form identification method and device
US8693790B2 (en) Form template definition method and form template definition apparatus
US11625871B2 (en) System and method for capturing and interpreting images into triple diagrams
CN103336961B (zh) 一种交互式的自然场景文本检测方法
CN110569774B (zh) 基于图像处理与模式识别的折线图图像自动数字化方法
CN107659799B (zh) 摄像装置、图像处理方法和存储介质
CN108965646B (zh) 图像处理装置、图像处理方法
CN115273115B (zh) 一种文档元素标注方法、装置、电子设备和存储介质
CN115761270A (zh) 色卡的检测方法、装置、电子设备及存储介质
CN113033562A (zh) 一种图像处理方法、装置、设备及存储介质
CN104156703B (zh) 基于色彩纹理的车牌定位方法及系统
CN106682670A (zh) 一种台标识别方法及系统
JP7111599B2 (ja) 図面認識装置、図面認識方法及びプログラム
CN105654087A (zh) 一种基于彩色模板的脱机手写字符提取方法
CN111783807B (zh) 图片提取方法、装置和计算机可读存储介质
CN112132141B (zh) 字符合并方法、装置、电子设备及存储介质
CN110598575A (zh) 表格版面分析与提取方法及相关装置
US20100238470A1 (en) Document image processing system and document image processing method
US20220301237A1 (en) Information processing apparatus and non-transitory computer readable medium
CN112686253A (zh) 一种用于电子白板的屏幕文字提取系统及方法
CN114677697B (zh) 一种空间邻域辅助的降水曲线提取方法及装置
CN114972784B (zh) 一种双色导线颜色提取与匹配方法
CN121617103B (zh) 一种小学生手写作文多模态ocr识别方法和系统

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17878053

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 171019)

122 Ep: pct application non-entry in european phase

Ref document number: 17878053

Country of ref document: EP

Kind code of ref document: A1