CN108446702B - Image character segmentation method, device, equipment and storage medium - Google Patents
Image character segmentation method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN108446702B CN108446702B CN201810210701.6A CN201810210701A CN108446702B CN 108446702 B CN108446702 B CN 108446702B CN 201810210701 A CN201810210701 A CN 201810210701A CN 108446702 B CN108446702 B CN 108446702B
- Authority
- CN
- China
- Prior art keywords
- character
- segmentation
- value
- character string
- boundary
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/14—Image acquisition
- G06V30/148—Segmentation of character regions
- G06V30/153—Segmentation of character regions using recognition of characters or words
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Character Input (AREA)
Abstract
The embodiment of the invention discloses an image character segmentation method, an image character segmentation device, image character segmentation equipment and a storage medium, wherein the method comprises the following steps: dividing a character string image to be segmented into setting regions with equal areas, and determining the number and the value of character points included in each setting region; determining the upper boundary of the character string in the character string image to be segmented according to the number and the value and the upper boundary judgment condition; determining the lower boundary of the character string in the character string image to be segmented according to the upper boundary and the lower boundary judgment condition; determining a character segmentation position according to a preset segmentation condition in a line range corresponding to the upper boundary and the lower boundary, and performing column-direction segmentation on a character string in a character string image to be segmented according to the character segmentation position; the upper boundary and the lower boundary of a single character obtained by column direction segmentation are determined according to the search rule, so that the problems of missing, wrong and poor segmentation effect existing when characters with different sizes are segmented by the existing character segmentation method are solved, and the segmentation accuracy and the segmentation effect of the characters with different sizes are improved.
Description
Technical Field
The embodiment of the invention relates to the technical field of image processing, in particular to an image character segmentation method, device, equipment and storage medium.
Background
With the development of image processing technology, users often need to extract and identify character information contained in an image, for example, in the case of identifying whether a banknote is true or false, crown word number information corresponding to the banknote needs to be extracted. However, in the existing image character recognition method, only characters with uniform size in an image are generally segmented, and when characters with different sizes are contained in image data, wrong segmentation and missing segmentation often occur, so that the character segmentation effect is poor and the accuracy is low.
Disclosure of Invention
The embodiment of the invention provides an image character segmentation method, device, equipment and storage medium, which improve the segmentation accuracy and segmentation effect of characters with different sizes.
In a first aspect, an embodiment of the present invention provides an image character segmentation method, including:
dividing a character string image to be segmented into setting regions with equal areas, and determining the number and the value of character points included in each setting region; the character string image to be segmented is a binary image;
determining the upper boundary of the character string in the character string image to be segmented according to the number and the value and the upper boundary judgment condition;
determining the lower boundary of the character string in the character string image to be segmented according to the upper boundary and the lower boundary judgment condition;
determining character segmentation positions according to preset segmentation conditions in the line ranges corresponding to the upper boundary and the lower boundary, and performing column-direction segmentation on character strings in the character string image to be segmented according to the character segmentation positions;
and determining the upper boundary and the lower boundary of the single character obtained by column-direction segmentation according to a search rule.
In a second aspect, an embodiment of the present invention further provides an image character segmentation apparatus, including:
the value determining module is used for dividing the character string image to be segmented into setting regions with equal areas and determining the number and the value of character points included in each setting region; the character string image to be segmented is a binary image;
the upper boundary determining module is used for determining the upper boundary of the character string in the character string image to be segmented according to the number and the value and an upper boundary judging condition;
the lower boundary determining module is used for determining the lower boundary of the character string in the character string image to be segmented according to the upper boundary and the lower boundary judging condition;
the column direction segmentation module is used for determining a character segmentation position according to a preset segmentation condition in a row range corresponding to the upper boundary and the lower boundary, and performing column direction segmentation on the character string in the character string image to be segmented according to the character segmentation position;
and the boundary determining module is used for determining the upper boundary and the lower boundary of the single character obtained by column direction segmentation according to the search rule.
In a third aspect, an embodiment of the present invention further provides an apparatus, where the apparatus includes:
one or more processors;
storage means for storing one or more programs;
when executed by the one or more processors, cause the one or more processors to implement the method for image character segmentation provided by any of the embodiments of the present invention.
In a fourth aspect, an embodiment of the present invention further provides a computer storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the image character segmentation method provided in any embodiment of the present invention.
The method comprises the steps of dividing a character string image to be segmented into setting regions with equal areas, and determining the number and the value of character points included in each setting region; determining an upper boundary and a lower boundary of a character string in the character string image to be segmented according to the number and the value, the upper boundary judgment condition and the lower boundary judgment condition; in the line range corresponding to the upper boundary and the lower boundary, performing column-direction segmentation on the character strings in the character string image to be segmented according to preset segmentation conditions; the upper boundary and the lower boundary of a single character obtained by column-direction segmentation are determined according to the search rule, so that the problems of missing segmentation, wrong segmentation and poor segmentation effect existing when characters with different sizes are segmented by the existing character segmentation method are solved, and the segmentation accuracy and the segmentation effect of the characters with different sizes are improved.
Drawings
FIG. 1 is a flowchart illustrating an image character segmentation method according to an embodiment of the present invention;
FIG. 2 is a flowchart of a method for segmenting image characters according to a second embodiment of the present invention;
FIG. 3 is a flowchart of an image character segmentation method according to a third embodiment of the present invention;
FIG. 4 is a schematic structural diagram of an image character segmentation apparatus according to a fourth embodiment of the present invention;
fig. 5 is a schematic structural diagram of an apparatus according to a fifth embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting of the invention.
It should be further noted that, for the convenience of description, only some but not all of the relevant aspects of the present invention are shown in the drawings. Before discussing exemplary embodiments in more detail, it should be noted that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although a flowchart may describe the operations (or steps) as a sequential process, many of the operations can be performed in parallel, concurrently or simultaneously. In addition, the order of the operations may be re-arranged. The process may be terminated when its operations are completed, but may have additional steps not included in the figure. The processes may correspond to methods, functions, procedures, subroutines, and the like.
Example one
Fig. 1 is a flowchart of an image character segmentation method according to an embodiment of the present invention, where this embodiment is applicable to a case of segmenting an image containing characters with different sizes, and the method may be executed by an image character segmentation apparatus, where the apparatus may be implemented by software and/or hardware, and the method includes the following operations:
The character string image to be segmented is a binary image, and the set region is a rectangular region set according to requirements, such as 10 × 20. The area size of the set region may be adaptively set according to the character string image to be segmented, which is not limited in the embodiment of the present invention. One to-be-segmented character string image usually corresponds to a set region including hundreds or even thousands of equal areas, and the number of the set regions is specifically determined according to the to-be-segmented character string image and the size of the set region. For example, if the size of the character string image to be divided is 100 × 800 and the size of the setting region is 10 × 20, the number of the setting regions is 400, and the setting regions are uniformly arranged in the character string image to be divided.
In the embodiment of the present invention, before segmenting the characters in the character string image to be segmented, an area containing the character string to be segmented needs to be cut from the whole image to form the character string image to be segmented. The method for cutting the image according to the requirement is a mature technology in the prior art, and the embodiment of the invention is not detailed. After the cut-out region is acquired, binarization processing needs to be performed on the cut-out region. Optionally, the cut-out region is binarized by using an OTSU algorithm. The OTSU algorithm divides the original image into two images, i.e., a foreground image and a background image, by using a threshold, and assigns a character point in the foreground to 1 and a background to 0, so as to facilitate subsequent processing. It should be noted that, for different images, the light environment is different, and the exposure degree of the image is also different. The segmentation effect that may be finally caused by applying the binarization processing under the same condition (i.e. dividing the foreground and the background by the same threshold) to all the images with different exposure degrees is not ideal, especially to the image with a relatively serious exposure. Therefore, considering that the binarization effect directly affects the subsequent segmentation effect, the cut region can be subjected to binarization processing for multiple times. And returning when the segmentation of a certain image subjected to the binarization processing is correct, or replacing another threshold value for binarization processing. When the threshold is replaced, the threshold may be selected to be 0.8 times or 0.6 times of the threshold of the OTSU algorithm, and the specific replacement threshold may be set according to a requirement, which is not limited in the embodiment of the present invention.
Furthermore, after the binarization processing is performed on the character string image to be segmented, an integral image of the binarized image needs to be calculated, so as to accelerate the subsequent calculation of the number and value of the row and column character points and the number and value of the area character points.
Specifically, after the character string image to be segmented is divided into the setting regions with equal areas, the number and the value of the character points included in each setting region need to be determined, so as to base the subsequent determination of the boundaries of the character strings and the characters.
And step 120, determining the upper boundary of the character string in the character string image to be segmented according to the number and the value and the upper boundary judgment condition.
The upper boundary determination condition is a determination condition set for determining an upper boundary of a character string in a character string image to be segmented, and needs to be designed according to the number and the value of character points.
In the embodiment of the present invention, when determining the upper boundary of the character string in the divided character string image, the determination condition may be determined according to the number and value of the character points included in each setting region and the upper boundary.
And step 130, determining the lower boundary of the character string in the character string image to be segmented according to the upper boundary and the lower boundary judgment condition.
The lower boundary determination condition is a determination condition set for determining a lower boundary of a character string in the character string image to be segmented, and also needs to be designed according to the number and value of character points.
In the embodiment of the present invention, after the upper boundary of the character string in the divided character string image is determined, the lower boundary of the character string in the character string image to be divided needs to be determined according to the upper boundary of the character string in the divided character string image, the number and the value of the character points included in each setting region, and the lower boundary determination condition.
And 140, determining character segmentation positions according to preset segmentation conditions in the line ranges corresponding to the upper boundary and the lower boundary, and performing column-direction segmentation on the character strings in the character string image to be segmented according to the character segmentation positions.
The preset segmentation condition can be set according to the number and value of the character points and the prior knowledge of the characters. In the embodiment of the present invention, the character dividing position refers to a column-wise dividing position of a single character.
In the embodiment of the invention, after the upper boundary and the lower boundary of the character string in the character string image to be segmented are determined, the segmentation position of the left boundary and the right boundary of a single character can be determined in the line range corresponding to the upper boundary and the lower boundary. The advantage of this processing is that the calculation amount of the segmentation position of a single character can be reduced, and the accuracy of the segmentation position can be ensured. Even if other character points exist in the upper or lower area of the column where the single character segmentation position of the character string image is located in the cut-out area, the determination of the segmentation position is not affected.
And 150, determining the upper boundary and the lower boundary of the single character obtained by column-direction segmentation according to a search rule.
Wherein the search rule searches the line where the upper and lower boundaries of the single character are located and the associated line to determine the upper and lower boundaries of the single character.
In the embodiment of the invention, the determination of the upper boundary and the lower boundary of the single character also needs to be performed in the line range corresponding to the upper boundary and the lower boundary of the character string in the character string image to be segmented, so that the calculation amount for searching the upper boundary and the lower boundary of the single character is reduced, and the accuracy of the upper boundary and the lower boundary of the single character is ensured.
The method comprises the steps of dividing a character string image to be segmented into setting regions with equal areas, and determining the number and the value of character points included in each setting region; determining an upper boundary and a lower boundary of a character string in the character string image to be segmented according to the number and the value, the upper boundary judgment condition and the lower boundary judgment condition; in the line range corresponding to the upper boundary and the lower boundary, performing column-direction segmentation on the character strings in the character string image to be segmented according to preset segmentation conditions; the upper boundary and the lower boundary of a single character obtained by column-direction segmentation are determined according to the search rule, so that the problems of missing segmentation, wrong segmentation and poor segmentation effect existing when characters with different sizes are segmented by the existing character segmentation method are solved, and the segmentation accuracy and the segmentation effect of the characters with different sizes are improved.
Example two
Fig. 2 is a flowchart of an image character segmentation method according to a second embodiment of the present invention, and as shown in fig. 2, the method according to this embodiment may include:
In the embodiment of the present invention, when the upper boundary of the character string in the character string image to be divided is determined according to the number and the value and the upper boundary determination condition, the set areas in the divided character string image to be divided need to be determined one by one according to the sequence of each column. Namely, traversing calculation is carried out on each column of set areas in the divided character string image to be segmented.
The first preset value may be the same as the second preset value, or may be different from the second preset value. Optionally, the first preset value and the second preset value may be set to 10, the third preset value is smaller than the fourth preset value, optionally, the third preset value may be 20, and the fourth preset value may be 50. It should be noted that the preset value may be adjusted according to actual requirements, which is not limited in the embodiment of the present invention.
Specifically, when one of the rows of setting areas of the character string image to be segmented is determined, traversal calculation may be started from the first setting area of the row of setting areas. Before calculation, an initial value may be set in an array manner for each column of setting areas, for example, aiStartY [ i ]. about 100, when the number and value included in two consecutive setting areas before the first current processing area, and the first setting area after the first current processing area all satisfy the range requirement of the corresponding preset value, the upper boundary position of the first current processing area may be taken as a valid position, and at this time, the initial value of the first current processing area is replaced by another value in an array manner, for example, aiFLag1[ i ]. about 1. It should be noted that, in each column setting area, there is at most one valid position that meets the condition, and there is no valid position in some column setting areas.
And 240, if the sum of the number of the effective positions in a plurality of rows of the setting area in which the number is continuously set and arranged is greater than a preset threshold value, taking the plurality of rows of the setting area as alternative processing areas.
The set number may be 20, 30 or 40, and may be specifically designed according to the size of the set area, the size of a single character, and the number of characters in the character string image to be segmented. For example, if the size of a single character is 100 × 60, the size of a set region is 10 × 20, and 9 characters are included in a character string, the set number may be set to 20 or 30. The preset threshold may be 8, 9, or 10, and may be specifically set according to an actual requirement, which is not limited in the embodiment of the present invention. The number of the candidate processing regions may be 1 or more.
In the embodiment of the present invention, after the effective positions in the set areas of all the columns are determined, the effective positions may be used as a determination basis to preliminarily determine the column range in which the upper boundary of the character string in the character string image to be segmented is located. And when the sum of the number of effective positions in a multi-column setting area with the continuously set number of rows is larger than a preset threshold value, taking the multi-column setting area as an alternative processing area. For example, if the sum of the number of valid positions in all the columns with column numbers 10-30 is 9, which is greater than the preset threshold 8, all the columns with column numbers 10-30 are set as a candidate processing area. And if the sum of the number of effective positions in all the columns with the column numbers of 15-35 is 10 and is greater than the preset threshold value 8, setting the area of all the columns with the column numbers of 15-35 as another alternative processing area.
The first target processing area is an area including an upper boundary of a character string in the character string image to be segmented, and if the 10 th column to the 30 th column set area is the first target processing area, it indicates that the upper boundary of the character string in the character string image to be segmented is located in the 10 th column to the 30 th column set area.
Accordingly, when there are a plurality of candidate processing regions, it is necessary to select a region with the highest accuracy as the upper boundary from the plurality of candidate processing regions. The specific method is to calculate the variance of the effective positions in all the candidate processing areas, and determine the candidate processing area with the minimum variance as the first target processing area. For example, the 9 valid positions of the first candidate processing region are 4, 5, 6, 7, 8, 6, 5 and 4, respectively, and the corresponding variance is 1.322876, and the 9 valid positions of the second candidate processing region are 4, 7, 6, 2, 7, 10, 6, 5 and 4, respectively, and the corresponding variance is 2.291288. And if the variance of the first alternative processing area is the minimum, taking all effective positions in the first alternative processing area as alternative upper boundaries so as to obtain final upper boundaries after the alternative upper boundaries are processed in the following process.
Optionally, the first preset value may be 2, and may be adaptively designed according to the size of the set area, which is not limited in the embodiment of the present invention. The first preset value indicates the position of the setting area, and is set to avoid the character points of the partial character being located above the upper boundary.
Specifically, after all the alternative upper boundaries are obtained, the alternative upper boundaries may be sorted according to a certain rule, such as ascending or descending, then the first intermediate value of the alternative upper boundaries is obtained, and the first preset threshold is subtracted on the basis, and the pixel position is converted to be the position of the upper boundary of the character string in the character string image to be segmented. For example, if the upper boundaries are 4, 5, 6, 7, 8, 6, 5, and 4, respectively, and the upper boundaries are sorted into 4, 5, 6, 7, and 8, the first intermediate value is 6, 2 is subtracted from 6 to obtain 4, and the pixel position is converted into 40 by multiplying 4 by 10. The finally determined position of the boundary on the character string in the character string image to be segmented is 40. It should be noted that the finally determined position of the boundary on the character string in the image of the character string to be segmented is a horizontal straight line.
And step 270, setting each row of setting areas corresponding to the effective positions in the first target processing area as second target processing areas.
Wherein the second target processing area is an area including a lower boundary of the character string in the character string image to be segmented. The second target processing area does not need to be determined according to the number of the effective positions, whether the number of the effective positions exceeds a threshold value or not and the variance of the effective positions, each row of setting areas corresponding to the effective positions in the first target processing area is used as the second target processing area, and the lower boundary of the character string in the character string image to be segmented is obtained in the second target processing area according to the lower boundary judgment condition.
The seventh preset value may be the same as the eighth preset value, or may be different from the eighth preset value. Optionally, the seventh preset value and the eighth preset value may be set to 10, the fifth preset value is greater than the sixth preset value, optionally, the fifth preset value may be 40, and the sixth preset value may be 20. It should be noted that the preset value may be adjusted according to actual requirements, which is not limited in the embodiment of the present invention.
Specifically, when determining one row of setting areas of the second target processing area, the traversal calculation may be started from the first setting area of the row of setting areas. When the number and the value of two consecutive setting areas before the second current processing area, and the first setting area after the second current processing area all satisfy the range requirement of the corresponding preset value, the lower boundary position of the second current processing area can be used as one of the alternative lower boundaries. In addition, in each column setting area, there is at most one candidate lower boundary that meets the condition, and there is no candidate lower boundary in some column setting areas.
Optionally, the second preset value may be 1, and may be adaptively designed according to the size of the set area, which is not limited in the embodiment of the present invention. The second preset value indicates the position of the setting area, and is set to avoid the character points of the partial character being located below the lower boundary.
Specifically, after all the alternative lower boundaries are obtained, the alternative lower boundaries may be sorted according to a certain rule, such as ascending or descending, then a second intermediate value of the alternative lower boundaries is taken, a second preset threshold is subtracted on the basis, and the second intermediate value is converted into a pixel position to be used as the position of the upper boundary of the character string in the character string image to be segmented. For example, the alternative lower boundaries are 14, 15, 16, 17, 18, 16, 15, and 14, respectively, which are sorted into 14, 15, 16, 17, and 18, then the second intermediate value is 16, 1 is added on the basis of 16 to obtain 17, and the result of multiplying 17 by 10 is converted into a pixel position of 170. 170 is the finally determined position of the lower boundary of the character string in the character string image to be segmented. It should be noted that the finally determined position of the lower boundary of the character string in the image of the character string to be segmented is also a horizontal straight line.
And step 2120, determining an upper boundary and a lower boundary of the single character obtained by column-direction segmentation according to a search rule.
According to the method and the device, one set area of the character string image to be segmented is used as a current processing area, whether the number and the value of character points in two continuous set areas in front of the current processing area and a first set area behind the current processing area meet the corresponding range of preset values or not is judged respectively, the upper boundary of the character string in the character string image to be segmented is further determined according to the number and the variance of effective positions on the basis, and the lower boundary is determined according to the upper boundary and a similar sum value principle adopted by the determination of the upper boundary, so that the upper boundary and the lower boundary corresponding to the character string are accurately determined, and the segmentation accuracy and the segmentation effect of characters with different sizes are improved.
EXAMPLE III
Fig. 3 is a flowchart of an image character segmentation method according to a third embodiment of the present invention, and as shown in fig. 3, the method according to the third embodiment may include:
And step 320, determining the upper boundary of the character string in the character string image to be segmented according to the number and the value and the upper boundary judgment condition.
And step 330, determining a lower boundary of the character string in the character string image to be segmented according to the upper boundary and the lower boundary judgment condition.
And 340, determining character segmentation positions according to preset segmentation conditions in the line ranges corresponding to the upper boundary and the lower boundary, and performing column-direction segmentation on character strings in the character string image to be segmented according to the character segmentation positions.
Correspondingly, step 340 specifically includes:
In the embodiment of the invention, when the column-direction segmentation position of each character is determined in the line range corresponding to the upper boundary and the lower boundary, the number and the value of the character points do not need to be calculated by taking a set area as a unit, and the number and the value of each column of the character points in the image can be directly calculated. Specifically, the position where the sum of the number of character points in each column is 0 may be determined as the first candidate division position.
In the first candidate segmentation position determined in step 341, part of the positions may not be left and right boundaries of the character, and therefore, the first candidate segmentation position needs to be filtered. The screening of the first candidate segmentation locations may be performed in two steps. First, the number of character points included between the first candidate segmentation positions and the width between the first candidate segmentation positions may be used as a basis for the screening. For the character string image to be segmented, the width of a single character and the sum of the number of character points included in each single character can be obtained by a priori knowledge. Therefore, when the number of character points included between the currently judged first candidate segmentation position and other candidate segmentation positions and/or the width between the first candidate segmentation positions is smaller than a certain threshold, the currently judged first candidate segmentation position is considered to be unreasonable, and the currently judged first candidate segmentation position is excluded. And taking all the first alternative segmentation positions which finally meet the conditions as second alternative segmentation positions to complete one-time screening.
And 343, acquiring the variance of the difference values of the intermediate positions of the continuous preset number of second candidate segmentation positions, and taking the continuous preset number of second candidate segmentation positions corresponding to the minimum variance as the character segmentation positions.
The preset number corresponds to the number of characters included in the character string image to be segmented, namely when the number of characters included in the character string image to be segmented is 9, the preset number is also 9. The middle position of the second alternative segmentation position is the center position of a single character.
Accordingly, after the second candidate segmentation position is obtained, the second candidate segmentation position still needs to be further screened. The specific method comprises the following steps: and calculating the difference value of the middle positions of the second candidate segmentation positions of the continuous preset number, and taking the second candidate position corresponding to the minimum square difference in the difference values of the continuous preset number as the final reasonable character segmentation position. And the character segmentation position (left and right boundaries) is determined from the second alternative segmentation position by adopting a variance-based consistency method, so that the accuracy is high, and the adaptability to noise is strong.
And step 350, determining the upper boundary and the lower boundary of the single character obtained by column-direction segmentation according to a search rule.
Correspondingly, step 350 may specifically include:
In the embodiment of the invention, when the upper boundary and the lower boundary of a single character are determined, the character segmentation position is required to be used as the left boundary and the right boundary of the single character for calculation, so that the calculation amount can be effectively reduced, and the accuracy of the upper boundary and the lower boundary can be improved. The specific method comprises the following steps: in the line range of the upper boundary and the lower boundary of the character string image to be segmented determined in the above steps, the number of character points of each line is calculated from between the left boundary and the right boundary of a single character as a line and is used as a row.
Accordingly, the line and the maximum position can be used as the reference position. In order to ensure the accuracy of the reference position, the upper and lower boundaries of a single character may be found upward or downward based on the reference position using the row of several consecutive rows and the maximum position as the reference position.
And 353, when the row of the continuous number of rows is 0 above the reference position, taking the first row in the continuous number of rows as the upper boundary of the single character.
Specifically, in finding the upper boundary of a single character, it is possible to find upward from the reference position, and when a row of a consecutive number of rows occurs and is 0, the first row of the consecutive number of rows is taken as the upper boundary of the single character.
Similarly, when the lower boundary of a single character is found, it is possible to find from the reference position downward, and when a row of a consecutive number of rows appears and is 0, the first row of the consecutive number of rows is taken as the lower boundary of the single character.
In an optional embodiment of the present invention, the taking the preset number of consecutive second candidate segmentation positions corresponding to the minimum variance as the segmentation positions includes: acquiring at least one group of continuous preset number of second alternative segmentation positions corresponding to variances meeting preset requirements as an optimal segmentation position set; segmenting the character strings in the character string image to be segmented according to the optimal segmentation position set; acquiring the height accumulated sum of each character in the segmented character string; and taking the minimum height accumulation in the optimal segmentation position set and the corresponding optimal segmentation position as the character segmentation position.
The preset requirement may be that the variance is smaller than a certain threshold. The optimal segmentation location set includes all segmentation locations for which variance meets the requirement. That is, the optimal segmentation position set comprises at least one scheme for performing column-wise segmentation on characters in the character string image to be segmented.
In the embodiment of the present invention, after the optimal segmentation position set is obtained, after the characters in the character string image to be segmented are sequentially segmented in a row direction by using each group of character segmentation positions in the optimal segmentation position set, an upper boundary and a lower boundary of a single character are respectively determined for the characters. In order to determine the character segmentation position with the optimal effect, the characters segmented by each group of character segmentation positions need to be scored. The score here is a highly accumulated sum of each character. Since the single character does not fill the whole line, if a certain character point exists in a certain line in the single character within the line range corresponding to the upper boundary and the lower boundary of the character string image to be segmented, the heights of the characters are not accumulated. After the scoring is finished, the optimal character segmentation position can be determined according to the score of each character.
The embodiment of the invention sequentially determines the character segmentation position, the upper boundary and the lower boundary of a single character in the line range corresponding to the upper boundary and the lower boundary of the character string image to be segmented, segments the character string image to be segmented through a plurality of groups of character segmentation positions, scores the segmented characters after determining the upper boundary and the lower boundary, and determines the optimal character segmentation position according to the score, thereby reducing the calculation amount of character segmentation and improving the accuracy of character segmentation.
Example four
Fig. 4 is a schematic diagram of an image character segmentation apparatus provided in the fourth embodiment of the present invention, which is capable of executing the image character segmentation method provided in any embodiment of the present invention, and has functional modules and beneficial effects corresponding to the execution method.
The device comprises:
the sum value determining module 410 is configured to divide the character string image to be segmented into setting regions with equal areas, and determine the number and value of character points included in each setting region; the character string image to be segmented is a binary image;
an upper boundary determining module 420, configured to determine an upper boundary of a character string in the character string image to be segmented according to the number and the value and an upper boundary determination condition;
a lower boundary determining module 430, configured to determine a lower boundary of a character string in the character string image to be segmented according to the upper boundary and the lower boundary determination condition;
the column direction segmentation module 440 is configured to determine a character segmentation position according to a preset segmentation condition in a row range corresponding to the upper boundary and the lower boundary, and perform column direction segmentation on a character string in the character string image to be segmented according to the character segmentation position;
and a boundary determining module 450, configured to determine an upper boundary and a lower boundary of the single character obtained by column-wise segmentation according to a search rule.
The method comprises the steps of dividing a character string image to be segmented into setting regions with equal areas, and determining the number and the value of character points included in each setting region; determining an upper boundary and a lower boundary of a character string in the character string image to be segmented according to the number and the value, the upper boundary judgment condition and the lower boundary judgment condition; in the line range corresponding to the upper boundary and the lower boundary, performing column-direction segmentation on the character strings in the character string image to be segmented according to preset segmentation conditions; the upper boundary and the lower boundary of a single character obtained by column-direction segmentation are determined according to the search rule, so that the problems of missing segmentation, wrong segmentation and poor segmentation effect existing when characters with different sizes are segmented by the existing character segmentation method are solved, and the segmentation accuracy and the segmentation effect of the characters with different sizes are improved.
Optionally, the upper boundary determining module 420 is further configured to obtain one of the set areas in each row of the set areas as a first current processing area; if the quantity and value corresponding to the last-but-one setting area before the first current processing area are smaller than a first preset value, the quantity and value corresponding to the last-but-one setting area before the first current processing area are smaller than a second preset value, the quantity and value corresponding to the first current processing area are larger than a third preset value, the quantity and value corresponding to the first setting area after the first current processing area are larger than a fourth preset value, and the upper boundary position of the first current processing area is taken as an effective position; if the sum of the number of the effective positions in a plurality of rows of the setting area in which the number is continuously set and arranged is greater than a preset threshold value, taking the plurality of rows of the setting area as alternative processing areas; determining the candidate processing area with the minimum variance of the effective positions in the candidate processing areas as a first target processing area, and acquiring the effective position corresponding to each row of the setting area in the first target processing area as an alternative upper boundary; and sequencing all the alternative upper boundaries in the first target processing area to obtain a first intermediate value, and converting the first intermediate value minus a first preset numerical value into a pixel position as an upper boundary of a character in the character string image to be segmented.
Optionally, the lower boundary determining module 430 is further configured to use each row of setting areas corresponding to the effective position in the first target processing area as a second target processing area; acquiring one setting area of each row of setting areas in the second target processing area as a second current processing area, wherein if the quantity and value corresponding to the last setting area before the second current processing area are smaller than a fifth preset value, the quantity and value corresponding to the last setting area before the second current processing area are smaller than a sixth preset value, the quantity and value corresponding to the second current processing area are larger than a seventh preset value, the quantity and value corresponding to the first setting area after the second current processing area are larger than an eighth preset value, and the lower boundary of the second target processing area is used as an alternative lower boundary; and sequencing all the alternative lower boundaries in the second target processing area to obtain a second intermediate value, adding a second preset numerical value to the second intermediate value, and converting the second intermediate value into a pixel position to be used as the lower boundary of the characters in the character string image to be segmented.
Optionally, the column-wise segmentation module 440 is further configured to determine the sum of the number of the character points in each column and the position of 0 as a first candidate segmentation position; screening the first alternative segmentation positions according to the number of character points included among the first alternative segmentation positions and the width among the first alternative segmentation positions to obtain second alternative segmentation positions; and acquiring the variance of the difference values of the middle positions of the continuous preset number of second alternative segmentation positions, and taking the continuous preset number of second alternative segmentation positions corresponding to the minimum variance as the character segmentation positions.
Optionally, the boundary determining module 450 is further configured to take the character segmentation position as a left boundary and a right boundary of a single character, and calculate the number of character points in each line between the left boundary and the right boundary as a line and; taking the row and the maximum position as a reference position; when a row of a consecutive number of rows appears above the reference position and is 0, taking a first row of the consecutive number of rows as an upper boundary of the single character; when a line of a consecutive number of lines appears below the reference position and is 0, a first line of the consecutive number of lines is taken as a lower boundary of the single character.
Optionally, the column-wise segmentation module 440 is further configured to obtain at least one group of second candidate segmentation positions, corresponding to the variance meeting the preset requirement, in a continuous preset number, as an optimal segmentation position set; segmenting the character strings in the character string image to be segmented according to the optimal segmentation position set; acquiring the height accumulated sum of each character in the segmented character string; and taking the minimum height accumulation in the optimal segmentation position set and the corresponding optimal segmentation position as the character segmentation position.
The image character segmentation device can execute the image character segmentation method provided by any embodiment of the invention, and has corresponding functional modules and beneficial effects of the execution method. For details of the technique not described in detail in this embodiment, reference may be made to the method for segmenting image characters provided in any embodiment of the present invention.
EXAMPLE five
Fig. 5 is a schematic structural diagram of an apparatus according to a fifth embodiment of the present invention, as shown in fig. 5, the apparatus includes a processor 510, a memory 520, an input device 530, and an output device 540; the number of the processors 510 in the device may be one or more, and one processor 510 is taken as an example in fig. 5; the processor 510, the memory 520, the input device 530 and the output device 540 of the apparatus may be connected by a bus or other means, as exemplified by the bus connection in fig. 5.
The memory 520, which is a computer-readable storage medium, may be used to store software programs, computer-executable programs, and modules, such as program instructions/modules corresponding to the image character segmentation method in the embodiments of the present invention (e.g., the sum value determination module 410, the upper boundary determination module 420, the lower boundary determination module 430, the column direction segmentation module 440, and the boundary determination module 450 in the image character segmentation apparatus). The processor 510 implements the above-described method of image character segmentation by executing software programs, instructions, and modules stored in the memory 520 to perform various functional applications of the device and data processing.
The memory 520 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application program required for at least one function; the storage data area may store data created according to the use of the terminal, and the like. Further, memory 520 may include high speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid state storage device. In some examples, memory 520 may further include memory located remotely from processor 510, which may be connected to devices through a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The input device 530 may be used to receive input numeric or character information and generate key signal inputs related to user settings and function control of the apparatus. The output device 540 may include a display device such as a display screen.
EXAMPLE six
An embodiment of the present invention further provides a storage medium containing computer-executable instructions, which when executed by a computer processor, perform a method for image character segmentation, the method including:
dividing a character string image to be segmented into setting regions with equal areas, and determining the number and the value of character points included in each setting region; the character string image to be segmented is a binary image;
determining the upper boundary of the character string in the character string image to be segmented according to the number and the value and the upper boundary judgment condition;
determining the lower boundary of the character string in the character string image to be segmented according to the upper boundary and the lower boundary judgment condition;
determining a character segmentation position according to a preset segmentation condition in a line range corresponding to the upper boundary and the lower boundary, and performing column-direction segmentation on a character string in the character string image to be segmented according to the character segmentation position;
and determining the upper boundary and the lower boundary of the single character obtained by column-direction segmentation according to a search rule.
Of course, the storage medium containing the computer-executable instructions provided by the embodiments of the present invention is not limited to the method operations described above, and may also perform related operations in the image character segmentation method provided by any embodiments of the present invention.
From the above description of the embodiments, it is obvious for those skilled in the art that the present invention can be implemented by software and necessary general hardware, and certainly, can also be implemented by hardware, but the former is a better embodiment in many cases. Based on such understanding, the technical solutions of the present invention may be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as a floppy disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a FLASH Memory (FLASH), a hard disk or an optical disk of a computer, and includes several instructions for enabling a computer device (which may be a personal computer, a server, or a network device) to execute the methods according to the embodiments of the present invention.
It should be noted that, in the embodiment of the image character segmentation apparatus, the included units and modules are only divided according to functional logic, but are not limited to the above division as long as the corresponding functions can be realized; in addition, specific names of the functional units are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the present invention.
It is to be noted that the foregoing is only illustrative of the preferred embodiments of the present invention and the technical principles employed. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, although the present invention has been described in greater detail by the above embodiments, the present invention is not limited to the above embodiments, and may include other equivalent embodiments without departing from the spirit of the present invention, and the scope of the present invention is determined by the scope of the appended claims.
Claims (8)
1. An image character segmentation method, comprising:
dividing a character string image to be segmented into setting regions with equal areas, and determining the number and the value of character points included in each setting region; the character string image to be segmented is a binary image;
determining the upper boundary of the character string in the character string image to be segmented according to the number and the value and the upper boundary judgment condition;
determining the lower boundary of the character string in the character string image to be segmented according to the upper boundary and the lower boundary judgment condition;
determining character segmentation positions according to preset segmentation conditions in the line ranges corresponding to the upper boundary and the lower boundary, and performing column-direction segmentation on character strings in the character string image to be segmented according to the character segmentation positions;
determining an upper boundary and a lower boundary of a single character obtained by column direction segmentation according to a search rule;
determining the upper boundary of the character string in the character string image to be segmented according to the number and the value and the upper boundary judgment condition, wherein the determining comprises the following steps:
acquiring one of the set areas in each row of the set areas as a first current processing area;
if the quantity and value corresponding to the last-but-one setting area before the first current processing area are smaller than a first preset value, the quantity and value corresponding to the last-but-one setting area before the first current processing area are smaller than a second preset value, the quantity and value corresponding to the first current processing area are larger than a third preset value, the quantity and value corresponding to the first setting area after the first current processing area are larger than a fourth preset value, and the upper boundary position of the first current processing area is taken as an effective position;
if the sum of the number of the effective positions in a plurality of rows of the setting area in which the number is continuously set and arranged is greater than a preset threshold value, taking the plurality of rows of the setting area as alternative processing areas;
determining the candidate processing area with the minimum variance of the effective positions in the candidate processing areas as a first target processing area, and acquiring the effective position corresponding to each row of the setting area in the first target processing area as an alternative upper boundary;
sequencing all the alternative upper boundaries in the first target processing area to obtain a first intermediate value, and converting the first intermediate value minus a first preset numerical value into a pixel position as an upper boundary of a character in the character string image to be segmented;
determining the lower boundary of the characters in the character string image to be segmented according to the upper boundary and the lower boundary judgment condition, wherein the determining the lower boundary comprises the following steps:
taking each row of setting areas corresponding to the effective positions in the first target processing area as a second target processing area;
acquiring one setting area of each row of setting areas in the second target processing area as a second current processing area, wherein if the quantity and value corresponding to the last setting area before the second current processing area are smaller than a fifth preset value, the quantity and value corresponding to the last setting area before the second current processing area are smaller than a sixth preset value, the quantity and value corresponding to the second current processing area are larger than a seventh preset value, the quantity and value corresponding to the first setting area after the second current processing area are larger than an eighth preset value, and the lower boundary of the second target processing area is used as an alternative lower boundary;
and sequencing all the alternative lower boundaries in the second target processing area to obtain a second intermediate value, adding a second preset numerical value to the second intermediate value, and converting the second intermediate value into a pixel position to be used as the lower boundary of the characters in the character string image to be segmented.
2. The method according to claim 1, wherein the determining the character segmentation position according to the preset segmentation condition comprises:
determining the position of 0 sum of the number of the character points in each column as a first alternative segmentation position;
screening the first alternative segmentation positions according to the number of character points included among the first alternative segmentation positions and the width among the first alternative segmentation positions to obtain second alternative segmentation positions;
and acquiring the variance of the difference values of the middle positions of the continuous preset number of second alternative segmentation positions, and taking the continuous preset number of second alternative segmentation positions corresponding to the minimum variance as the character segmentation positions.
3. The method of claim 1, wherein determining the upper boundary and the lower boundary of a single character in the character string image to be segmented according to a search rule comprises:
taking the character segmentation positions as a left boundary and a right boundary of a single character, and calculating the number of character points of each line between the left boundary and the right boundary as a line and a row;
taking the row and the maximum position as a reference position;
when a row of a consecutive number of rows appears above the reference position and is 0, taking a first row of the consecutive number of rows as an upper boundary of the single character;
when a line of a consecutive number of lines appears below the reference position and is 0, a first line of the consecutive number of lines is taken as a lower boundary of the single character.
4. The method according to claim 2, wherein the taking the preset number of consecutive second candidate segmentation positions corresponding to the minimum variance as segmentation positions comprises:
acquiring at least one group of continuous preset number of second alternative segmentation positions corresponding to variances meeting preset requirements as an optimal segmentation position set;
segmenting the character strings in the character string image to be segmented according to the optimal segmentation position set;
acquiring the height accumulated sum of each character in the segmented character string;
and taking the minimum height accumulation in the optimal segmentation position set and the corresponding optimal segmentation position as the character segmentation position.
5. An image character segmentation apparatus, comprising:
the value determining module is used for dividing the character string image to be segmented into setting regions with equal areas and determining the number and the value of character points included in each setting region; the character string image to be segmented is a binary image;
the upper boundary determining module is used for determining the upper boundary of the character string in the character string image to be segmented according to the number and the value and an upper boundary judging condition;
the lower boundary determining module is used for determining the lower boundary of the character string in the character string image to be segmented according to the upper boundary and the lower boundary judging condition;
the column direction segmentation module is used for determining a character segmentation position according to a preset segmentation condition in a row range corresponding to the upper boundary and the lower boundary, and performing column direction segmentation on the character string in the character string image to be segmented according to the character segmentation position;
the boundary determining module is used for determining the upper boundary and the lower boundary of the single character obtained by column direction segmentation according to the search rule;
the upper boundary determining module is further configured to obtain one of the set areas in each row of the set areas as a first current processing area; if the quantity and value corresponding to the last-but-one setting area before the first current processing area are smaller than a first preset value, the quantity and value corresponding to the last-but-one setting area before the first current processing area are smaller than a second preset value, the quantity and value corresponding to the first current processing area are larger than a third preset value, the quantity and value corresponding to the first setting area after the first current processing area are larger than a fourth preset value, and the upper boundary position of the first current processing area is taken as an effective position; if the sum of the number of the effective positions in a plurality of rows of the setting areas in which the number is continuously set and arranged is larger than a preset threshold value, taking the plurality of rows of the setting areas as alternative processing areas; determining the candidate processing area with the minimum variance of the effective positions in the candidate processing areas as a first target processing area, and acquiring the effective position corresponding to each row of the setting area in the first target processing area as an alternative upper boundary; sequencing all the alternative upper boundaries in the first target processing area to obtain a first intermediate value, and converting the first intermediate value minus a first preset numerical value into a pixel position as an upper boundary of a character in the character string image to be segmented;
the lower boundary determining module is further configured to use each row of setting areas corresponding to the effective positions in the first target processing area as a second target processing area; acquiring one setting area of each row of setting areas in the second target processing area as a second current processing area, wherein if the quantity and value corresponding to the last setting area before the second current processing area are smaller than a fifth preset value, the quantity and value corresponding to the last setting area before the second current processing area are smaller than a sixth preset value, the quantity and value corresponding to the second current processing area are larger than a seventh preset value, the quantity and value corresponding to the first setting area after the second current processing area are larger than an eighth preset value, and the lower boundary of the second target processing area is used as an alternative lower boundary; and sequencing all the alternative lower boundaries in the second target processing area to obtain a second intermediate value, adding a second preset numerical value to the second intermediate value, and converting the second intermediate value into a pixel position to be used as the lower boundary of the characters in the character string image to be segmented.
6. The apparatus of claim 5, comprising:
the column direction segmentation module is further used for determining the sum of the number of the character points in each column and the position of 0 as a first alternative segmentation position;
screening the first alternative segmentation positions according to the number of character points included among the first alternative segmentation positions and the width among the first alternative segmentation positions to obtain second alternative segmentation positions;
acquiring the variance of the difference value of the middle positions of the continuous preset number of second alternative segmentation positions, and taking the continuous preset number of second alternative segmentation positions corresponding to the minimum variance as the character segmentation positions;
the boundary determining module is further configured to take the character segmentation position as a left boundary and a right boundary of a single character, and calculate the number of character points in each row between the left boundary and the right boundary as a row and a column;
taking the row and the maximum position as a reference position;
when a row of a consecutive number of rows appears above the reference position and is 0, taking a first row of the consecutive number of rows as an upper boundary of the single character;
when a line of a consecutive number of lines appears below the reference position and is 0, taking a first line of the consecutive number of lines as a lower boundary of the single character;
the column direction segmentation module is further used for acquiring at least one group of second alternative segmentation positions which meet the preset requirement and correspond to the variances in a continuous preset number to serve as an optimal segmentation position set;
segmenting the character strings in the character string image to be segmented according to the optimal segmentation position set;
acquiring the height accumulated sum of each character in the segmented character string;
and taking the minimum height accumulation in the optimal segmentation position set and the corresponding optimal segmentation position as the character segmentation position.
7. An electronic device, characterized in that the device comprises:
one or more processors;
a storage device for storing one or more programs,
when executed by the one or more processors, cause the one or more processors to implement the method of image character segmentation as claimed in any one of claims 1-4.
8. A computer storage medium on which a computer program is stored which, when being executed by a processor, carries out the method of image character segmentation according to any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810210701.6A CN108446702B (en) | 2018-03-14 | 2018-03-14 | Image character segmentation method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810210701.6A CN108446702B (en) | 2018-03-14 | 2018-03-14 | Image character segmentation method, device, equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108446702A CN108446702A (en) | 2018-08-24 |
| CN108446702B true CN108446702B (en) | 2022-05-31 |
Family
ID=63195068
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810210701.6A Active CN108446702B (en) | 2018-03-14 | 2018-03-14 | Image character segmentation method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108446702B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109299718B (en) * | 2018-09-21 | 2021-09-24 | 新华三信息安全技术有限公司 | Character recognition method and device |
| CN110378347B (en) * | 2019-07-04 | 2021-10-08 | 北京爱医生智慧医疗科技有限公司 | Method and device for extracting key information of medical examination sheet |
| CN110502293B (en) * | 2019-07-10 | 2022-02-01 | 维沃移动通信有限公司 | Screen capturing method and terminal equipment |
| CN115050015B (en) * | 2022-08-15 | 2023-09-12 | 恒银金融科技股份有限公司 | Accurate segmentation method for character area of financial bill account |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101251892A (en) * | 2008-03-07 | 2008-08-27 | 北大方正集团有限公司 | Method and apparatus for cutting character |
| CN102496019A (en) * | 2011-12-08 | 2012-06-13 | 银江股份有限公司 | License plate character segmenting method |
| CN103136845A (en) * | 2013-01-23 | 2013-06-05 | 浙江大学 | Renminbi (RMB) counterfeit identifying method based on crown-word image characters |
| CN103455815A (en) * | 2013-08-27 | 2013-12-18 | 电子科技大学 | Self-adaptive license plate character segmentation method in complex scene |
| CN104239879A (en) * | 2014-09-29 | 2014-12-24 | 小米科技有限责任公司 | Character segmentation method and device |
| CN104616009A (en) * | 2015-02-13 | 2015-05-13 | 广州广电运通金融电子股份有限公司 | Character cutting and recognizing method |
| CN105243730A (en) * | 2015-10-28 | 2016-01-13 | 深圳怡化电脑股份有限公司 | Banknote identification method and system |
| CN105550686A (en) * | 2015-12-09 | 2016-05-04 | 小米科技有限责任公司 | Character segmentation method and device |
| US9378435B1 (en) * | 2014-06-10 | 2016-06-28 | David Prulhiere | Image segmentation in optical character recognition using neural networks |
| CN106250362A (en) * | 2015-06-05 | 2016-12-21 | 富士通株式会社 | Text segmentation device and text segmenting method |
| CN106384113A (en) * | 2016-11-07 | 2017-02-08 | 湖南源信光电科技有限公司 | Odd-and-even-line license plate character dividing method based on projection and template matching |
| CN107220981A (en) * | 2017-06-22 | 2017-09-29 | 深圳怡化电脑股份有限公司 | Character segmentation method, device, equipment and storage medium |
| CN107437294A (en) * | 2017-08-01 | 2017-12-05 | 深圳怡化电脑股份有限公司 | A kind of character segmentation method, device, equipment and storage medium |
-
2018
- 2018-03-14 CN CN201810210701.6A patent/CN108446702B/en active Active
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101251892A (en) * | 2008-03-07 | 2008-08-27 | 北大方正集团有限公司 | Method and apparatus for cutting character |
| CN102496019A (en) * | 2011-12-08 | 2012-06-13 | 银江股份有限公司 | License plate character segmenting method |
| CN103136845A (en) * | 2013-01-23 | 2013-06-05 | 浙江大学 | Renminbi (RMB) counterfeit identifying method based on crown-word image characters |
| CN103455815A (en) * | 2013-08-27 | 2013-12-18 | 电子科技大学 | Self-adaptive license plate character segmentation method in complex scene |
| US9378435B1 (en) * | 2014-06-10 | 2016-06-28 | David Prulhiere | Image segmentation in optical character recognition using neural networks |
| CN104239879A (en) * | 2014-09-29 | 2014-12-24 | 小米科技有限责任公司 | Character segmentation method and device |
| CN104616009A (en) * | 2015-02-13 | 2015-05-13 | 广州广电运通金融电子股份有限公司 | Character cutting and recognizing method |
| CN106250362A (en) * | 2015-06-05 | 2016-12-21 | 富士通株式会社 | Text segmentation device and text segmenting method |
| CN105243730A (en) * | 2015-10-28 | 2016-01-13 | 深圳怡化电脑股份有限公司 | Banknote identification method and system |
| CN105550686A (en) * | 2015-12-09 | 2016-05-04 | 小米科技有限责任公司 | Character segmentation method and device |
| CN106384113A (en) * | 2016-11-07 | 2017-02-08 | 湖南源信光电科技有限公司 | Odd-and-even-line license plate character dividing method based on projection and template matching |
| CN107220981A (en) * | 2017-06-22 | 2017-09-29 | 深圳怡化电脑股份有限公司 | Character segmentation method, device, equipment and storage medium |
| CN107437294A (en) * | 2017-08-01 | 2017-12-05 | 深圳怡化电脑股份有限公司 | A kind of character segmentation method, device, equipment and storage medium |
Non-Patent Citations (4)
| Title |
|---|
| Research of improving the accuracy of license plate character segmentation;Qiao S等;《2010 Fifth International Conference on Frontier of Computer Science and Technology》;20101231;第489-493页 * |
| Segmentation of characters on car license plates;He X等;《2008 IEEE 10th Workshop on Multimedia Signal Processing》;20081231;第399-402页 * |
| 一种新的车牌图像字符分割与识别算法;陈振学等;《微电子学与计算机》;20071231;第24卷(第2期);第42-44页 * |
| 基于结构特征的纸币号码识别方法;苑玮琦等;《计算机工程与应用》;20141231;第50卷(第8期);第118-121页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108446702A (en) | 2018-08-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108446702B (en) | Image character segmentation method, device, equipment and storage medium | |
| CN110516208B (en) | System and method for extracting PDF document form | |
| US20190188528A1 (en) | Text detection method and apparatus, and storage medium | |
| CN111259854B (en) | Method and device for identifying structured information of table in text image | |
| CN110473238B (en) | Image registration method and device, electronic equipment and storage medium | |
| CN110619333A (en) | Text line segmentation method, text line segmentation device and electronic equipment | |
| CN116403094B (en) | Embedded image recognition method and system | |
| JP2019512127A (en) | String distance calculation method and apparatus | |
| CN111967545B (en) | Text detection method and device, electronic equipment and computer storage medium | |
| CN109724988B (en) | PCB defect positioning method based on multi-template matching | |
| CN112906695B (en) | Form recognition method adapting to multi-class OCR recognition interface and related equipment | |
| CN105335741A (en) | Smudged serial number classification method and system | |
| CN112307803A (en) | Digital geological outcrop crack extraction method and device | |
| CN115937065A (en) | Foreign matter detection method, device and equipment of display module and storage medium | |
| CN112580624B (en) | Multi-directional text region detection method and device based on boundary prediction | |
| CN111340000A (en) | Method and system for extracting and optimizing PDF document table | |
| CN114357958A (en) | Table extraction method, device, equipment and storage medium | |
| CN108460386B (en) | Character image cutting method, device, equipment and storage medium | |
| CN115457581A (en) | Table extraction method and device and computer equipment | |
| JP2012003358A (en) | Background determination device, method, and program | |
| CN114332667A (en) | Corn plant type identification method and device, electronic equipment and storage medium | |
| CN114897801A (en) | AOI defect detection method, device and equipment and computer medium | |
| CN110619273B (en) | Efficient iris recognition method and recognition device | |
| CN110570437B (en) | Electric power channel automatic inspection data processing method based on boundary identification | |
| CN109409370B (en) | Remote desktop character recognition method and device |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |