CN111213157A - Express information input method and system based on intelligent terminal - Google Patents
Express information input method and system based on intelligent terminal Download PDFInfo
- Publication number
- CN111213157A CN111213157A CN201780095880.7A CN201780095880A CN111213157A CN 111213157 A CN111213157 A CN 111213157A CN 201780095880 A CN201780095880 A CN 201780095880A CN 111213157 A CN111213157 A CN 111213157A
- Authority
- CN
- China
- Prior art keywords
- information
- image
- module
- character
- express
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 78
- 238000007781 pre-processing Methods 0.000 claims abstract description 36
- 230000011218 segmentation Effects 0.000 claims abstract description 24
- 238000012545 processing Methods 0.000 claims abstract description 21
- 238000004458 analytical method Methods 0.000 claims abstract description 15
- 238000004891 communication Methods 0.000 claims description 26
- 238000000605 extraction Methods 0.000 claims description 23
- 238000010606 normalization Methods 0.000 claims description 21
- 238000012937 correction Methods 0.000 claims description 15
- 238000003709 image segmentation Methods 0.000 claims description 10
- 238000000926 separation method Methods 0.000 claims description 6
- 239000000284 extract Substances 0.000 claims description 2
- 230000003287 optical effect Effects 0.000 description 10
- 238000005286 illumination Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 238000004364 calculation method Methods 0.000 description 5
- 239000013598 vector Substances 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 4
- 230000008859 change Effects 0.000 description 4
- 230000005484 gravity Effects 0.000 description 4
- 238000011161 development Methods 0.000 description 3
- 230000018109 developmental process Effects 0.000 description 3
- 230000004069 differentiation Effects 0.000 description 3
- 238000012706 support-vector machine Methods 0.000 description 3
- 238000000354 decomposition reaction Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000011430 maximum method Methods 0.000 description 2
- 238000007639 printing Methods 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 description 1
- 240000004282 Grewia occidentalis Species 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000032823 cell division Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000012916 structural analysis Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 230000001131 transforming effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/10—Image acquisition
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Character Input (AREA)
- Character Discrimination (AREA)
Abstract
The invention provides an express information input method and an express information input system based on an intelligent terminal, wherein the input method comprises the following steps: entering a mail page of the intelligent terminal; acquiring an information image containing express delivery information; preprocessing the information image; performing layout analysis and region segmentation on the preprocessed information image; performing character recognition on each region of the information image after region segmentation, and extracting the express information; and recording the express information into the corresponding position of the mail page. According to the invention, the express information is automatically input without handwriting or manual input, the working efficiency is improved, and the processing time of a single express is reduced; meanwhile, the automatic information input reduces the error incidence rate caused by manual input.
Description
The invention relates to the field of intelligent terminals, in particular to an express information input method and an express information input system based on an intelligent terminal.
Since the first smart phone comes to the market, smart phone manufacturers are continuously improving product designs, more and more functions which cannot be imagined come to our sides, and with the richness of functions of smart phones and the rapid development of mobile internet, smart phones replace many common electronic devices around our sides, and change our life style and peripheral industries.
With the continuous perfection of electronic commerce transaction platforms and the rapid development of traditional communication, mobile communication and other technologies, more and more people obtain the required commodities in an online shopping mode, and the types of the commodities can relate to the aspects of daily life of people. The buyer only needs to place an order through a client product (a website or an application (App) in the mobile terminal) of the e-commerce platform, and then can enter a series of processes such as seller delivery, logistics company delivery and the like, and finally the buyer can receive the ordered goods without going out of the house. Among them, the filling of the address of the recipient/sender and the distribution of the goods are very critical, and once the delivery error occurs, the loss may be caused to the buyer or seller. At present, comparatively traditional express delivery manifest of posting the mode for adopting paper needs the manual personal information such as addressee, sender address of filling in, on the one hand, writes by hand inefficiency, seriously reduces work efficiency, and often need repeated work meaningless, and on the other hand, express delivery information's feedback needs to take notes the express delivery note number. Electronic mails which are created later are faster than original mails by means of scanning two-dimensional codes and the like, but personal information such as addressees and addresses of senders still needs to be manually input, and the information is still quite complicated to input.
Currently, information entry methods based on scanned images have been widely used. In the book and newspaper industry and other industries which have great dependence on paper reading materials, a scanner is adopted to scan a book, scanned picture contents are combined into a complete book to be stored, and then the book file is converted into a PDF file. The information input based on the scanning image has the characteristics of short input time, large input information amount, small input information loss and the like.
In view of the rapid development of current e-commerce and the continuous increase of business volume of express delivery industry, the mode of manually inputting express delivery information by express delivery practitioners is more and more difficult to adapt to the requirement of the industry for large quantity and short time. Therefore, an express information input method and an express information input system which are convenient and rapid to use, high in information acquisition and input accuracy, stable in performance and high in applicability are urgently needed, the defect that express information needs to be manually input in the existing express industry is overcome, and working efficiency is improved.
Therefore, the express information input method and the express information input system based on the intelligent terminal are provided, the express information is automatically input without handwriting or manual input, the working efficiency is improved, and the processing time of single express is reduced; meanwhile, the automatic information input reduces the error incidence rate caused by manual input.
Disclosure of Invention
In order to overcome the technical defects, the invention aims to provide an express delivery information input method and an express delivery information input system based on an intelligent terminal.
In one aspect of the invention, an express information input method based on an intelligent terminal is disclosed, which comprises the following steps:
entering a mail page of the intelligent terminal;
acquiring an information image containing express delivery information;
preprocessing the information image;
performing layout analysis and region segmentation on the preprocessed information image;
performing character recognition on each region of the information image after region segmentation, and extracting the express information;
and recording the express information into the corresponding position of the mail page.
Preferably, the step of acquiring an information image containing express delivery information includes:
starting a camera of the intelligent terminal, and shooting and acquiring an information image containing express delivery information;
and/or
And calling the photo album application of the intelligent terminal to acquire an information image containing express delivery information.
Preferably, the step of preprocessing the information image comprises:
carrying out binarization processing or gray level processing on the information image;
and performing inclination correction by taking the edge of the information image or the line direction of the characters as a reference.
Preferably, the step of performing character recognition on each region of the information image after region segmentation and extracting the express delivery information includes:
cutting the information image into image lines and separating single characters from the image lines;
extracting statistical or structural features from the single character, including refinement and normalization;
and finding the character class with the highest similarity to the single character from the learned feature library.
Preferably, the express delivery information entry method further includes:
and storing the express delivery information.
On the other hand, the invention discloses an express information input system based on an intelligent terminal, which comprises the following components: the system comprises a page entry module, an image acquisition module, an image preprocessing module, an image segmentation module, an information identification module and an information input module;
the page entering module enters a mail page of the intelligent terminal;
the image acquisition module is in communication connection with the page entry module and acquires an information image containing express delivery information;
the image preprocessing module is in communication connection with the image acquisition module and is used for preprocessing the information image;
the image segmentation module is in communication connection with the image preprocessing module and is used for performing layout analysis and region segmentation on the preprocessed information image;
the information identification module is in communication connection with the image segmentation module, performs character identification on each region of the information image after region segmentation, and extracts the express information;
the information input module is in communication connection with the page entering module and the information identification module and inputs the express information into the corresponding position of the mail page.
Preferably, the image acquisition module starts a camera of the intelligent terminal, and shoots and acquires an information image containing express delivery information;
and/or
The image acquisition module calls the photo album application of the intelligent terminal to acquire an information image containing express delivery information.
Preferably, the image preprocessing module comprises: an image preprocessing unit and an inclination correcting unit;
the image preprocessing unit is used for carrying out binarization processing or gray level processing on the information image;
the inclination correction unit performs inclination correction with respect to the edge of the information image or the line direction of the text.
Preferably, the information identifying module includes: the device comprises a character separation unit, a feature extraction unit and a character matching unit;
the character separating unit cuts the information image into image lines and separates single characters from the image lines;
the feature extraction unit is in communication connection with the character separation unit and is used for extracting statistical features or structural features from the single character, wherein the statistical features or the structural features comprise thinning and normalization;
and the character matching unit is in communication connection with the feature extraction unit and finds the character class with the highest similarity with the single character from the learned feature library.
Preferably, the express delivery information entry system further comprises an information storage module, which is in communication connection with the information identification module and stores the express delivery information.
After the technical scheme is adopted, compared with the prior art, the method has the following beneficial effects:
1. according to the express information input method and the input system, the express information is automatically input without handwriting or manual input, the working efficiency is improved, and the processing time of a single express is reduced; meanwhile, the automatic information input reduces the error incidence rate caused by manual input.
Fig. 1 is a schematic flow chart of an express delivery information entry method according to a preferred embodiment of the present invention;
fig. 2 is a schematic flow chart of an image preprocessing step of the express delivery information entry method of fig. 1;
fig. 3 is a schematic flow chart of a express delivery information identification step of the express delivery information entry method of fig. 1;
fig. 4 is a schematic structural diagram of an express delivery information entry system according to a preferred embodiment of the present invention.
Reference numerals:
100-express delivery information entry system; 11-page entry module; 12-an image acquisition module; 13-an image pre-processing module; 14-an image segmentation module; 15-an information identification module; 16-an information entry module; 17-information storage module.
The advantages of the invention are further illustrated in the following description of specific embodiments in conjunction with the accompanying drawings.
Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the invention, as detailed in the appended claims.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. As used in this specification and the appended claims, the singular forms "a", "an", "the", and the like are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.
In the description of the present invention, unless otherwise specified and limited, it is to be noted that the term "connected" is to be interpreted broadly, and may be, for example, a mechanical connection or an electrical connection, or a communication between two elements, or may be a direct connection or an indirect connection through an intermediate medium, and a specific meaning of the term may be understood by those skilled in the art according to specific situations.
In the following description, suffixes such as "module" or "unit" used to represent elements are used only for facilitating the explanation of the present invention, and have no specific meaning in themselves. Thus, "module" and "unit" may be used in a mixture.
The information entry method and the information entry system of the present invention may be applied to an intelligent terminal, which may be in various forms, for example, the intelligent terminal described in the present invention may include a mobile terminal such as a mobile phone, a smart phone, a notebook computer, a PDA (personal digital assistant), a PAD (tablet computer), a PMP (portable multimedia player), a navigation device, a smart watch, and the like, and a fixed terminal such as a digital TV, a desktop computer, and the like. In the following, the present invention will be described assuming that the terminal is a mobile terminal and that the mobile terminal is a smartphone. However, it will be understood by those skilled in the art that the configuration according to the embodiment of the present invention can be applied to a fixed type terminal in addition to elements particularly used for moving purposes. For convenience of description, the embodiments of the present invention are described with reference to a smart phone as an example, and other application scenarios may be referred to each other.
Referring to fig. 1, a method for entering express delivery information based on an intelligent terminal according to a preferred embodiment of the present invention includes the following steps:
s100: entering a mail page of the intelligent terminal;
s200: acquiring an information image containing express delivery information;
s300: preprocessing the information image;
s400: performing layout analysis and region segmentation on the preprocessed information image;
s500: performing character recognition on each region of the information image after region segmentation, and extracting the express information;
s600: and recording the express information into the corresponding position of the mail page.
Specifically, the method comprises the following steps:
-S100: entering a mail page of the intelligent terminal;
when the intelligent terminal is used for electronic mail sending, an express mail sending page is entered, a newly added address is added, and when the information of a receiver/sender needs to be input, characters or other prompts of shooting or uploading images and automatic identification are displayed to prompt a user to shoot information images containing express information.
-S200: acquiring an information image containing express delivery information; in a preferred embodiment, S200: the step of obtaining an information image containing express delivery information comprises the following steps:
starting a camera of the intelligent terminal, and shooting and acquiring an information image containing express delivery information;
and/or
And calling the photo album application of the intelligent terminal to acquire an information image containing express delivery information.
The method for acquiring the information image can be various, and the image can be shot on the spot by calling a camera of the intelligent terminal or randomly selected from a local or cloud photo album of the intelligent terminal in a screenshot mode.
Here, the express delivery information includes, but is not limited to: name of sender/receiver, telephone, location, detailed address, etc.
-S300: preprocessing the information image;
referring to fig. 2, in a preferred embodiment, S300: the step of preprocessing the information image comprises:
s310: carrying out binarization processing or gray level processing on the information image;
s320: and performing inclination correction by taking the edge of the information image or the line direction of the characters as a reference.
-S400: and performing layout analysis and region segmentation on the preprocessed information image.
Preprocessing, layout analysis and region segmentation of the image:
-non-uniform illumination correction
The problem of uneven illumination is an important link for identifying and extracting express information in an information image. High light and shadow are two major factors affecting image uniformity. The highlight region of the information image can be automatically corrected by the color shift ratio of each component of the image in the R, G, B space. The Retinex method can be used for correcting the shadow of an optical image and enhancing the image.
In reality, image acquisition is affected by uneven illumination, and partial high-brightness areas may exist in the image. Highlight region correction can detect normal illumination regions using pixel and spatial distribution characteristics of an information image, and then automatically correct non-uniform illumination in the image using color shift ratios of the respective components.
Graying
There are three main methods for image graying:
(1) maximum method: namely, the maximum one of the gray values of the three channels of R, G and B is selected as the gray value of the current pixel point. The mathematical expression is as follows:
G＝max(R，G，B)
(2) average value method: directly adding the gray values of the R, G and B channels to obtain an average value as the gray value of the current pixel point. The mathematical expression is as follows:
G＝(R+G+B)/3
(3) weighted average method: on the basis of an average value method, the coefficient of each original channel is converted into different values according to the light sensing principle of human eyes, so that the result which is closer to the observation result of the human eyes is obtained. The mathematical expression is as follows:
G＝k0R+k1G+k2B
wherein k is0、k1、k2The following relationship is satisfied:
k0+k1+k2＝1
the maximum method usually results in a high-brightness gray image, the average method usually blurs the edge details of the image to obtain an image with soft edges, and the weighted average method adjusts k0、k1、k2The value of (c) can result in an image that retains more high frequency information (i.e., edge information) of the image. Therefore, the invention adopts a weighted average method to graye the information image after the non-uniformity correction.
-binarization
The information image subjected to non-uniform correction and image graying reduces the influence of non-uniform illumination and image color on character recognition. The express information in the information image is mainly characters, and the main characteristics of the characters are structural characteristics. The structural feature is a feature quantity independent of the image gradation, so the information image also needs to be subjected to binarization processing, i.e., reduction of the gradation image to a (0, 1) image. Binarization is generally represented by the following equation:
binarization of grayscale images requires efficient differentiation of character pixels from the image. The binarization requirement of the traditional printing font image requires that the structural characteristics of the original characters are kept as much as possible on one hand, and the binaryzation of the characters cannot be blank on the other hand. The background of the traditional printed characters is relatively single and smooth, and the background and the characters can be distinguished by a simple global threshold value method.
-layout analysis
The layout analysis of the image includes the tilt correction of the image and the segmentation of the lines and rows of the image. The image tilt correction method mainly includes a projective drawing method, a straight line fitting method, a Hough change method, and the like. The invention can be adjusted by paying attention to the placing angle and the like when the image is collected, and the condition of image inclination can not be considered for simplifying the processing. Therefore, the main content of the layout analysis of the information image is the row-column division of the information image. Layout analysis is one of the key technologies for optical image-based recognition systems. Through analyzing the layout of the information image, the position of express information such as the name, the sex, the address, the telephone number and the like of a receiver/sender in the information image is determined, and the method has important significance for accurately analyzing the express information. Although the layout of the information image of the general express delivery information is relatively fixed, the layout is relatively complex. Firstly, the typesetting of the information image must have some differences, and a method with better fitness must be selected. Secondly, the font size of the same line in the information image may be different, for example, in the first line of name information, the name font is smaller than that of the "name", which brings great trouble to line extraction in layout analysis. The phenomenon that numbers and Chinese character fonts coexist exists in the secondary image, and great difficulty is brought to the division of character blocks.
The layout analysis method of the image includes a top-up method and a bottom-up method. The top-down approach starts with the macroscopic direction of the image and gradually distinguishes the image into different modules by analyzing the global features of the image. And finally, dividing the image into the structural elements by iteration division again and again. In the invention, in the dividing process of the character area, firstly, the starting position and the ending position of the character line are determined, and secondly, the position of a single character is determined in each line. The top-down method is effective for character recognition with a relatively fixed layout, but for the situation of a relatively complex layout, it is difficult to accurately segment structural elements such as characters, tables, images and the like in an image due to neglect of a large amount of image details.
The bottom-up method is that starting from basic structural elements of the image, the basic structural elements are gradually combined into a character, an image or a table through structural analysis of local elements; and then, analyzing the position relation among characters, tables or images to obtain the row and column information in the image layout. And analyzing the structure in the image row by row and column by column so as to finish the extraction work of the whole layout information. A large amount of iterative operations are designed by a bottom-down method, the calculation process is complex, the calculation speed is low, and the application is less in practice. At present, a large number of character recognition methods based on optical characters mainly combine a top-down method and a bottom-up method, so that a balance is obtained between recognition speed and performance.
-S500: performing character recognition on each region of the information image after region segmentation, and extracting the express letter:
referring to fig. 3, in a preferred embodiment, S500: performing character recognition on each region of the information image after region segmentation, wherein the step of extracting the express information comprises the following steps:
s510: cutting the information image into image lines and separating single characters from the image lines;
s520: extracting statistical or structural features from the single character, including refinement and normalization;
s530: and finding the character class with the highest similarity to the single character from the learned feature library.
Extracting character features in the information image:
to automatically recognize express delivery information from an optical image, the express delivery information in the optical image needs to be classified efficiently and accurately for each character. However, it is not practical to directly use the image obtained after image preprocessing for matching. Firstly, the image to be recognized itself contains a lot of information, taking an image of 64 × 64 size as an example, if the image to be recognized is directly matched, 4096-dimensional feature vectors are formed, and the search space of the recognizer is greatly increased. Secondly, the storage space of the character template required for matching directly with the character image can be very large. At present, Chinese characters recorded are about 9 ten thousand characters, although the commonly used Chinese characters are only about 3500 characters, along with the improvement of the cultural level of people, the probability of rarely used characters appearing in express information, particularly in a name of a receiver/sender, is higher and higher, so that if an original image of a character is taken for storage, the storage amount of the character with a single font is very large, and the font of different provinces is slightly different. Finally, the directly used image is affected by sampling noise of an image sensor, an image acquisition angle and the like, and a large matching error is brought. Therefore, the character features in the optical image need to be extracted after the image preprocessing to reduce the search space of the character recognizer.
Character feature extraction is mainly to extract features with different essentials among characters from original optical image data. The extraction of character features needs to follow some rules as follows:
(1) distinctiveness: i.e. there is a large distinction between different characters. For example, there is a need for greater differentiation between kanji characters and numbers, as well as for different differentiation between kanji characters. Different chinese characters should have a larger distance in the feature space. Such features enable differential processing of different characters in the presence of sampling noise or other disturbances in the image.
(2) Reliability: i.e. the same character has the same or very similar feature vectors even in case of noise or rotation, scaling etc.
(3) Independence: i.e. requires no correlation between different features of the same character. The independence between the feature components ensures that the judgment of other feature quantities is not influenced by the unchanged single feature quantity.
In the process of extracting the character features, the feature requirements are met, and meanwhile, the number of the features is required to be as small as possible. The characteristic components as few as possible can ensure the effective information input space of the recognizer on one hand, reduce the storage space required by the matching template on the other hand, and simultaneously reduce the search space of the recognizer and accelerate the recognition process.
-character normalization processing
The sizes of the character block images obtained by segmentation after the image overall preprocessing have large differences. The size of the image block does not affect the structural characteristics of the character, however, the stroke length of the Chinese character is also a very important characteristic of the Chinese character. The length features of the strokes of the Chinese characters are in a direct proportion relation with the size of the image, so the character image obtained after character segmentation needs to be normalized before feature extraction. The character normalization process includes position normalization and size normalization.
The character position normalization has two methods of gravity center normalization and frame normalization. The center-of-gravity normalization firstly calculates the position of the center of gravity of the character image of the printing form, and after the position of the center of gravity of the image is obtained through calculation, the center of gravity of the character is moved to the position of the center of the image, so that the position normalization of the character image after segmentation is completed. Frame normalization firstly calculates the upper, lower, left and right frames of a character image, and after the central position of the image is obtained through calculation, the central position of the character is moved to the center of the image, so that the character image normalization is completed.
There are two ways for normalizing the image size, one is normalization according to the size of the character image frame: that is, the image is enlarged or reduced to a predetermined specific size according to the size of the outer frame of the image. The method for carrying out normalization operation according to the size of the image frame is simple in operation and small in operation amount. Another method of image size normalization needs to take into account the distribution characteristics of the image. One of the simpler methods in the distribution characteristics of an image is the distribution variance of the image.
Structural features of the character
The structural features of a kanji character include stroke features and part features.
The stroke characteristics of a Chinese character include the type of stroke and the length of the stroke. The types of strokes mainly include: horizontal, vertical, left-falling, right-falling, turning and hooking. The statistical data shows the proportion of 6 strokes in Chinese character, the four basic strokes of horizontal stroke, left falling stroke, vertical stroke and right falling stroke in Chinese character are the most, and the folding stroke and the hooking stroke can be considered to be formed by the four basic strokes approximately. Therefore, the strokes of Chinese characters can be used as an important characteristic for recognition. The structural characteristics of the character can also be statistically analyzed in the horizontal, vertical, left-falling and right-falling directions.
The points on the strokes can be used as the characteristic points of the Chinese characters, and meanwhile, in order to further distinguish the structure, the important points on the background of the characters can also be used as the characteristic points of the characters. The points on the stroke and the points in the character background can together form an important feature vector for Chinese character recognition. The feature points in the stroke include end points, break points, branch points, etc.
The stroke relation structure of Chinese characters is simple, and the relation between components is very complex. Different combinations of the number and location of the components may also occur. The number of the components is seen as single-body character, double-body character, three-body character, four-body character and the like. The independent font, the left and right font, the upper and lower font and the surrounding font are arranged according to the spatial position relation of the components.
The four-direction decomposition of Chinese characters means that according to the structural characteristics of Chinese characters, the strokes of Chinese characters are decomposed from four directions of horizontal, vertical, left-falling and right-falling. The simplest four-direction decomposition determines the direction of a Chinese character stroke by determining points in eight fields of pixel points. The eight fields of Chinese character stroke direction distinguishing method includes 'and' method and 'or' method.
The above method assumes that the strokes of the chinese character are single pixel points. Therefore, the input Chinese character image is required to be an image subjected to stroke thinning or stroke skeleton extraction. However, in practical use, the thinning of the strokes of the Chinese characters or the skeleton extraction easily causes the condition that the strokes are lost for the character image with thinner strokes, thereby reducing the recognition probability of the characters.
-statistical characteristics of characters
The statistical features of the optical image-based typographic characters include both global features and local features. Unlike the structural features of the characters, the statistical features of the characters are obtained from the binary image, and some of the statistical features of the characters are obtained even directly from the gray image through corresponding transformation.
The global feature of the character image is essentially only to treat the character image as a normal image, and the character is only an object having a certain feature therein. Therefore, the global feature extraction method of the character image is similar to the feature extraction method of the general image. The method for extracting the global characteristics of the character image mainly comprises the following steps:
(1) change domain feature component: and transforming the binarized character image to other feature spaces, and taking coefficients of corresponding vectors in the feature spaces as features. Common transforms include a two-dimensional fourier transform (2-D fourier transform), a hadamard transform (hadamard transform), and a Hough transform (Hough transform). The two-dimensional fourier transform is to transform image information of a spatial domain into a two-dimensional frequency domain, and to transform a spatial position change having a large correlation into a frequency domain composed of frequency components having normalized orthogonality, thereby obtaining characteristics different between target (i.e., characters). The Hadamard transform is a commonly used feature transform in remote sensing images, and realizes the transform from image space relation to multispectral domain by using symmetric and orthogonal Hadamard matrix, thus achieving the purpose of extracting target features and carrying out classification and identification on the remote sensing images. The Hough transform is one of basic methods for recognizing a target with a specific shape from an optical image, and the Hough transform used for extracting target features after expansion mainly transforms the image from a spatial domain into a feature space consisting of different basic shapes, and the features of the target are formed by using coefficients of different geometric shapes.
(2) Moment invariant Feature (Moment Feature): invariant moment is an important method for detecting and identifying targets in optical image processing. The central moment and the origin moment of the image can distinguish the geometric shape information of the projection of the target on the imaging plane, but the geometric shape of the projection plane has no scale, rotation or affine invariance.
(3) Global projection feature: the images are projected to several reference directions, respectively, and only strokes perpendicular to the reference directions are taken to be projected to the reference directions. Compared with the stroke extraction method based on the structural features, the method is simple in operation and can realize rapid extraction. The global projection characteristics can reflect the complexity of the whole Chinese character, the main direction of the strokes, the possible connection relation among the strokes and the like to a certain degree. To simplify the calculation, projections in four directions are usually taken, i.e., the lateral direction, the vertical direction, the positive 45 degree direction, and the negative 45 degree direction.
(4) Background characteristics: the background portion and strokes of the chinese characters may also be used as a global feature of the image of the chinese characters. The blank dots (non-stroke dots) located on the two diagonal lines of the image are usually selected to count the stroke density of the dots in each direction of the character as the global background feature of the image.
The local features of the Chinese characters need to divide the image into different local regions first, and the features of the image are counted in different region ranges. The features of the image may be one or a combination of the global features previously described. The method for dividing the key points of the local feature extraction into local regions. The local region division method mainly includes a mesh method (including a fixed mesh method and an elastic mesh method), a cell division method, a direction line element division method, a four-corner feature division method, a Gabor division method, and the like.
-character classifier
The extraction work of express information can be completed only by sending character feature vectors of single character images obtained through image preprocessing after normalization processing and character feature extraction into a Chinese character classifier. The classifier divides the input unknown samples into different types and completes the task of one-to-one correspondence between the samples to be identified and the types to which the samples belong. The image classifier mainly comprises an Euclidean distance classifier, a neural network classifier, a support vector machine classifier and a genetic algorithm classifier.
The Euclidean distance classifier is the simplest and intuitive classification method, and takes the distance of a point in a high-dimensional space as a main basis of sample similarity measurement. The smaller the distance value, the higher the degree of similarity between the samples to be measured.
The support vector machine classifier and the neural network have good curve fitting capability and target classification capability, and have a large number of applications in target identification and detection. The disadvantages of neural networks are also apparent. The current structure of the neural network has no reliable rule, so the convergence speed of the algorithm is very low, the initial value selection of the network has great influence on the performance of the algorithm, and the algorithm is easy to converge to a minimum value. Support vector machines use a special type of operation plane. Such a hyperplane is referred to as an optimal classification hyperplane.
-S600: recording the express information into a corresponding position of the mail page;
-S700: and storing the express delivery information.
Referring to fig. 4, the present invention also discloses an intelligent terminal based express delivery information entry system 100, where the express delivery information entry system 100 includes: the system comprises a page entering module 11, an image obtaining module 12, an image preprocessing module 13, an image segmentation module 14, an information identification module 15 and an information input module 16;
the page entering module 11 enters a mail page of the intelligent terminal;
the image acquisition module 12 is in communication connection with the page entry module 11, and acquires an information image containing express delivery information;
the image preprocessing module 13 is in communication connection with the image acquisition module 12 and is used for preprocessing the information image;
the image segmentation module 14 is in communication connection with the image preprocessing module 13, and performs layout analysis and region segmentation on the preprocessed information image;
the information identification module 15 is in communication connection with the image segmentation module 14, and performs character identification on each region of the information image after region segmentation to extract the express delivery information;
the information recording module 16 is in communication connection with the page entering module 11 and the information identification module 15, and records the express delivery information into the corresponding position of the mail page.
In a preferred embodiment, the image obtaining module 12 starts a camera of the intelligent terminal, and shoots and obtains an information image containing express delivery information;
and/or
The image obtaining module 12 calls an album application of the intelligent terminal to obtain an information image containing express delivery information.
In a preferred embodiment, the image preprocessing module 13 includes: an image preprocessing unit and an inclination correcting unit;
the image preprocessing unit is used for carrying out binarization processing or gray level processing on the information image;
the inclination correction unit performs inclination correction with respect to the edge of the information image or the line direction of the text.
In a preferred embodiment, the information identification module 15 includes: the device comprises a character separation unit, a feature extraction unit and a character matching unit;
the character separating unit cuts the information image into image lines and separates single characters from the image lines;
the feature extraction unit is in communication connection with the character separation unit and is used for extracting statistical features or structural features from the single character, wherein the statistical features or the structural features comprise thinning and normalization;
and the character matching unit is in communication connection with the feature extraction unit and finds the character class with the highest similarity with the single character from the learned feature library.
In a preferred embodiment, the express delivery information entry system 100 further includes an information storage module 17, which is connected to the information identification module 16 in a communication manner and stores the express delivery information.
It should be noted that the embodiments of the present invention have been described in terms of preferred embodiments, and not by way of limitation, and that those skilled in the art can make modifications and variations of the embodiments described above without departing from the spirit of the invention.
Claims (10)
- An express information input method based on an intelligent terminal is characterized by comprising the following steps:entering a mail page of the intelligent terminal;acquiring an information image containing express delivery information;preprocessing the information image;performing layout analysis and region segmentation on the preprocessed information image;performing character recognition on each region of the information image after region segmentation, and extracting the express information;and recording the express information into the corresponding position of the mail page.
- The courier information entry method of claim 1,the step of obtaining an information image containing express delivery information comprises the following steps:starting a camera of the intelligent terminal, and shooting and acquiring an information image containing express delivery information;and/orAnd calling the photo album application of the intelligent terminal to acquire an information image containing express delivery information.
- The courier information entry method of claim 1,the step of preprocessing the information image comprises:carrying out binarization processing or gray level processing on the information image;and performing inclination correction by taking the edge of the information image or the line direction of the characters as a reference.
- The courier information entry method of claim 1,performing character recognition on each region of the information image after region segmentation, wherein the step of extracting the express information comprises the following steps:cutting the information image into image lines and separating single characters from the image lines;extracting statistical or structural features from the single character, including refinement and normalization;and finding the character class with the highest similarity to the single character from the learned feature library.
- The courier information entry method of claim 1,the express delivery information entry method further comprises the following steps:and storing the express delivery information.
- An express delivery information input system based on an intelligent terminal is characterized in that,the express delivery information entry system comprises: the system comprises a page entry module, an image acquisition module, an image preprocessing module, an image segmentation module, an information identification module and an information input module;the page entering module enters a mail page of the intelligent terminal;the image acquisition module is in communication connection with the page entry module and acquires an information image containing express delivery information;the image preprocessing module is in communication connection with the image acquisition module and is used for preprocessing the information image;the image segmentation module is in communication connection with the image preprocessing module and is used for performing layout analysis and region segmentation on the preprocessed information image;the information identification module is in communication connection with the image segmentation module, performs character identification on each region of the information image after region segmentation, and extracts the express information;the information input module is in communication connection with the page entering module and the information identification module and inputs the express information into the corresponding position of the mail page.
- The courier information entry system of claim 6,the image acquisition module starts a camera of the intelligent terminal, shoots and acquires an information image containing express information;and/orThe image acquisition module calls the photo album application of the intelligent terminal to acquire an information image containing express delivery information.
- The courier information entry system of claim 6,the image preprocessing module comprises: an image preprocessing unit and an inclination correcting unit;the image preprocessing unit is used for carrying out binarization processing or gray level processing on the information image;the inclination correction unit performs inclination correction with respect to the edge of the information image or the line direction of the text.
- The courier information entry system of claim 6,the information identification module includes: the device comprises a character separation unit, a feature extraction unit and a character matching unit;the character separating unit cuts the information image into image lines and separates single characters from the image lines;the feature extraction unit is in communication connection with the character separation unit and is used for extracting statistical features or structural features from the single character, wherein the statistical features or the structural features comprise thinning and normalization;and the character matching unit is in communication connection with the feature extraction unit and finds the character class with the highest similarity with the single character from the learned feature library.
- The courier information entry system of claim 6,the express information entry system further comprises an information storage module which is in communication connection with the information identification module and used for storing the express information.
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/CN2017/105743 WO2019071476A1 (en) | 2017-10-11 | 2017-10-11 | Express information input method and system based on intelligent terminal |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111213157A true CN111213157A (en) | 2020-05-29 |
Family
ID=66100310
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201780095880.7A Pending CN111213157A (en) | 2017-10-11 | 2017-10-11 | Express information input method and system based on intelligent terminal |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN111213157A (en) |
| WO (1) | WO2019071476A1 (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110443538A (en) * | 2019-06-20 | 2019-11-12 | 苏州视印智能***有限公司 | A kind of addressee automatic information identifying system and method |
| CN112668574A (en) * | 2020-12-28 | 2021-04-16 | 深圳前海百递网络有限公司 | Parcel image processing method and device, computer equipment and storage medium |
| CN114780685A (en) * | 2022-04-28 | 2022-07-22 | 贵州电网有限责任公司 | Method for automatically identifying defect information input condition and supplementing defect information through unmanned aerial vehicle |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20060007204A (en) * | 2004-07-19 | 2006-01-24 | 한국과학기술원 | Document image processing and verification system for digitalizing a large volume of data and method thereof |
| CN1877598A (en) * | 2005-06-06 | 2006-12-13 | 英华达(上海)电子有限公司 | Method for gathering and recording business card information in mobile phone by using image recognition |
| CN104881770A (en) * | 2015-06-03 | 2015-09-02 | 秦志勇 | Express bill information identification system and express bill information identification method |
| CN106203416A (en) * | 2016-07-08 | 2016-12-07 | 武汉工程大学 | A kind of express delivery document information input method based on scanogram and input device |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5335290A (en) * | 1992-04-06 | 1994-08-02 | Ricoh Corporation | Segmentation of text, picture and lines of a document image |
| CN104408450A (en) * | 2014-11-21 | 2015-03-11 | 深圳天源迪科信息技术股份有限公司 | Identity card identifying method, device and system |
-
2017
- 2017-10-11 WO PCT/CN2017/105743 patent/WO2019071476A1/en active Application Filing
- 2017-10-11 CN CN201780095880.7A patent/CN111213157A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20060007204A (en) * | 2004-07-19 | 2006-01-24 | 한국과학기술원 | Document image processing and verification system for digitalizing a large volume of data and method thereof |
| CN1877598A (en) * | 2005-06-06 | 2006-12-13 | 英华达(上海)电子有限公司 | Method for gathering and recording business card information in mobile phone by using image recognition |
| CN104881770A (en) * | 2015-06-03 | 2015-09-02 | 秦志勇 | Express bill information identification system and express bill information identification method |
| CN106203416A (en) * | 2016-07-08 | 2016-12-07 | 武汉工程大学 | A kind of express delivery document information input method based on scanogram and input device |
Non-Patent Citations (2)
| Title |
|---|
| 赵珀璋等: "计算机中文信息处理", 30 June 1989, 宇航出版社, pages: 345 - 347 * |
| 钟义信等: "智能理论与技术—人工智能与神经网络", 31 December 1992, 人民邮电出版社, pages: 89 - 92 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110443538A (en) * | 2019-06-20 | 2019-11-12 | 苏州视印智能***有限公司 | A kind of addressee automatic information identifying system and method |
| CN112668574A (en) * | 2020-12-28 | 2021-04-16 | 深圳前海百递网络有限公司 | Parcel image processing method and device, computer equipment and storage medium |
| CN114780685A (en) * | 2022-04-28 | 2022-07-22 | 贵州电网有限责任公司 | Method for automatically identifying defect information input condition and supplementing defect information through unmanned aerial vehicle |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2019071476A1 (en) | 2019-04-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109829453B (en) | Method and device for recognizing characters in card and computing equipment | |
| US8744196B2 (en) | Automatic recognition of images | |
| US20210342571A1 (en) | Automated signature extraction and verification | |
| CN109919147A (en) | The method of text identification in drop for clothing image | |
| US20240037969A1 (en) | Recognition of handwritten text via neural networks | |
| WO2010092952A1 (en) | Pattern recognition device | |
| US11367310B2 (en) | Method and apparatus for identity verification, electronic device, computer program, and storage medium | |
| CN111488881A (en) | Method, device and storage medium for removing handwritten content in text image | |
| CN108090728B (en) | Express information input method and system based on intelligent terminal | |
| US20170351914A1 (en) | Method and device for classifying scanned documents | |
| Konya et al. | Fast seamless skew and orientation detection in document images | |
| CN113901952A (en) | Print form and handwritten form separated character recognition method based on deep learning | |
| CN111368632A (en) | Signature identification method and device | |
| CN111213157A (en) | Express information input method and system based on intelligent terminal | |
| CN113158895A (en) | Bill identification method and device, electronic equipment and storage medium | |
| US10115036B2 (en) | Determining the direction of rows of text | |
| CN114005127A (en) | Image optical character recognition method based on deep learning, storage device and server | |
| CN114581928A (en) | Form identification method and system | |
| WO2022082431A1 (en) | Systems and methods for extracting information from paper media based on depth information | |
| CN116994269A (en) | Seal similarity comparison method and seal similarity comparison system in image document | |
| CN114386413A (en) | Handling digitized handwriting | |
| Hadi et al. | A novel approach of skew estimation and correction in persian manuscript text using radon transform | |
| CN114627457A (en) | Ticket information identification method and device | |
| CN113887484B (en) | Card type file image identification method and device | |
| TWI773444B (en) | Image recognition system and method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |