CN111709419A - Method, system and equipment for positioning banknote serial number and readable storage medium - Google Patents
Method, system and equipment for positioning banknote serial number and readable storage medium Download PDFInfo
- Publication number
- CN111709419A CN111709419A CN202010521386.6A CN202010521386A CN111709419A CN 111709419 A CN111709419 A CN 111709419A CN 202010521386 A CN202010521386 A CN 202010521386A CN 111709419 A CN111709419 A CN 111709419A
- Authority
- CN
- China
- Prior art keywords
- image
- area
- edge
- stroke width
- paper money
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/62—Text, e.g. of license plates, overlay texts or captions on TV images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/255—Detecting or recognising potential candidate objects based on visual cues, e.g. shapes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07D—HANDLING OF COINS OR VALUABLE PAPERS, e.g. TESTING, SORTING BY DENOMINATIONS, COUNTING, DISPENSING, CHANGING OR DEPOSITING
- G07D7/00—Testing specially adapted to determine the identity or genuineness of valuable papers or for segregating those which are unacceptable, e.g. banknotes that are alien to a currency
- G07D7/004—Testing specially adapted to determine the identity or genuineness of valuable papers or for segregating those which are unacceptable, e.g. banknotes that are alien to a currency using digital security elements, e.g. information coded on a magnetic thread or strip
- G07D7/0047—Testing specially adapted to determine the identity or genuineness of valuable papers or for segregating those which are unacceptable, e.g. banknotes that are alien to a currency using digital security elements, e.g. information coded on a magnetic thread or strip using checkcodes, e.g. coded numbers derived from serial number and denomination
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07D—HANDLING OF COINS OR VALUABLE PAPERS, e.g. TESTING, SORTING BY DENOMINATIONS, COUNTING, DISPENSING, CHANGING OR DEPOSITING
- G07D7/00—Testing specially adapted to determine the identity or genuineness of valuable papers or for segregating those which are unacceptable, e.g. banknotes that are alien to a currency
- G07D7/20—Testing patterns thereon
- G07D7/2016—Testing patterns thereon using feature extraction, e.g. segmentation, edge detection or Hough-transformation
Abstract
The application provides a method, a system and equipment for positioning a banknote serial number and a readable storage medium, wherein edge detection is performed on a gray image of a banknote image to obtain an edge image; performing stroke width conversion on an original image according to image edge points of an edge image and a gradient direction to obtain a stroke width image, wherein the stroke width image comprises a character area; forming a character area of the stroke width image into a connected area; and respectively screening each communication area according to the geometric characteristics of the paper money crown word area to determine the paper money crown word area. The paper money serial number can be quickly and accurately positioned.
Description
Technical Field
The present application relates to the field of image processing technologies, and in particular, to a method, a system, a device, and a readable storage medium for positioning a serial number of a banknote.
Background
This section is intended to provide a background or context to the embodiments of the application that are recited in the claims. The description herein is not admitted to be prior art by inclusion in this section.
In order to maintain social stability and promote the development of economic globalization, the method has important practical significance for effectively supervising the circulation of the paper money. The crown word number is used as a unique identification mark of the paper money, and the monitoring of the crown word number is an important link for monitoring the circulation process of the paper money. The serial number on the RMB refers to a string of codes at the lower left corner of the paper money, and the serial number can be used as 'identity' information of each paper money. At present, most of ATM machines of banks can realize the inquiry of the serial number of the RMB with the denomination of 100 yuan, and the paper money is judged through automatic identification and detection, thereby bringing convenience to the life of people. The characteristic that the paper money is easy to damage in the circulation process is adopted, so that the accuracy of identifying the serial number of the paper money is improved, and the method is an important research direction. Meanwhile, the premise that the crown word number characters can be correctly identified is that the position of the crown word number region needs to be accurately and completely determined. Therefore, the research of a high-speed and effective banknote serial number area positioning method is an indispensable step in the process of monitoring the banknote serial number.
The existing text region positioning technology and some deep learning frameworks specially designed for text characteristics can hardly achieve real-time effect on equipment with poor performance in terms of efficiency. In certain application scenes, good effects can be achieved, but similar characteristics can be obtained for certain scenes, so that misjudgment can be easily caused for an image containing many other special background textures, such as a banknote. Besides, the banknote image has the characteristics of low noise and high resolution, and the requirement of the application scene on time efficiency is high.
In the prior art, a scheme with positioning accuracy and timeliness does not exist.
Disclosure of Invention
The embodiment of the application provides a method, a system and equipment for positioning a paper money serial number and a readable storage medium, and improves the accuracy and timeliness of positioning the paper money serial number.
In a first aspect, an embodiment of the present application provides a method for positioning a banknote serial number, where the method includes:
carrying out edge detection on the gray level image of the paper money image to obtain an edge image;
performing stroke width conversion on an original image according to image edge points of an edge image and a gradient direction to obtain a stroke width image, wherein the stroke width image comprises a character area;
forming a character area of the stroke width image into a connected area;
and respectively screening each communication area according to the geometric characteristics of the paper money crown word area to determine the paper money crown word area.
Optionally, after determining the banknote crown number region, the method further comprises:
judging whether boundary expansion is needed according to the boundary information of the paper currency crown word number area, and if so, performing boundary expansion on the existing area in the paper currency crown word number area;
and performing relocation operation on the character area after the boundary expansion, wherein the relocation operation comprises the steps of acquiring boundary information of the character area by using a sliding window and determining the boundary information of the banknote crown word number area by using a projection method.
Optionally, the performing of boundary expansion on the basis of the existing region in the banknote crown word region includes:
when W is less than WdThen, the left and right boundaries of the paper money crown word number area are respectively expanded by Wd-W; and/or
When H is less than HdThen, the upper and lower boundaries of the paper money crown word number area are respectively expanded by Hd-H; wherein, WdThe maximum width of the region of the banknote crown number, HdThe height of the paper money is the maximum value of the paper money crown number area, W is the width of the paper money crown number area, and H is the height of the paper money crown number area.
Optionally, after the edge detection is performed on the grayscale image of the banknote image to obtain an edge image, the method further includes:
noise filtering is performed on the edge image by using the prior information to filter out at least one of small-area noise, large-area noise and linear noise.
Optionally, the performing stroke width transformation on the original image according to the image edge points of the edge image and the gradient direction to obtain a stroke width image includes:
performing convolution operation on the original image by using a transverse operator and a longitudinal operator to obtain transverse and longitudinal image gradients;
calculating the gradient direction of the image edge points according to the transverse and longitudinal image gradients;
and determining the stroke width image of the image edge points according to the image edge points and the gradient directions corresponding to the image edge points.
Optionally, the transverse image gradient GxAnd the longitudinal image gradient is calculated according to the following formula:
where I represents the original image.
Optionally, the gradient direction of the image edge point is calculated according to the following formula:
wherein G isxRepresenting transverse image gradients, GyRepresenting the longitudinal image gradient.
Optionally, the determining the stroke width image of the image edge point according to the image edge point and the gradient direction corresponding to the image edge point includes:
determining a ray l according to an image edge point p and a gradient direction theta corresponding to the image edge point;
along the ray direction, gradually increasing the coefficient n within the limit of the maximum stroke width, and determining an image edge point q;
determining the width of the character stroke according to the image edge points p and q;
searching in the image edge points with the set stroke distance, and determining a stroke width image of each image edge point;
wherein the ray l is calculated according to the following formula:
l＝p+n·θ,(n≥0)。
optionally, the forming the character region of the stroke width image into a connected region includes:
utilizing an open operation on a character area of the stroke width image to filter small-noise stroke pixels;
filling up gaps between character areas by using corrosion operation;
connecting the single character areas constitutes the connected area.
Optionally, the respectively screening the communication areas according to the geometric features of the banknote crown word number area to determine the banknote crown word number area includes:
determining a circumscribed rectangle frame of each connected region;
screening each communicated area according to the height and width of the corresponding circumscribed rectangle frame:
and determining the screened communication area as a paper currency crown word number area.
Optionally, before performing edge detection on the grayscale image of the banknote image, the method further comprises:
and performing inclination correction processing, orientation detection processing and gray-scale processing on the banknote image.
In a second aspect, an embodiment of the present application provides a system for positioning a serial number of a banknote, where the system includes:
the edge detection module is used for carrying out edge detection on the gray level image of the paper money image to obtain an edge image;
the stroke width changing module is used for carrying out stroke width conversion on the original image according to the image edge points of the edge image and the gradient direction to obtain a stroke width image, and the stroke width image comprises a character area;
the communication module is used for forming a communication area in the character area of the stroke width image;
and the communicated region filtering module is used for respectively screening each communicated region according to the geometric characteristics of the paper money crown word number region to determine the paper money crown word number region.
Optionally, the system further comprises:
the edge correction module is used for judging whether the boundary expansion is needed according to the boundary information of the paper currency crown word number area, and if so, performing the boundary expansion on the basis of the existing area in the paper currency crown word number area; and performing relocation operation on the character area after the boundary expansion, wherein the relocation operation comprises the steps of acquiring boundary information of the character area by using a sliding window and determining the boundary information of the banknote crown word number area by using a projection method.
Optionally, the performing of boundary expansion on the basis of the existing region in the banknote crown word region includes:
when W is less than WdThen, the left and right boundaries of the paper money crown word number area are respectively expanded by Wd-W; and/or
When H is less than HdThen, the upper and lower boundaries of the paper money crown word number area are respectively expanded by Hd-H; wherein, WdThe maximum width of the region of the banknote crown number, HdThe height of the paper money is the maximum value of the paper money crown number area, W is the width of the paper money crown number area, and H is the height of the paper money crown number area.
Optionally, the edge detection module is further configured to:
noise filtering is performed on the edge image by using the prior information to filter out at least one of small-area noise, large-area noise and linear noise.
Optionally, the stroke width changing module is specifically configured to:
performing convolution operation on the original image by using a transverse operator and a longitudinal operator to obtain transverse and longitudinal image gradients;
calculating the gradient direction of the image edge points according to the transverse and longitudinal image gradients;
and determining the stroke width image of the image edge points according to the image edge points and the gradient directions corresponding to the image edge points.
Optionally, the transverse image gradient GxAnd the longitudinal image gradient is calculated according to the following formula:
where I represents the original image.
Optionally, the gradient direction of the image edge point is calculated according to the following formula:
wherein G isxRepresenting transverse image gradients, GyRepresenting the longitudinal image gradient.
Optionally, the determining the stroke width image of the image edge point according to the image edge point and the gradient direction corresponding to the image edge point includes:
determining a ray l according to an image edge point p and a gradient direction theta corresponding to the image edge point;
along the ray direction, gradually increasing the coefficient n within the limit of the maximum stroke width, and determining an image edge point q;
determining the width of the character stroke according to the image edge points p and q;
searching in the image edge points with the set stroke distance, and determining a stroke width image of each image edge point;
wherein the ray l is calculated according to the following formula:
l＝p+n·θ,(n≥0)。
optionally, the communication module is specifically configured to:
utilizing an open operation on a character area of the stroke width image to filter small-noise stroke pixels;
filling up gaps between character areas by using corrosion operation;
connecting the single character areas constitutes the connected area.
Optionally, the connected region filtering module is specifically configured to:
determining a circumscribed rectangle frame of each connected region;
screening each communicated area according to the height and width of the corresponding circumscribed rectangle frame:
and determining the screened communication area as a paper currency crown word number area.
Optionally, the initial processing includes performing a skew correction process, an orientation detection process, and a graying process on the banknote image.
In a third aspect, an embodiment of the present application provides a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor implements any one of the methods of the first aspect when executing the computer program.
In a fourth aspect, the present application provides a computer-readable storage medium storing a computer program for executing the method of any one of the above first aspects.
In summary, the method, the system, the device and the readable storage medium for positioning the serial number of the banknote provided by the application obtain an edge image by performing edge detection on a gray image of a banknote image; further, stroke width conversion is carried out on the original image according to the image edge points of the edge image and the gradient direction to obtain a stroke width image, the stroke width image comprises a character area, so that the limitation of the global maximum stroke width is increased and the limitation angle is widened aiming at the specific characteristics of the paper money image, and the calculation efficiency of the algorithm is improved. Further, forming a character area of the stroke width image into a connected area; and respectively screening each communication area according to the geometric characteristics of the paper money crown word area to determine the paper money crown word area. The characteristics of the paper money image sample are better fitted, and the accuracy of positioning the paper money serial numbers is improved.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts. In the drawings:
fig. 1 is a schematic flow chart of a method for positioning a serial number of a banknote provided in an embodiment of the present application;
fig. 2 is a flowchart of a positioning method provided in an embodiment of the present application;
FIG. 3 is a schematic diagram of a stroke width provided in an embodiment of the present application;
FIG. 4 is a schematic illustration of the position of a horizontal sliding window in a vertical projection view provided in an embodiment of the present application;
fig. 5 is a block diagram of a positioning system for a banknote serial number provided in an embodiment of the present application.
Fig. 6 is a schematic structural diagram of a computer device suitable for implementing the method for positioning the serial number of the banknote according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and the described embodiments are only a part of the embodiments of the present application, but not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
The principles and spirit of the present application are explained in detail below with reference to several representative embodiments of the present application.
Although the present application provides method operational steps or apparatus configurations as illustrated in the following examples or figures, more or fewer operational steps or modular units may be included in the methods or apparatus based on conventional or non-inventive efforts. In the case of steps or structures which do not logically have the necessary cause and effect relationship, the execution sequence of the steps or the module structure of the apparatus is not limited to the execution sequence or the module structure shown in the embodiment or the drawings of the present application. The described methods or modular structures, when applied in an actual device or end product, may be executed sequentially or in parallel according to embodiments or the methods or modular structures shown in the figures.
In the prior art, text region positioning refers to a technology for correctly outputting the position of a text region for an input image, and for a commonly used positioning algorithm of a natural scene text region, the positioning algorithm can be roughly divided into a positioning algorithm based on connected region analysis, a positioning algorithm based on edge detection analysis and a positioning algorithm based on deep learning.
The positioning algorithm based on the connected region analysis is to distinguish the text region and the background region based on the consistency region in the text information characteristic, thereby achieving the purpose of positioning. Because fewer candidate connected regions are generally obtained, and such methods can accommodate rotation, scaling and font, such methods have become one of the mainstream algorithms in the field of natural scene text localization.
The edge detection-based method utilizes the difference of the contrast ratio between a text area and a background area, most of the difference is reflected in the edge area, and meanwhile, the edge area has rich gradient information, so the method generally adopts an edge detection operator to detect the edge information of an image, and then removes the non-text area by utilizing heuristic rules or morphological processing on the edge image.
The difficulty of the edge detection algorithm lies in the balance of the edge detection on the precision, the false edge is generated due to the fact that the detection precision is excessively improved, and the edge detection is inaccurate due to the fact that the noise resistance is improved.
With the development of artificial intelligence, the method of deep learning is also gradually applied to text positioning, and since text can also be regarded as a special target, applying a framework of target detection (e.g. R-CNN, SSD) to text detection is the first idea of text detection. Secondly, there are some deep learning frameworks designed specifically for text characteristics, such as CTPN, EAST. However, the method is based on the detection of a text line with a large text area in a natural scene, and the detection effect on a small target is still to be improved; meanwhile, most network structures are deep and complex in hierarchy, and for equipment with poor performance, the real-time effect is difficult to achieve in efficiency.
The method can achieve good effect in a certain application scene, but because the above characteristics are not only unique to text characteristics, but also have similar characteristics for some scenes, erroneous judgment can be easily caused for the image containing many other special background textures, such as paper money. Besides, the banknote image has the characteristics of low noise and high resolution, and the requirement of the application scene on time efficiency is high, so the method is not suitable.
For a high-noise low-resolution image such as a banknote image, a positioning method with stable performance and better noise resistance is required to meet the requirement of actual production. Strokes are important features of characters, stroke widths are important features for distinguishing character areas from non-character areas, and accuracy of stroke width extraction has great influence on accuracy of final character area positioning. Therefore, the embodiment of the application provides an improved method for positioning the crown word number of the paper currency by stroke width conversion by using the premise that the stroke width is the unique characteristic of the text character, combining the unique characteristic of the paper currency picture sample and applying the related technology of picture processing.
Fig. 1 shows a flow of a method for positioning a banknote serial number according to an embodiment of the present application, which eliminates interference of complex banknote texture, reduces subsequent calculation amount, and improves accuracy of an algorithm. Aiming at specific characteristics of the paper money image, the limitation of the global maximum stroke width is increased, the limitation angle is widened, the formation of a communication area is realized by adopting morphological operation, the characteristics of the paper money image sample are better fitted, and the positioning accuracy is improved. The method comprises the following steps:
step 101: and carrying out edge detection on the gray level image of the paper money image to obtain an edge image.
Step 102: and carrying out stroke width conversion on the original image according to the image edge points of the edge image and the gradient direction to obtain a stroke width image, wherein the stroke width image comprises a character area.
Step 103: the character areas of the stroke width image are formed into connected areas.
Step 104: and respectively screening each communication area according to the geometric characteristics of the paper money crown word area to determine the paper money crown word area.
In some embodiments, the method further comprises:
step 105: and judging whether the boundary expansion is needed according to the boundary information of the paper currency crown word number area, and if so, performing the boundary expansion on the basis of the existing area in the paper currency crown word number area. Specific ways of boundary extension include, but are not limited to, the following: when W is less than WdThen, the left and right boundaries of the paper money crown word number area are respectively expanded by Wd-W; and/or when H < HdThen, the upper and lower boundaries of the paper money crown word number area are respectively expanded by Hd-H; wherein, WdThe maximum width of the region of the banknote crown number, HdThe height of the paper money is the maximum value of the paper money crown number area, W is the width of the paper money crown number area, and H is the height of the paper money crown number area.
Step 106: and performing relocation operation on the character area after the boundary expansion, wherein the relocation operation comprises the steps of acquiring boundary information of the character area by using a sliding window and determining the boundary information of the banknote crown word number area by using a projection method. And by further edge correction, two steps of expansion and relocation are adopted, so that the positioning result is further improved.
In some embodiments, the initial processing involved in step 101 includes, but is not limited to, a tilt correction process, an orientation detection process, and a graying process.
In some embodiments, the performing edge noise filtering on the initially processed banknote image in step 101 may specifically include: noise filtering is performed on the edge image by using the prior information to filter out at least one of small-area noise, large-area noise and linear noise. By adopting the edge filtering method, the interference of complicated texture of the paper money is greatly eliminated, the subsequent calculated amount is reduced, and the accuracy of the algorithm is improved.
In some embodiments, in step 102, first, performing a convolution operation on an original image by using a horizontal operator and a vertical operator to obtain horizontal and vertical image gradients; further, calculating the gradient direction of the image edge points according to the transverse and longitudinal image gradients; further, determining the stroke width image of the image edge points according to the image edge points and the gradient directions corresponding to the image edge points.
The transverse image gradient G involved in step 102xAnd the longitudinal image gradient can be calculated according to the following equations (1) and (2):
where I represents the original image.
The gradient direction of the image edge point referred to in step 102 can be calculated according to the following formula (3):
wherein G isxRepresenting transverse image gradients, GyRepresenting the longitudinal image gradient.
In step 102, determining the stroke width image of the image edge point according to the image edge point and the gradient direction corresponding to the image edge point may include the following steps: determining a ray l according to an image edge point p and a gradient direction theta corresponding to the image edge point; along the ray direction, gradually increasing the coefficient n within the limit of the maximum stroke width, and determining an image edge point q; determining the width of the character stroke according to the image edge points p and q; searching in the image edge points with the set stroke distance, and determining a stroke width image of each image edge point; wherein the ray l is calculated according to the following formula (4):
p + n θ, (n ≧ 0) … … formula (4)
In some embodiments, in step 103, the forming the connected region in the obtained stroke width image by using morphological operation may take the following steps: utilizing an open operation on a character area of the stroke width image to filter small-noise stroke pixels; filling up gaps between character areas by using corrosion operation; connecting the single character areas constitutes the connected area. Aiming at specific characteristics of the paper money image, the limitation of the global maximum stroke width is increased, the limitation angle is widened, the formation of a communication area is realized by adopting morphological operation, the characteristics of the paper money image sample are better fitted, and the positioning accuracy is improved.
In some embodiments, in step 104, the screening of all connected regions according to the prior information to determine the banknote crown word number region may take the following steps: determining a circumscribed rectangle frame of each connected region; screening each communicated area according to the height and width of the corresponding circumscribed rectangle frame: and determining the screened communication area as a paper currency crown word number area.
In order to make the banknote serial number positioning method provided in the embodiment of the present application clearer and more definite, fig. 2 is an overall flowchart of the banknote serial number positioning method.
Step 201: the edge detection and the edge filtering are further performed on the gray-scaled banknote image subjected to the skew correction and the orientation determination. Specifically, the edge detection operator obtains an edge detection map and filters the initial edge map.
The background of the banknote image of each country is complex, a large number of the banknote image are pattern areas, the number of pixel points of the obtained edge image is often large, and the required crown word number area only occupies a small part of the whole image, so that the method provided by the embodiment of the application adopts an edge filtering strategy before formal stroke width conversion. Starting from the following two aspects, the edge pixels of the non-character strokes are eliminated, and the calculation pressure of a subsequent algorithm is relieved.
In a first aspect, both too small areas and too large areas of noise are cleaned. According to the statistical data in the historical experimental process, the edge image of each character is theoretically a closed curve, the closed curve belongs to the same connected region, and the number of pixel points of the connected region of the edge of each character is generally not less than 13. However, in the process of edge detection, the situation that edge connection lines are broken due to excessive noise interference is considered, and in order to avoid the situation that edge images broken in a small area are judged by mistake, edge pixel points with the number of edge pixels smaller than 5 are set to be removed, and the edge pixels are considered not to form the stroke edges of the characters. In addition, the maximum height or width of each character is 10 pixels by statistics, so that it is considered that when the number of pixels in the connected region is greater than 100, the connected region cannot form the stroke edge of the character.
In a second aspect, linear noise is removed. In theory, the edges of the character are all circular loops, so that for the connected region in the edge image with the difference of no more than 3 pixels, the connected region is considered to be close to a horizontal straight line in the edge image, and cannot form the stroke edge of the character. For a vertical straight line, since the width of the character 1 may be less than or equal to 2, only when the left-right difference of the connected region is 0, it is considered that the connected region presents a vertical straight line in the image and cannot constitute the stroke edge of the character.
Step 202: stroke Width Transformation (SWT) is performed. And performing edge filtering on the original edge image by using the related prior information, and then determining a stroke width transformation graph according to a new stroke rule.
After the filtered edge image is obtained, Sobel operators can be adopted to carry out filtering operation on the image, the method is divided into a transverse operator and a longitudinal operator which carry out convolution operation on the image respectively to obtain gradient images in two directions, and if I is used for representing an original image, a transverse G is adoptedXAnd a longitudinal direction GYThe two gradient direction calculation formulas are as described in the above formulas (1) and (2). After the gradients in the horizontal and vertical directions are obtained, the gradient direction θ of the pixel point can be calculated according to the formula (3). Assuming that each edge pixel point is p, and the corresponding gradient direction is θ, a ray l ═ p + n · θ (n ≧ 0l) can be determined, where + n · (n denotes a multiplication operator, along the direction of the ray, within the limit of the maximum stroke width, n is gradually increased, and another edge point q is found, as shown in fig. 3, where the length of the line segment formed by the middle vertical shadow is the stroke width of the character.
Because the target area is the character area of the number of the crown word and the stroke width of the character is in a certain range, all edge pixel points in the gradient direction do not need to be searched in the searching process, and only the range that the stroke distance does not exceed 5 pixels is needed to be searched, so that the calculation efficiency of the algorithm is greatly improved. In addition, because the characters in the crown word size area are numbers and capital letters, most strokes are curves which are not horizontal and vertical like the strokes of the text characters, the angle limit of the gradient direction of the composing strokes is relaxed and is set as dp + dq < pi/2. And finally, carrying out second search to obtain a stroke width transformation diagram.
Step 203: and forming a connected region by using morphological operation, and screening the connected region by using related prior information.
After obtaining the stroke width image, the character areas need to be connected, so that the final overall area can be determined conveniently. Since the image affine may cause many holes between the capital characters, and the interval between capital characters is much larger than the interval between text characters, if the method of forming the character candidate region under the condition that the stroke width ratio in the eight-direction field does not exceed 3 is adopted, characters with far part spacing are omitted, so that the morphological operation is proposed to solve the problem.
Firstly, filtering out some fine noise stroke pixels by utilizing open operation, then filling gaps among characters by using corrosion operation, and connecting scattered single character areas into a large connected area to form a connected area of a crown word number area. Since the crown word number region generally appears on the banknote image in a horizontal regular manner, a rectangular element with the length of 20 and the height of 2 is adopted when morphological operation is carried out, and the connection of two single character regions is ensured to be the maximum possible.
After all connected areas are obtained, the minimum rectangular frame externally connected with each connected area is obtained, and the rectangular height is recorded as HCWidth of WCSince the crown word size region has certain geometric characteristics, the connected region can be screened according to the following formula (5):
the minimum height of a single character is 5 pixels statistically obtained in experiments, so that the height of the circumscribed rectangle is limited firstly when the connected region is screened. And secondly, the crown word number region is a horizontal long rectangular region, so that the connected region is screened by adopting the length-width ratio. The screening formula has only the lowest limit and does not have the upper limit, because the crown word number character areas are hidden between pattern backgrounds, and when morphological operation corrosion is carried out, the areas are easily included into the communication areas, so that the areas are too high or too wide. To avoid missing this, there is no upper limit within these constraints. After screening, the approximate area left behind, which is considered to be the crown word number area, is a preliminary result of the location of the crown word number area.
Step 204: and the positioning area is corrected by adopting expansion and repositioning processing, and a small amount of noise similar to the stroke width is processed, so that a more accurate and close positioning result is obtained.
After the steps, most samples can be completely and closely attached to the boundary of the character area, but the occurrence position of noise in the circulation process of the paper money is not fixed, partial noise may cause the phenomena of character stroke distortion and the like in the affine process of an image, errors may be generated when morphological operations are performed, and some characters are omitted. In order to effectively solve the problems of missing characters or incomplete character areas and the like, after the corresponding crown word number area is obtained, further expansion and repositioning operations are carried out on the edge image. The expansion and repositioning operations are based on a heuristic mode of image information of the paper currency crown word number area, and the paper currency crown word number positioning algorithm is optimized through analysis of a large number of paper currency samples and experimental results, so that the robustness of the algorithm is improved, and the algorithm can obtain more ideal positioning results for various samples with different texture background interferences.
The first step of edge correction is to expand the boundary on the basis of the existing positioning area. Setting the maximum width of the crown word number region as WdMaximum height of HdThe width W and height H of the area obtained by the preliminary positioning. When W is less than WdThen, the left and right boundaries are expanded outward by Wd-W; for the same reason, when H is less than HdThen extend the upper and lower boundaries to Hd-H to ensure that the region contains all the capitalized characters in its entirety.
The repositioning is an operation of, after the expansion operation, causing the boundary of the character location to recede to the place containing all characters, and although the integrity of the characters is ensured, the location area is not compact enough, so that the character area needs to be repositioned again. The repositioning process is carried out in two steps, including sliding window to obtain the character area and determining accurate boundary information via projection.
First, for the crown word region, part of pattern noise may be included in the edge portion of the region during the expansion process, and in order to remove the interference information of this portion, based on the prior information that the character region is known to be relatively continuous, it is proposed to determine the approximate range of the character region by fixing the sliding window of the maximum rectangular regionA method. Because the gray value of the character edge pixel point in the edge image is 1, and the background is 0, when the gray value and the maximum value in the fixed area are known, the area is the character area at the maximum probability. Specifically, it is necessary to first take a vertical projection of this edge region and then use a length WdThe sliding window moves on the vertically projected image, and the coordinate information obtained when the maximum pixel value sum is obtained by calculation is the left and right boundaries of the approximate character region determined here, as shown in fig. 4; for the same reason, the length of use is HdThe sliding window of (2) finds the upper and lower boundaries of the character area on the horizontally projected image.
Second, the size of each crown region is not consistent, and we need to accurately determine the position of the crown region based on the largest rectangular region obtained through the sliding window. Since the crown region obtained at this time is small through the extraction of the previous step and can be regarded as not containing noise, we need only simply project here to specifically determine the boundaries of the crown region up, down, left, and right.
In summary, the method, the system, the device and the readable storage medium for positioning the serial number of the banknote provided by the application obtain an edge image by performing edge detection on a gray image of a banknote image; further, stroke width conversion is carried out on the original image according to the image edge points of the edge image and the gradient direction to obtain a stroke width image, the stroke width image comprises a character area, so that the limitation of the global maximum stroke width is increased and the limitation angle is widened aiming at the specific characteristics of the paper money image, and the calculation efficiency of the algorithm is improved. Further, forming a character area of the stroke width image into a connected area; and respectively screening each communication area according to the geometric characteristics of the paper money crown word area to determine the paper money crown word area. The characteristics of the paper money image sample are better fitted, and the accuracy of positioning the paper money serial numbers is improved.
Based on the same technical concept, the embodiment of the present application further provides a system for positioning a serial number of a banknote, as shown in fig. 5, the system includes:
and the edge detection module 501 is configured to perform edge detection on the grayscale image of the banknote image to obtain an edge image.
The stroke width changing module 502 is configured to perform stroke width conversion on the original image according to the image edge points of the edge image and the gradient direction to obtain a stroke width image, where the stroke width image includes a character region.
And a communicating module 503, configured to form a communicating region from the character region of the stroke width image.
And the connected region filtering module 504 is configured to respectively filter each connected region according to the geometric features of the banknote serial number region, and determine the banknote serial number region.
Optionally, the system further comprises: the edge correction module is used for judging whether the boundary expansion is needed according to the boundary information of the paper currency crown word number area, and if so, performing the boundary expansion on the basis of the existing area in the paper currency crown word number area; and performing relocation operation on the character area after the boundary expansion, wherein the relocation operation comprises the steps of acquiring boundary information of the character area by using a sliding window and determining the boundary information of the banknote crown word number area by using a projection method.
Optionally, the performing of boundary expansion on the basis of the existing region in the banknote crown word region includes:
when W is less than WdThen, the left and right boundaries of the paper money crown word number area are respectively expanded by Wd-W; and/or when H < HdThen, the upper and lower boundaries of the paper money crown word number area are respectively expanded by Hd-H; wherein, WdThe maximum width of the region of the banknote crown number, HdThe height of the paper money is the maximum value of the paper money crown number area, W is the width of the paper money crown number area, and H is the height of the paper money crown number area.
Optionally, the edge detection module 501 is further configured to: noise filtering is performed on the edge image by using the prior information to filter out at least one of small-area noise, large-area noise and linear noise.
Optionally, the stroke width changing module 502 is specifically configured to: performing convolution operation on the original image by using a transverse operator and a longitudinal operator to obtain transverse and longitudinal image gradients; calculating the gradient direction of the image edge points according to the transverse and longitudinal image gradients; and determining the stroke width image of the image edge points according to the image edge points and the gradient directions corresponding to the image edge points.
In some embodiments, the transverse image gradient GxAnd the longitudinal image gradient is calculated according to equations (1) and (2).
In some embodiments, the gradient direction θ of the image edge point is calculated according to equation (3).
In some embodiments, the determining the stroke width image of the image edge point according to the image edge point and the gradient direction corresponding to the image edge point includes: determining a ray l according to an image edge point p and a gradient direction theta corresponding to the image edge point; along the ray direction, gradually increasing the coefficient n within the limit of the maximum stroke width, and determining an image edge point q; determining the width of the character stroke according to the image edge points p and q; searching in the image edge points with the set stroke distance, and determining a stroke width image of each image edge point; wherein the ray l is calculated according to formula (4).
In some embodiments, the communication module 503 is specifically configured to: utilizing an open operation on a character area of the stroke width image to filter small-noise stroke pixels; filling up gaps between character areas by using corrosion operation; connecting the single character areas constitutes the connected area.
In some embodiments, the connected region filtering module 504 is specifically configured to: determining a circumscribed rectangle frame of each connected region; screening each communicated area according to the height and width of the corresponding circumscribed rectangle frame: and determining the screened communication area as a paper currency crown word number area.
In some embodiments, the initial processing includes performing a skew correction process, an orientation detection process, and a graying process on the banknote image.
In terms of hardware, in order to provide an embodiment of the electronic device for implementing all or part of the contents in the method for positioning the serial number of the banknote, the electronic device specifically includes the following contents:
a processor (processor), a memory (memory), a communication Interface (Communications Interface), and a bus; the processor, the memory and the communication interface complete mutual communication through the bus; the communication interface is used for realizing information transmission among related equipment such as a server, a device, a distributed message middleware cluster device, various databases, a user terminal and the like; the electronic device may be a desktop computer, a tablet computer, a mobile terminal, and the like, but the embodiment is not limited thereto. In this embodiment, the electronic device may refer to the embodiment of the method for positioning a serial number of a banknote and the embodiment of the method and apparatus for positioning a serial number of a banknote in the embodiments, which are incorporated herein by reference, and repeated details are not repeated herein.
Fig. 6 is a schematic block diagram of a system configuration of an electronic device 9600 according to an embodiment of the present invention. As shown in fig. 6, the electronic device 9600 can include a central processor 9100 and a memory 9140; the memory 9140 is coupled to the central processor 9100. Notably, this FIG. 6 is exemplary; other types of structures may also be used in addition to or in place of the structure to implement telecommunications or other functions.
In one embodiment, the function of the method of positioning the banknote serial number may be integrated into the central processor 9100. For example, the central processor 9100 may be configured to control as follows:
step 101: and carrying out edge detection on the gray level image of the paper money image to obtain an edge image.
Step 102: and carrying out stroke width conversion on the original image according to the image edge points of the edge image and the gradient direction to obtain a stroke width image, wherein the stroke width image comprises a character area.
Step 103: the character areas of the stroke width image are formed into connected areas.
Step 104: and respectively screening each communication area according to the geometric characteristics of the paper money crown word area to determine the paper money crown word area.
As can be seen from the foregoing description, the electronic device according to the embodiment of the present invention obtains a filtered edge image by performing edge noise filtering on the initially processed banknote image; performing stroke width transformation according to the image edge points of the edge image and the gradient direction to obtain a stroke width image; forming a connected region in the obtained stroke width image by using morphological operation; and screening all the communicated areas according to the prior information to determine the paper money crown word number area. The paper money serial number can be quickly and accurately positioned.
In another embodiment, the device for positioning the banknote serial number may be configured separately from the central processor 9100, for example, the device for positioning the banknote serial number may be configured as a chip connected to the central processor 9100, and the function of the positioning method of the banknote serial number is realized by the control of the central processor.
As shown in fig. 6, the electronic device 9600 may further include: a communication module 9110, an input unit 9120, an audio processor 9130, a display 9160, and a power supply 9170. It is noted that the electronic device 9600 also does not necessarily include all of the components shown in fig. 6; further, the electronic device 9600 may further include components not shown in fig. 6, which may be referred to in the art.
As shown in fig. 6, a central processor 9100, sometimes referred to as a controller or operational control, can include a microprocessor or other processor device and/or logic device, which central processor 9100 receives input and controls the operation of the various components of the electronic device 9600.
The memory 9140 can be, for example, one or more of a buffer, a flash memory, a hard drive, a removable media, a volatile memory, a non-volatile memory, or other suitable device. The information relating to the failure may be stored, and a program for executing the information may be stored. And the central processing unit 9100 can execute the program stored in the memory 9140 to realize information storage or processing, or the like.
The input unit 9120 provides input to the central processor 9100. The input unit 9120 is, for example, a key or a touch input device. Power supply 9170 is used to provide power to electronic device 9600. The display 9160 is used for displaying display objects such as images and characters. The display may be, for example, an LCD display, but is not limited thereto.
The memory 9140 can be a solid state memory, e.g., Read Only Memory (ROM), Random Access Memory (RAM), a SIM card, or the like. There may also be a memory that holds information even when power is off, can be selectively erased, and is provided with more data, an example of which is sometimes called an EPROM or the like. The memory 9140 could also be some other type of device. Memory 9140 includes a buffer memory 9141 (sometimes referred to as a buffer). The memory 9140 may include an application/function storage portion 9142, the application/function storage portion 9142 being used for storing application programs and function programs or for executing a flow of operations of the electronic device 9600 by the central processor 9100.
The memory 9140 can also include a data store 9143, the data store 9143 being used to store data, such as contacts, digital data, pictures, sounds, and/or any other data used by an electronic device. The driver storage portion 9144 of the memory 9140 may include various drivers for the electronic device for communication functions and/or for performing other functions of the electronic device (e.g., messaging applications, contact book applications, etc.).
The communication module 9110 is a transmitter/receiver 9110 that transmits and receives signals via an antenna 9111. The communication module (transmitter/receiver) 9110 is coupled to the central processor 9100 to provide input signals and receive output signals, which may be the same as in the case of a conventional mobile communication terminal.
Based on different communication technologies, a plurality of communication modules 9110, such as a cellular network module, a bluetooth module, and/or a wireless local area network module, may be provided in the same electronic device. The communication module (transmitter/receiver) 9110 is also coupled to a speaker 9131 and a microphone 9132 via an audio processor 9130 to provide audio output via the speaker 9131 and receive audio input from the microphone 9132, thereby implementing ordinary telecommunications functions. The audio processor 9130 may include any suitable buffers, decoders, amplifiers and so forth. In addition, the audio processor 9130 is also coupled to the central processor 9100, thereby enabling recording locally through the microphone 9132 and enabling locally stored sounds to be played through the speaker 9131.
An embodiment of the present invention also provides a computer-readable storage medium capable of implementing all the steps in the method for positioning a banknote serial number, where the execution main body in the above embodiment may be a server, and the computer-readable storage medium has a computer program stored thereon, where the computer program, when executed by a processor, implements all the steps in the method for positioning a banknote serial number in the above embodiment.
As can be seen from the above description, the computer-readable storage medium provided in the embodiment of the present invention obtains a filtered edge image by performing edge noise filtering on the initially processed banknote image; performing stroke width transformation according to the image edge points of the edge image and the gradient direction to obtain a stroke width image; forming a connected region in the obtained stroke width image by using morphological operation; and screening all the communicated areas according to the prior information to determine the paper money crown word number area. The paper money serial number can be quickly and accurately positioned.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, apparatus, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (devices), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The principle and the implementation mode of the invention are explained by applying specific embodiments in the invention, and the description of the embodiments is only used for helping to understand the method and the core idea of the invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present invention.
Claims (12)
1. A method for positioning a banknote serial number, the method comprising:
carrying out edge detection on the gray level image of the paper money image to obtain an edge image;
performing stroke width conversion on an original image according to image edge points of an edge image and a gradient direction to obtain a stroke width image, wherein the stroke width image comprises a character area;
forming a character area of the stroke width image into a connected area;
and respectively screening each communication area according to the geometric characteristics of the paper money crown word area to determine the paper money crown word area.
2. The method of claim 1, wherein after determining the banknote crown size region, the method further comprises:
judging whether boundary expansion is needed according to the boundary information of the paper currency crown word number area, and if so, performing boundary expansion on the existing area in the paper currency crown word number area;
and performing relocation operation on the character area after the boundary expansion, wherein the relocation operation comprises the steps of acquiring boundary information of the character area by using a sliding window and determining the boundary information of the banknote crown word number area by using a projection method.
3. The method of claim 2, wherein said performing a boundary extension on the basis of an existing region in said banknote crown mark region comprises:
when W is less than WdThen, the left and right boundaries of the paper money crown word number area are respectively expanded by Wd-W; and/or
When H is less than HdThen, the upper and lower boundaries of the paper money crown word number area are respectively expanded by Hd-H; wherein, WdThe maximum width of the region of the banknote crown number, HdThe height of the paper money is the maximum value of the paper money crown number area, W is the width of the paper money crown number area, and H is the height of the paper money crown number area.
4. The method of claim 1, wherein after performing edge detection on the gray scale image of the banknote image to obtain an edge image, further comprising:
noise filtering is performed on the edge image by using the prior information to filter out at least one of small-area noise, large-area noise and linear noise.
5. The method of claim 1, wherein the stroke width transforming the original image according to the image edge points of the edge image and the gradient direction to obtain the stroke width image comprises:
performing convolution operation on the original image by using a transverse operator and a longitudinal operator to obtain transverse and longitudinal image gradients;
calculating the gradient direction of the image edge points according to the transverse and longitudinal image gradients;
and determining the stroke width image of the image edge points according to the image edge points and the gradient directions corresponding to the image edge points.
6. The method of claim 5, wherein determining the stroke width image of the image edge points based on the image edge points and the gradient directions corresponding to the image edge points comprises:
determining a ray l according to an image edge point p and a gradient direction theta corresponding to the image edge point;
along the ray direction, gradually increasing the coefficient n within the limit of the maximum stroke width, and determining an image edge point q;
determining the width of the character stroke according to the image edge points p and q;
searching in the image edge points with the set stroke distance, and determining a stroke width image of each image edge point;
wherein the ray l is calculated according to the following formula:
l＝p+n·θ,(n≥0)。
7. the method of claim 1, wherein forming the character regions of the stroke width image into connected regions comprises:
utilizing an open operation on a character area of the stroke width image to filter small-noise stroke pixels;
filling up gaps between character areas by using corrosion operation;
connecting the single character areas constitutes the connected area.
8. The method of claim 1, wherein the step of screening each connected region according to the geometric characteristics of the banknote crown size region to determine the banknote crown size region comprises:
determining a circumscribed rectangle frame of each connected region;
screening each communicated area according to the height and width of the corresponding circumscribed rectangle frame:
and determining the screened communication area as a paper currency crown word number area.
9. The method of claim 1, wherein prior to edge detecting the gray scale image of the banknote image, the method further comprises:
and performing inclination correction processing, orientation detection processing and gray-scale processing on the banknote image.
10. A system for locating the crown sizes of banknotes, characterized in that it comprises:
the edge detection module is used for carrying out edge detection on the gray level image of the paper money image to obtain an edge image;
the stroke width changing module is used for carrying out stroke width conversion on the original image according to the image edge points of the edge image and the gradient direction to obtain a stroke width image, and the stroke width image comprises a character area;
the communication module is used for forming a communication area in the character area of the stroke width image;
and the communicated region filtering module is used for respectively screening each communicated region according to the geometric characteristics of the paper money crown word number region to determine the paper money crown word number region.
11. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the method of any of claims 1 to 9 when executing the computer program.
12. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program for executing the method of any one of claims 1 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010521386.6A CN111709419A (en) | 2020-06-10 | 2020-06-10 | Method, system and equipment for positioning banknote serial number and readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010521386.6A CN111709419A (en) | 2020-06-10 | 2020-06-10 | Method, system and equipment for positioning banknote serial number and readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111709419A true CN111709419A (en) | 2020-09-25 |
Family
ID=72539033
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010521386.6A Pending CN111709419A (en) | 2020-06-10 | 2020-06-10 | Method, system and equipment for positioning banknote serial number and readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111709419A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111046866A (en) * | 2019-12-13 | 2020-04-21 | 哈尔滨工程大学 | Method for detecting RMB crown word number region by combining CTPN and SVM |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103942797A (en) * | 2014-04-24 | 2014-07-23 | 中国科学院信息工程研究所 | Scene image text detection method and system based on histogram and super-pixels |
| WO2016086877A1 (en) * | 2014-12-03 | 2016-06-09 | 夏普株式会社 | Text detection method and device |
| WO2017016240A1 (en) * | 2015-07-24 | 2017-02-02 | 广州广电运通金融电子股份有限公司 | Banknote serial number identification method |
| CN106446920A (en) * | 2016-09-05 | 2017-02-22 | 电子科技大学 | Stroke width transformation method based on gradient amplitude constraint |
| CN107464335A (en) * | 2017-08-03 | 2017-12-12 | 恒银金融科技股份有限公司 | A kind of paper money number localization method |
-
2020
- 2020-06-10 CN CN202010521386.6A patent/CN111709419A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103942797A (en) * | 2014-04-24 | 2014-07-23 | 中国科学院信息工程研究所 | Scene image text detection method and system based on histogram and super-pixels |
| WO2016086877A1 (en) * | 2014-12-03 | 2016-06-09 | 夏普株式会社 | Text detection method and device |
| WO2017016240A1 (en) * | 2015-07-24 | 2017-02-02 | 广州广电运通金融电子股份有限公司 | Banknote serial number identification method |
| CN106446920A (en) * | 2016-09-05 | 2017-02-22 | 电子科技大学 | Stroke width transformation method based on gradient amplitude constraint |
| CN107464335A (en) * | 2017-08-03 | 2017-12-12 | 恒银金融科技股份有限公司 | A kind of paper money number localization method |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111046866A (en) * | 2019-12-13 | 2020-04-21 | 哈尔滨工程大学 | Method for detecting RMB crown word number region by combining CTPN and SVM |
| CN111046866B (en) * | 2019-12-13 | 2023-04-18 | 哈尔滨工程大学 | Method for detecting RMB crown word number region by combining CTPN and SVM |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20190188528A1 (en) | Text detection method and apparatus, and storage medium | |
| WO2015171518A1 (en) | Method and apparatus of extracting particular information from standard card | |
| WO2020082731A1 (en) | Electronic device, credential recognition method and storage medium | |
| CN111259878A (en) | Method and equipment for detecting text | |
| CN107545223B (en) | Image recognition method and electronic equipment | |
| CN112906695B (en) | Form recognition method adapting to multi-class OCR recognition interface and related equipment | |
| CN112200117A (en) | Form identification method and device | |
| CN109492642A (en) | Licence plate recognition method, device, computer equipment and storage medium | |
| CN109447117A (en) | The double-deck licence plate recognition method, device, computer equipment and storage medium | |
| JP2022550195A (en) | Text recognition method, device, equipment, storage medium and computer program | |
| CN112418165A (en) | Small-size target detection method and device based on improved cascade neural network | |
| CN114529773A (en) | Form identification method, system, terminal and medium based on structural unit | |
| CN113609984A (en) | Pointer instrument reading identification method and device and electronic equipment | |
| CN111709419A (en) | Method, system and equipment for positioning banknote serial number and readable storage medium | |
| CN112507938A (en) | Geometric feature calculation method, geometric feature recognition method and geometric feature recognition device for text primitives | |
| CN111062262B (en) | Invoice recognition method and invoice recognition device | |
| CN110674811B (en) | Image recognition method and device | |
| CN106056575B (en) | A kind of image matching method based on like physical property proposed algorithm | |
| CN110929738A (en) | Certificate card edge detection method, device, equipment and readable storage medium | |
| CN113378865B (en) | Image pyramid matching method and device | |
| CN115527023A (en) | Image detection method, image detection device, electronic equipment and storage medium | |
| CN114495132A (en) | Character recognition method, device, equipment and storage medium | |
| CN104112135B (en) | Text image extraction element and method | |
| CN111599080A (en) | Spliced paper money detection method and device, financial machine tool equipment and storage medium | |
| CN113850238B (en) | Document detection method and device, electronic equipment and storage medium |
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 |