WO2016037523A1 - 一种基于清分机积灰条件下的钞票识别方法及清分机 - Google Patents

一种基于清分机积灰条件下的钞票识别方法及清分机 Download PDF

Info

Publication number
WO2016037523A1
WO2016037523A1 PCT/CN2015/087901 CN2015087901W WO2016037523A1 WO 2016037523 A1 WO2016037523 A1 WO 2016037523A1 CN 2015087901 W CN2015087901 W CN 2015087901W WO 2016037523 A1 WO2016037523 A1 WO 2016037523A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
spectrum image
banknote
positioning
edges
Prior art date
Application number
PCT/CN2015/087901
Other languages
English (en)
French (fr)
Inventor
王洋
梁添才
Original Assignee
广州广电运通金融电子股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 广州广电运通金融电子股份有限公司 filed Critical 广州广电运通金融电子股份有限公司
Priority to US15/328,814 priority Critical patent/US9928677B2/en
Priority to EP15840300.6A priority patent/EP3193313A4/en
Priority to RU2017106097A priority patent/RU2643493C1/ru
Publication of WO2016037523A1 publication Critical patent/WO2016037523A1/zh
Priority to ZA2017/01209A priority patent/ZA201701209B/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/13—Edge detection
    • G—PHYSICS
    • G07—CHECKING-DEVICES
    • G07D—HANDLING OF COINS OR VALUABLE PAPERS, e.g. TESTING, SORTING BY DENOMINATIONS, COUNTING, DISPENSING, CHANGING OR DEPOSITING
    • G07D11/00—Devices accepting coins; Devices accepting, dispensing, sorting or counting valuable papers
    • G07D11/10—Mechanical details
    • G07D11/14—Inlet or outlet ports
    • G—PHYSICS
    • G07—CHECKING-DEVICES
    • G07D—HANDLING OF COINS OR VALUABLE PAPERS, e.g. TESTING, SORTING BY DENOMINATIONS, COUNTING, DISPENSING, CHANGING OR DEPOSITING
    • G07D11/00—Devices accepting coins; Devices accepting, dispensing, sorting or counting valuable papers
    • G07D11/50—Sorting or counting valuable papers
    • G—PHYSICS
    • G07—CHECKING-DEVICES
    • G07D—HANDLING OF COINS OR VALUABLE PAPERS, e.g. TESTING, SORTING BY DENOMINATIONS, COUNTING, DISPENSING, CHANGING OR DEPOSITING
    • G07D13/00—Handling of coins or of valuable papers, characterised by a combination of mechanisms not covered by a single one of groups G07D1/00 - G07D11/00
    • 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
    • G07D7/06—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 wave or particle radiation
    • G07D7/12—Visible light, infrared or ultraviolet radiation
    • G07D7/1205—Testing spectral properties
    • 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
    • G07D7/2008—Testing patterns thereon using pre-processing, e.g. de-blurring, averaging, normalisation or rotation
    • 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
    • G—PHYSICS
    • G07—CHECKING-DEVICES
    • G07F—COIN-FREED OR LIKE APPARATUS
    • G07F7/00—Mechanisms actuated by objects other than coins to free or to actuate vending, hiring, coin or paper currency dispensing or refunding apparatus
    • G07F7/04—Mechanisms actuated by objects other than coins to free or to actuate vending, hiring, coin or paper currency dispensing or refunding apparatus by paper currency
    • 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
    • G07D7/16—Testing the dimensions

Definitions

  • Embodiments of the present invention relate to the field of banknote identification, and in particular, to a banknote identification method and a sorting machine based on a dust collecting condition of a sorting machine.
  • the clearing machine is such a financial machine. It uses a computer and pattern recognition technology to realize the functional technologies such as banknote counterfeiting and banknote multi-channel transmission.
  • the clearing machine sorts the banknotes at high speed during the working process, and it is easy to generate friction when it contacts the banknotes during the running process.
  • the ink on the surface of the banknotes and the adhesives in the process of use will fall off with the mechanical movement.
  • the frequency of use of the machine is high and it is not cleaned in time, the ink dust will accumulate on both sides of the acquisition module, causing the image signal collected by the acquisition module to be abnormal, which leads to low detection rate and low recognition accuracy of the sorting machine.
  • the existing sorting machine can only solve the above problems by manual cleaning, but this method is cumbersome to operate, and the user does not know the proper cleaning time, thereby greatly reducing the user experience.
  • the invention provides a banknote recognition method and a sorting machine based on the ash accumulating condition of the sorting machine, and the effective region boundary judgment is performed by using the gradation difference between the foreground and the background of the sensor image signal, and the signal characteristics of the plurality of sensors are comprehensively adopted. Modify the detection direction and the secondary scan search edge, and finally weight The way of newly locating the effective boundary of the image area can greatly improve the detection rate and recognition accuracy of the sorting machine.
  • step S2 positioning the four edges of the reflected spectrum image and determining whether the positioning is successful, if yes, obtaining a positioning image and performing steps S3 and S4, otherwise performing step S5;
  • step S4 determining whether the forward image of the reflected spectrum image is normal, if step S7 is performed, otherwise step S5 is performed;
  • the step S2 includes:
  • step S21 includes:
  • the four edges include a left edge, a right edge, an upper edge, and a lower edge;
  • the step S22 includes:
  • pixgray(i,j) is the gray level value of the gray line position
  • notegray(i,j) is the foreground gray value of the banknote
  • backgray(i,j) is the background gray value of the banknote
  • Threshold is the edge detection threshold
  • the step S22 includes:
  • pixgray(i,j) is the gray value of the gray position
  • notegray(i,j) is the foreground gray value of the banknote
  • backgray(i,j) is the background gray value of the banknote
  • Threshold is the edge detection threshold
  • the step S4 includes:
  • sum(j) is accumulated, and sum(j) is satisfied (0 ⁇ j ⁇ 1/5W);
  • step S5 If SUM>T 1 , it is determined that the forward image is an abnormal edge detection image and step S5 is performed, otherwise step S7 is performed;
  • the step S7 includes:
  • the banknotes are identified by face value, face, authenticity identification and clear function.
  • An acquisition module configured to collect a reflection spectrum image and a transmission spectrum image of the banknote
  • a positioning determining module configured to locate four edges of the reflected spectrum image and determine whether the positioning is successful
  • a first rotation mapping module configured to perform angular rotation mapping on the positioning image to obtain a forward image of the reflected spectrum image
  • a second determining module configured to determine whether a forward image of the reflected spectrum image is normal
  • a positioning module for positioning four edges of the transmission spectrum image
  • a second rotation mapping module configured to map four edges of the transmission spectrum image to the reflection spectrum image and perform angular rotation mapping to obtain a forward image of the reflection spectrum image
  • An identification module for identifying the banknote
  • the clearing machine comprises: a depositing and dispensing port, a cash dispensing port, a cash withdrawal port, a conveying track and an identification module, wherein the identification module comprises: two sets of oppositely disposed CIS image sensors, a transmission light source plate, a storage module, Detection module and display module;
  • Two sets of the CIS image sensors are respectively disposed on two sides;
  • Two sets of the transmissive light source plates are respectively disposed on two sides;
  • the CIS image sensor is configured to generate and receive a reflected spectrum image
  • the CIS image sensor and the transmissive light source panel cooperate to generate and receive the transmission spectrum image
  • the storage module is configured to store the reflected spectrum image and the transmission spectrum image.
  • the effective region boundary is judged by using the grayscale difference between the foreground and the background of the sensor image signal, and the signal characteristics of various sensors, the detection direction and the secondary scan search edge are comprehensively used. Finally, in the manner of repositioning the effective boundary of the image region, the banknote recognition method and the sorting machine based on the dust collecting condition of the sorting machine can greatly improve the detection rate and the recognition accuracy of the sorting machine.
  • FIG. 1 is a flow chart of a first embodiment of a banknote identification method based on a dust collecting condition of a sorting machine according to the present invention
  • FIG. 2 is a flow chart of a second embodiment of a banknote identification method based on the ashing condition of a sorting machine according to the present invention
  • FIG. 3 is a reflection spectrum image of an embodiment of a banknote identification method based on the ashing condition of a sorter according to the present invention
  • FIG. 4 is a transmission spectrum image of an embodiment of a banknote recognition method based on a ashing condition of a sorter according to the present invention
  • FIG. 5 is a schematic view showing the positioning of a reflection spectrum image in a second embodiment of the banknote identification method based on the ashing condition of the sorter;
  • FIG. 6 is a schematic diagram of the reflection spectrum image positioning satisfying the discrimination condition 1 in the second embodiment of the banknote recognition method based on the ash deposition condition of the present invention
  • FIG. 7 is a schematic diagram of a forward image after the reflection spectral image is successfully positioned in the second embodiment of the banknote recognition method according to the ashing condition of the sorter according to the present invention
  • FIG. 8 is a schematic diagram of the reflection spectrum image positioning satisfying the discrimination condition 2 in the second embodiment of the banknote recognition method according to the ashing condition of the sorter;
  • FIG. 9 is a schematic diagram showing the failure of the reflection spectrum image localization in the second embodiment of the banknote identification method based on the ashing condition of the sorter;
  • FIG. 10 is a schematic diagram of positioning a reflection spectrum image by positioning a transmission spectrum image in a second embodiment of the banknote recognition method based on the ash deposition condition of the present invention
  • FIG. 11 is a schematic diagram of a forward image after positioning a reflection spectrum image by positioning a transmission spectrum image in a second embodiment of the banknote recognition method according to the ashing condition of the sorter;
  • Figure 12 is a schematic structural view of a first embodiment of the sorting machine of the present invention.
  • Figure 13 is a schematic structural view of a second embodiment of the sorting machine of the present invention.
  • Figure 14 is a schematic view showing the structure of an identification module in the second embodiment of the sorting machine of the present invention.
  • the invention provides a banknote recognition method and a sorting machine based on the ash accumulating condition of the sorting machine, and the effective region boundary judgment is performed by using the gradation difference between the foreground and the background of the sensor image signal, and the signal characteristics of the plurality of sensors are comprehensively adopted. Modifying the detection direction and the secondary scan search edge, and finally repositioning the effective boundary of the image area, can greatly improve the detection rate and recognition accuracy of the sorting machine.
  • the method of the embodiment of the present invention can be used not only for detecting banknotes, but also for detecting sheet-like valuable magnetic files such as checks, and the apparatus of the embodiment of the present invention can be applied to both ATM machines and applications.
  • the bill processing apparatus such as the sorting machine
  • the method of the embodiment of the present invention will be described below by taking the sorting machine as an example.
  • the clearing machine is only taken as an example, it should not be construed as limiting the method of the present invention.
  • a first embodiment of a banknote identification method based on a ash-growth condition in an embodiment of the present invention includes:
  • the spectrum collected by the sensor in the clearing machine includes white light spectrum signal, reflected spectrum signal, transmission spectrum signal, ultraviolet signal, magnetic signal and thickness signal.
  • the invention mainly uses the reflection spectrum image and the transmission spectrum image of the banknote as the target image to detect and recognize the banknote, so that the reflection spectrum image and the transmission spectrum image of the banknote can be collected first.
  • step S2 positioning the four edges of the reflected spectrum image and determining whether the positioning is successful, if yes, obtaining a positioning image and performing steps S3 and S4, otherwise performing step S5;
  • the four edges of the reflection spectrum image may be positioned to determine the image area of the banknote, and steps S3 and S4 are performed after the positioning is determined to be successful, and step S5 is performed after the positioning failure is determined.
  • step S4 determining whether the forward image of the reflected spectrum image is normal, if step S7 is performed, otherwise step S5 is performed;
  • the positive of the reflected spectrum image can be further determined. Whether the image is normal, that is, whether the spectral image contained in the forward image described above is complete or out of range, if step S7 is performed, otherwise step S5 is performed.
  • step S6 and S7 are performed, otherwise step S8 is performed.
  • the transmission spectrum image is successfully positioned, and the four edges of the transmission spectrum image can be mapped to the reflection spectrum image and angularly rotated to obtain a forward image of the reflection spectrum image.
  • the identification of the banknote can be performed.
  • the invention is based on the banknote identification method under the ash accumulation condition of the sorting machine and the sorting machine can greatly improve the detection rate and the recognition accuracy of the sorting machine.
  • the first embodiment of the banknote identification method based on the ash-growth condition of the present invention is briefly described above.
  • the second embodiment of the banknote identification method based on the ash-growth condition of the present invention is described in detail below. 2.
  • the second embodiment of the banknote identification method based on the dust collecting condition of the sorting machine in the embodiment of the present invention comprises:
  • the spectrum collected by the sensor in the clearing machine includes white light spectrum signal, reflected spectrum signal, transmission spectrum signal, ultraviolet signal, magnetic signal and thickness signal.
  • the invention mainly uses the reflection spectrum image and the transmission spectrum image of the banknote as the target image to detect and recognize the banknote, so that the reflection spectrum image and the transmission spectrum image of the banknote can be collected first.
  • step 202 positioning the four edges of the reflected spectrum image and determining whether the positioning is successful, if yes, obtaining a positioning image and performing steps 203 and 204, otherwise performing step 205;
  • the four edges of the reflected spectral image can be located to determine the image area of the banknote, and steps 203 and 204 are performed after the positioning is determined to be successful, and step 205 is performed after the positioning failure is determined.
  • a boundary search is performed from the top, bottom, left and right directions to the center of the reflected spectrum image, and the four edges of the reflected spectrum image are positioned.
  • the step 202 may include: determining, by the step 2021, whether the four edges of the reflected spectrum image are successfully located, and if yes, obtaining the positioning image and performing steps 203 and 204; otherwise, performing step 205. And step 2022 performs angular rotation mapping on the positioning image to obtain a forward image of the reflected spectrum image.
  • the step 2021 may specifically include: the four edges include a left edge, a right edge, an upper edge, and a lower edge; searching from the left side of the reflected spectrum image, if the pixel point is satisfied
  • the search range of the above-mentioned reflection spectrum image is only 1/2 of the width of the reflection spectrum image. , not limited here.
  • step 2021 may specifically include: if the image composed of the four edges is satisfied
  • step 2021 may further include: if the image composed of the four edges is satisfied
  • pixgray(i, j) is the gray value of the gray position
  • notegray (i, j) ) is the foreground gray value of the banknote
  • backgray (i, j) is the background gray value of the banknote
  • Threshold is the edge detection threshold
  • step 204 determining whether the forward image of the reflected spectrum image is normal, if step 207 is performed, otherwise step 205 is performed;
  • step 207 After obtaining the forward image of the reflected spectrum image, it may be further determined whether the forward image of the reflected spectrum image is normal, that is, whether the spectral image included in the forward image is complete or out of range, if step 207 is performed, otherwise steps are performed. 205.
  • the edge of the normal banknote is successfully positioned, the entire spectrum image of the foreground will be extracted completely. If the image captured by the acquisition module covers a large amount of dust, which causes the edge to be searched, the edge of the dust is positioned as the left or right border of the banknote, and the extracted spectrum is extracted. Part of the image is the background image and part of the banknote image foreground. It needs to be discriminated by the discriminant formula to reduce the false recognition rate.
  • the determining whether the forward image of the reflected spectrum image is normal may specifically include: if the pixel in the forward image of the reflected spectral image satisfies
  • sum(j) is accumulated, and sum(j) is satisfied (0 ⁇ j ⁇ 1/5W);
  • the four edges of the transmission spectrum image can be located, and the positioning of the reflection spectrum image can be achieved by positioning the transmission spectrum image. If the transmission spectrum image is successfully positioned, steps 206 and 207 are performed, otherwise step 208 is performed.
  • the present invention mainly uses the transmission spectrum image to perform reverse search, effectively Avoid image interference caused by device dust and banknotes.
  • the point-to-point information mapping is used to map the boundary points searched by the transmission spectrum to the reflection spectrum image.
  • the search is performed from the left and right sides of the whole image to the two sides, and the upper and lower search remains unchanged.
  • a forward spectrum of the reflected spectrum is obtained by mapping and angular rotation.
  • the transmission spectrum image is successfully positioned, and the four edges of the transmission spectrum image can be mapped to the reflection spectrum image and angularly rotated to obtain a forward image of the reflection spectrum image.
  • the identification of the banknote can be performed.
  • the above identification of the banknote may specifically include: performing banknote denomination, face-to-face, authenticity identification, and clearing function identification on the banknote.
  • the invention is based on the banknote identification method under the ash accumulation condition of the sorting machine and the sorting machine can greatly improve the detection rate and the recognition accuracy of the sorting machine.
  • the second embodiment of the banknote identification method based on the ashing condition of the sorting machine of the present invention is briefly described above.
  • the first embodiment of the sorting machine of the present invention will be described in detail below. Referring to FIG. 12, the clearing machine in the embodiment of the present invention is described.
  • the first embodiment includes:
  • the collecting module 1201 is configured to collect a reflection spectrum image and a transmission spectrum image of the banknote
  • the positioning determining module 1202 is configured to locate four edges of the reflected spectrum image and determine whether the positioning is successful;
  • a first rotation mapping module 1203, configured to perform angular rotation mapping on the positioning image to obtain a forward image of the reflected spectrum image
  • the second determining module 1204 is configured to determine whether the forward image of the reflected spectrum image is normal
  • a positioning module 1205, configured to locate four edges of the transmission spectrum image
  • a second rotation mapping module 1206, configured to map four edges of the transmission spectrum image to the reflection spectrum image and perform angular rotation mapping to obtain a forward image of the reflection spectrum image;
  • An identification module 1207 configured to identify the banknote
  • the present invention is also based on the sorting machine product.
  • the features of the first embodiment and the second embodiment of the banknote recognition method under gray conditions are not described here.
  • the second embodiment of the sorting machine includes: The port 131, the cash withdrawal port 132, the withdrawal slot 133, the transport track 134 and the identification module 135, wherein the identification module 135 comprises: two sets of oppositely disposed CIS image sensors 1351 and a transmission light source panel 1352, a storage module, a detection module and a display module ;
  • Two sets of the CIS image sensors 1351 are respectively disposed on two sides;
  • Two sets of the transmissive light source plates 1352 are respectively disposed on two sides;
  • the CIS image sensor 1351 is configured to generate and receive a reflected spectrum image
  • the CIS image sensor 1351 and the transmissive light source panel 1352 cooperate to generate and receive the transmission spectrum image
  • the storage module is configured to store the reflected spectrum image and the transmission spectrum image.
  • FIG. 13 is a schematic structural diagram of the sorting machine of the present invention.
  • the workflow is summarized as follows: the banknotes are driven by the mechanical device to drive the banknotes into the identification module 135, and the image is scanned and recognized in the identification module 135.
  • the spectral image is sent to the memory, and the image in the memory is detected and recognized by the recognition algorithm, and finally the recognition result is sent to the upper computer to control the cash flow.
  • the light emitted from the LED light source array is directly directed to the surface of the banknote, and the light reflected from the surface is focused by the self-focusing rod lens array and imaged on the photosensor array. It is converted into electric charge, and the light intensity of different parts of the scanning surface is different, so the light intensity received by the sensor unit of different positions (ie, the pixels of CIS) is different.
  • the analog switch is controlled to be sequentially turned on by the shift register, and the electrical signals of the pixels are sequentially output in the form of analog signals, thereby obtaining a light reflection image signal for scanning the banknote.
  • the transmissive light source panel 1352 is mounted at a position directly opposite the CIS image sensor 1351.
  • the light source of the transmission light source panel 1352 displays light emitted through the surface of the banknote by the CIS image sensor 1351.
  • the output of the analog signal is finally generated to generate a transmission spectrum signal.
  • the whole process is completed in an instant and takes about tens of microseconds.
  • the reflected spectral image and the transmitted spectral image are received almost simultaneously, and each pixel point corresponds one-to-one.
  • the signal characteristics of the transmitted spectral image are used for secondary search and detection, and mapped to the reflected spectral image. Can effectively solve the influence of boundary ash.
  • Each device is equipped with a two-stage CIS image sensor 1351 and a transmission light source plate 1352 for scanning the front and back images of the banknote to improve the recognition efficiency.
  • the invention is based on the banknote identification method under the ash accumulation condition of the sorting machine and the sorting machine can greatly improve the detection rate and the recognition accuracy of the sorting machine.

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Toxicology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • Theoretical Computer Science (AREA)
  • Inspection Of Paper Currency And Valuable Securities (AREA)

Abstract

一种基于清分机积灰条件下的钞票识别方法及清分机，通过利用传感器图像信号前景与背景的灰度差值进行有效区域边界判断，并综合采用多种传感器的信号特征、修改检测方向及二次扫描搜索边缘，最后重新定位图像区域的有效边界的方式，能够大大提高清分机的检测率及识别准确率。清分机包括：入钞口（131）、出钞口（132）、退钞口（133）、传送轨道（134）及识别模块（135），其中所述识别模块（135）包括：两组相对设置的CIS图像传感器（1351）及透射光源板（1352）、存储模块、检测模块及显示模块。

Description

一种基于清分机积灰条件下的钞票识别方法及清分机
本申请要求于2014年09月11日提交中国专利局、申请号为201410460813.9、发明名称为“一种基于清分机积灰条件下的钞票识别方法及清分机”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本发明实施例涉及钞票识别领域，具体涉及一种基于清分机积灰条件下的钞票识别方法及清分机。
背景技术
伴随电子及计算机技术的飞速进步及应用，出于效率和成本的要求，人们越来越倾向于用金融机具代替传统的手工纸币分拣。清分机正是这样的金融机具，它综合运用了计算机、模式识别技术，实现了纸币鉴伪、纸币多通道传送等功能技术。
清分机在工作过程中高速对钞票进行分拣，在运行过程中与钞票接触时易产生摩擦，钞票表面油墨与使用过程中的粘着物会随机械运动而脱落。当机器使用频率较高而又没有及时清理时，这种油墨粉尘会集聚在采集模块两侧，造成采集模块采集的图像信号异常，进而导致清分机的检测率低及识别准确率低。
现有的清分机一般只能通过手工清理的方式解决上述的问题，但是这种方式操作起来较为繁琐，而且用户并不知道适当的清理时间，从而大大降低了用户的使用体验。
发明内容
本发明提供了一种基于清分机积灰条件下的钞票识别方法及清分机，通过利用传感器图像信号前景与背景的灰度差值进行有效区域边界判断，并综合采用多种传感器的信号特征、修改检测方向及二次扫描搜索边缘，最后重 新定位图像区域的有效边界的方式，能够大大提高清分机的检测率及识别准确率。
本发明实施例提供的基于清分机积灰条件下的钞票识别方法，包括：
S1、采集钞票的反射光谱图像及透射光谱图像；
S2、定位所述反射光谱图像的四个边缘并判断是否定位成功，若是则得到定位图像并执行步骤S3及S4，否则执行步骤S5；
S3、对所述定位图像进行角度旋转映射，得到所述反射光谱图像的正向图像；
S4、判断所述反射光谱图像的正向图像是否正常，若是执行步骤S7，否则执行步骤S5；
S5、定位所述透射光谱图像的四个边缘，若成功则执行步骤S6及S7，否则执行步骤S8；
S6、将所述透射光谱图像的四个边缘映射至所述反射光谱图像并进行角度旋转映射，得到所述反射光谱图像的正向图像；
S7、对所述钞票进行识别；
S8、退出所述钞票。
可选的，
所述步骤S2包括：
S21、定位所述反射光谱图像的四个边缘；
S22、判断所述反射光谱图像的四个边缘是否定位成功，若是则得到定位图像并执行步骤S3及S4，否则执行步骤S5。
可选的，所述步骤S21包括：
所述四个边缘包括左边缘、右边缘、上边缘及下边缘；
从所述反射光谱图像的左侧开始搜索，若像素点满足
Figure PCTCN2015087901-appb-000001
则停止搜索，并标记所述像素点的坐标，得到一系列标记像素点坐标，将所述像素点进行直线拟合，完成对左边缘的定位；
使用对所述左边缘定位的方法对所述右边缘、所述上边缘及所述下边缘 进行定位；
[根据细则26改正16.10.2015]　
其中notegray(i,j)为所述反射光谱图像第i行、第j列像素点的灰度值，H为所述反射光谱图像的高度，W为所述反射光谱图像的宽度，Threshold为边缘检测判断阈值。
可选的，
所述步骤S22包括：
若所述四个边缘组成的图像满足
Figure PCTCN2015087901-appb-000004
则确定所述反射光谱图像的四个边缘定位成功，得到定位图像并执行步骤S3及S4，否则执行步骤S5；
其中pixgray(i,j)为积灰线位置像素灰度值，notegray(i,j)为钞票前景灰度值，backgray(i,j)为钞票背景灰度值，Threshold为边缘检测阈值。
可选的，
所述步骤S22包括：
若所述四个边缘组成的图像满足
Figure PCTCN2015087901-appb-000005
则确定所述反射光谱图像的四个边缘定位成功，得到定位图像并执行步骤S3及S4，否则执行步骤S5；
其中pixgray(i,j)为积灰位置像素灰度值，notegray(i,j)为钞票前景灰度值，backgray(i,j)为钞票背景灰度值，Threshold为边缘检测阈值。
可选的，
所述步骤S4包括：
若所述反射光谱图像的正向图像中的像素点满足
notegray(i,W-j)-notegray(i,j)>Threshold(0<i<H,0<j<1/5W)，
或notegray(i,j)-notegray(i,W-j)>Threshold(0<i<H,0<j<1/5W)；
则sum(j)累加，sum(j)满足(0<j<1/5W)；
若sum(j)>T，则所述正向图像为积灰图像，统计变量SUM累加；
若SUM>T1，则判定所述正向图像为异常检边图像并执行步骤S5，否则执行步骤S7；
[根据细则26改正16.10.2015]　
其中notegray(i,j)为反射光谱图像第i行、第j列像素灰度值，H为反射光谱图像高度，W为反射光谱图像宽度，Threshold为设定阈值，T为单列积灰点数阈值，T1为积灰列数阈值。
可选的，
所述步骤S7包括：
对所述钞票进行钞票面值、面向、真伪鉴别及清分功能识别。
本发明实施例提供的清分机，包括：
采集模块，用于采集钞票的反射光谱图像及透射光谱图像；
定位判断模块，用于定位所述反射光谱图像的四个边缘并判断是否定位成功；
第一旋转映射模块，用于对所述定位图像进行角度旋转映射，得到所述反射光谱图像的正向图像；
第二判断模块，用于判断所述反射光谱图像的正向图像是否正常；
定位模块，用于定位所述透射光谱图像的四个边缘；
第二旋转映射模块，用于将所述透射光谱图像的四个边缘映射至所述反射光谱图像并进行角度旋转映射，得到所述反射光谱图像的正向图像；
识别模块，用于对所述钞票进行识别；
退出模块，用于退出所述钞票。
本发明实施例提供的清分机，包括：入钞口、出钞口、退钞口、传送轨道及识别模块，其中识别模块包括：两组相对设置的CIS图像传感器及透射光源板、存储模块、检测模块及显示模块；
两组所述CIS图像传感器分别设置在两个侧面上；
两组所述透射光源板分别设置在两个侧面上；
所述CIS图像传感器用于产生并接收反射光谱图像；
所述CIS图像传感器及所述透射光源板配合用于产生并接收所述透射光谱图像；
所述存储模块用于存储所述反射光谱图像及所述透射光谱图像。
通过利用传感器图像信号前景与背景的灰度差值进行有效区域边界判断，并综合采用多种传感器的信号特征、修改检测方向及二次扫描搜索边缘， 最后重新定位图像区域的有效边界的方式，本发明基于清分机积灰条件下的钞票识别方法及清分机能够大大提高清分机的检测率及识别准确率。
附图说明
图1为本发明基于清分机积灰条件下的钞票识别方法第一实施例流程图；
图2为本发明基于清分机积灰条件下的钞票识别方法第二实施例流程图；
图3为本发明基于清分机积灰条件下的钞票识别方法实施例的反射光谱图像；
图4为本发明基于清分机积灰条件下的钞票识别方法实施例的透射光谱图像；
图5为本发明基于清分机积灰条件下的钞票识别方法第二实施例中反射光谱图像的定位示意图；
图6为本发明基于清分机积灰条件下的钞票识别方法第二实施例中反射光谱图像定位满足判别条件1的示意图；
图7为本发明基于清分机积灰条件下的钞票识别方法第二实施例中反射光谱图像定位成功后的正向图像的示意图；
图8为本发明基于清分机积灰条件下的钞票识别方法第二实施例中反射光谱图像定位满足判别条件2的示意图；
图9为本发明基于清分机积灰条件下的钞票识别方法第二实施例中反射光谱图像定位失败的示意图；
图10为本发明基于清分机积灰条件下的钞票识别方法第二实施例中通过定位透射光谱图像来定位反射光谱图像的示意图；
图11为本发明基于清分机积灰条件下的钞票识别方法第二实施例中通过定位透射光谱图像来定位反射光谱图像成功后的正向图像的示意图；
图12为本发明清分机第一实施例的结构示意图；
图13为本发明清分机第二实施例的结构示意图；
图14为本发明清分机第二实施例中识别模块的结构示意图。
具体实施方式
本发明提供了一种基于清分机积灰条件下的钞票识别方法及清分机，通过利用传感器图像信号前景与背景的灰度差值进行有效区域边界判断，并综合采用多种传感器的信号特征、修改检测方向及二次扫描搜索边缘，最后重新定位图像区域的有效边界的方式，能够大大提高清分机的检测率及识别准确率。
需要说明的是，本发明实施例的方法不但可以用于检测钞票，还可以用于检测支票等薄片类有价磁性文件，本发明实施例的装置既可应用于ATM机中，也可以应用于清分机等票据处理设备中，下面以清分机为例对本发明实施例的方法进行说明，虽然仅以清分机为例进行说明，但是不应将此作为本发明方法、的限定。
请参阅图1，本发明实施例中基于清分机积灰条件下的钞票识别方法的第一实施例包括：
S1、采集钞票的反射光谱图像及透射光谱图像；
清分机内传感器采集的光谱包括白光光谱信号、反射光谱信号、透射光谱信号、紫外信号、磁信号以及厚度信号等。本发明主要通过钞票的反射光谱图像、透射光谱图像为目标图像，对钞票进行检测与识别处理，因此可以首先采集钞票的反射光谱图像及透射光谱图像。
S2、定位反射光谱图像的四个边缘并判断是否定位成功，若是则得到定位图像并执行步骤S3及S4，否则执行步骤S5；
得到钞票的反射光谱图像之后，可以定位反射光谱图像的四个边缘，以确定钞票的图像区域，并在确定定位成功后执行步骤S3及S4，在确定定位失败后执行步骤S5。
S3、对定位图像进行角度旋转映射，得到反射光谱图像的正向图像；
确定反射光谱图像的四个边缘定位成功，可以对定位过程中得到的定位图像进行角度旋转映射，得到上述反射光谱图像的正向图像。
S4、判断反射光谱图像的正向图像是否正常，若是执行步骤S7，否则执行步骤S5；
得到反射光谱图像的正向图像之后，可以进一步判断反射光谱图像的正 向图像是否正常，亦即判断上述的正向图像中包含的光谱图像是否完整或超范围，若是执行步骤S7，否则执行步骤S5。
S5、定位透射光谱图像的四个边缘，若成功则执行步骤S6及S7，否则执行步骤S8；
在反射光谱图像定位失败或得到的正向图像异常时，可以定位透射光谱图像的四个边缘，并通过透射光谱图像的定位来实现对反射光谱图像的定位。若透射光谱图像定位成功则执行步骤S6及S7，否则执行步骤S8。
S6、将透射光谱图像的四个边缘映射至反射光谱图像并进行角度旋转映射，得到反射光谱图像的正向图像；
确定透射光谱图像定位成功，可以将透射光谱图像的四个边缘映射至反射光谱图像并进行角度旋转映射，得到反射光谱图像的正向图像。
S7、对钞票进行识别；
通过步骤S6得到反射光谱图像的正向图像或确定反射光谱图像的正向图像正常时，可以进行对钞票的识别。
S8、退出钞票。
确定透射光谱图像定位失败，退出钞票。
通过利用传感器图像信号前景与背景的灰度差值进行有效区域边界判断，并综合采用多种传感器的信号特征、修改检测方向及二次扫描搜索边缘，最后重新定位图像区域的有效边界的方式，本发明基于清分机积灰条件下的钞票识别方法及清分机能够大大提高清分机的检测率及识别准确率。
上面简单介绍了本发明基于清分机积灰条件下的钞票识别方法的第一实施例，下面对本发明基于清分机积灰条件下的钞票识别方法的第二实施例进行详细的描述，请参阅图2，本发明实施例中基于清分机积灰条件下的钞票识别方法第二实施例包括：
201、采集钞票的反射光谱图像及透射光谱图像；
清分机内传感器采集的光谱包括白光光谱信号、反射光谱信号、透射光谱信号、紫外信号、磁信号以及厚度信号等。本发明主要通过钞票的反射光谱图像、透射光谱图像为目标图像，对钞票进行检测与识别处理，因此可以首先采集钞票的反射光谱图像及透射光谱图像。
202、定位反射光谱图像的四个边缘并判断是否定位成功，若是则得到定位图像并执行步骤203及204，否则执行步骤205；
得到钞票的反射光谱图像之后，可以定位反射光谱图像的四个边缘，以确定钞票的图像区域，并在确定定位成功后执行步骤203及204，在确定定位失败后执行步骤205。请参阅图5，分别从上下、左右四个方向向反射光谱图像中心进行边界搜索，将反射光谱图像的四个边缘定位出来。
步骤202具体可以包括：步骤2021判断所述反射光谱图像的四个边缘是否定位成功，若是则得到定位图像并执行步骤203及204，否则执行步骤205。及步骤2022对定位图像进行角度旋转映射，得到反射光谱图像的正向图像。
其中步骤2021具体可以包括：所述四个边缘包括左边缘、右边缘、上边缘及下边缘；从所述反射光谱图像的左侧开始搜索，若像素点满足
Figure PCTCN2015087901-appb-000007
[根据细则26改正16.10.2015]　
则停止搜索，并标记所述像素点的坐标，得到一系列标记像素点坐标，将所述像素点进行直线拟合，完成对左边缘的定位；使用对所述左边缘定位的方法对所述右边缘、所述上边缘及所述下边缘进行定位；其中notegray(i,j)为所述反射光谱图像第i行、第j列像素点的灰度值，H为所述反射光谱图像的高度，W为所述反射光谱图像的宽度，Threshold为边缘检测判断阈值。其中右边缘、上边缘及下边缘的满足条件分别为：
Figure PCTCN2015087901-appb-000010
右边缘；
Figure PCTCN2015087901-appb-000011
上边缘；
Figure PCTCN2015087901-appb-000012
下边缘。
需要说明的是，定位反射光谱图像的四个边缘时，四个边缘的定位不存在必然的先后顺序，而且为了节省搜索时间，上述反射光谱图像的搜索范围仅为反射光谱图像宽度的1/2，在此处不做限定。
请参阅图6及图7，其中步骤2021具体可以包括：若所述四个边缘组成的图像满足
判别条件1，
Figure PCTCN2015087901-appb-000013
则确定所述反射光谱图像的四个边缘定位成功，，亦即钞票的前景和背景的灰度符合条件，得到定位图像并执行步骤203及204，否则执行步骤205；其中pixgray(i,j)为积灰线位置像素灰度值，notegray(i,j)为钞票前景灰度值，backgray(i,j)为钞票背景灰度值，Threshold为边缘检测阈值。
请参阅图8及图9，步骤2021具体还可以包括：若所述四个边缘组成的图像满足
判别条件2，
Figure PCTCN2015087901-appb-000014
则确定所述反射光谱图像的四个边缘定位成功，得到定位图像并执行步骤203及204，否则执行步骤205；其中pixgray(i,j)为积灰位置像素灰度值，notegray(i,j)为钞票前景灰度值，backgray(i,j)为钞票背景灰度值，Threshold为边缘检测阈值。
204、判断反射光谱图像的正向图像是否正常，若是执行步骤207，否则执行步骤205；
得到反射光谱图像的正向图像之后，可以进一步判断反射光谱图像的正向图像是否正常，亦即判断上述的正向图像中包含的光谱图像是否完整或超范围，若是执行步骤207，否则执行步骤205。
正常钞票的边缘定位成功，会将前景整幅光谱图像提取完整，倘若采集模块采集的图像覆盖大量积灰，导致搜索边缘时，误将积灰边缘定位成钞票的左或右边界，提取的光谱图像一部分为背景图像，一部分为钞票图像前景。需通过判别公式对其进行判别，降低误识别率。
反射光谱图像的正向图像是否正常的判断具体可以包括：若所述反射光谱图像的正向图像中的像素点满足
notegray(i,W-j)-notegray(i,j)>Threshold(0<i<H,0<j<1/5W)，
或notegray(i,j)-notegray(i,W-j)>Threshold(0<i<H,0<j<1/5W)；
则sum(j)累加，sum(j)满足(0<j<1/5W)；
若sum(j)>T，则所述正向图像为积灰图像，统计变量SUM累加；
[根据细则26改正16.10.2015]　
若SUM>T1，则判定所述正向图像为异常检边图像并执行步骤S5，否则执行步骤S7；其中notegray(i,j)为反射光谱图像第i行、第j列像素灰度值，H为反射光谱图像高度，W为反射光谱图像宽度，Threshold为设定阈值，T为单列积灰点数阈值(若钞票为人民币，则T＝3/4H)，T1为积灰列数阈值(若钞票为人民币，则T1＝3)。
205、定位透射光谱图像的四个边缘，若成功则执行步骤206及207，否则执行步骤208；
在反射光谱图像定位失败或得到的正向图像异常时，可以定位透射光谱图像的四个边缘，并通过透射光谱图像的定位来实现对反射光谱图像的定位。若透射光谱图像定位成功则执行步骤206及207，否则执行步骤208。
请参阅图10及图11，由于钞票鉴伪信息特征能够通过反射光谱图像反应在钞票的前景图像中，对边界搜索产生一定的干扰，本发明主要是利用透射光谱图像左右进行反向搜索，有效避开装置积灰及钞票自身特性引起的图像干扰。利用点对点信息映射，将透射光谱搜索到的边界点一一映射到反射光谱图像。为了节省搜索时间，分别从整幅图像的左、右各1/3处向两侧搜索，上下搜索保持不变。通过映射与角度旋转得到反射光谱正向图像。
206、将透射光谱图像的四个边缘映射至反射光谱图像并进行角度旋转映射，得到反射光谱图像的正向图像；
确定透射光谱图像定位成功，可以将透射光谱图像的四个边缘映射至反射光谱图像并进行角度旋转映射，得到反射光谱图像的正向图像。
207、对钞票进行识别；
通过步骤206得到反射光谱图像的正向图像或确定反射光谱图像的正向图像正常时，可以进行对钞票的识别。
上述对钞票的识别具体可以包括：对钞票进行钞票面值、面向、真伪鉴别及清分功能识别。
208、退出钞票。
确定透射光谱图像定位失败，退出钞票。
通过利用传感器图像信号前景与背景的灰度差值进行有效区域边界判断，并综合采用多种传感器的信号特征、修改检测方向及二次扫描搜索边缘，最后重新定位图像区域的有效边界的方式，本发明基于清分机积灰条件下的钞票识别方法及清分机能够大大提高清分机的检测率及识别准确率。
上面简单介绍了本发明基于清分机积灰条件下的钞票识别方法的第二实施例，下面对本发明清分机的第一实施例进行详细的描述，请参阅图12，本发明实施例中清分机的第一实施例包括：
采集模块1201，用于采集钞票的反射光谱图像及透射光谱图像；
定位判断模块1202，用于定位所述反射光谱图像的四个边缘并判断是否定位成功；
第一旋转映射模块1203，用于对所述定位图像进行角度旋转映射，得到所述反射光谱图像的正向图像；
第二判断模块1204，用于判断所述反射光谱图像的正向图像是否正常；
定位模块1205，用于定位所述透射光谱图像的四个边缘；
第二旋转映射模块1206，用于将所述透射光谱图像的四个边缘映射至所述反射光谱图像并进行角度旋转映射，得到所述反射光谱图像的正向图像；
识别模块1207，用于对所述钞票进行识别；
退出模块1208，用于退出所述钞票。
由于本发明清分机的第一实施例为与本发明基于清分机积灰条件下的钞票识别方法的第一实施例及第二实施例相对应的实施例，因此也具有本发明基于清分机积灰条件下的钞票识别方法的第一实施例及第二实施例所具有的特点，在此处不再累述。
上面简单介绍了本发明清分机的第一实施例，下面对本发明清分机的第二实施例进行详细的描述，请参阅图13，本发明实施例中清分机的第二实施例包括：入钞口131、出钞口132、退钞口133、传送轨道134及识别模块135，其中识别模块135包括：两组相对设置的CIS图像传感器1351及透射光源板1352、存储模块、检测模块及显示模块；
两组所述CIS图像传感器1351分别设置在两个侧面上；
两组所述透射光源板1352分别设置在两个侧面上；
所述CIS图像传感器1351用于产生并接收反射光谱图像；
所述CIS图像传感器1351及所述透射光源板1352配合用于产生并接收所述透射光谱图像；
所述存储模块用于存储所述反射光谱图像及所述透射光谱图像。
请参阅图13，为本发明的清分机的结构示意图，工作流程概括为：钞票在入钞口通过机械装置的传动下带动钞票进入识别模块135，在识别模块135中进行图像扫描并将获取的光谱图像送至存储器，通过识别算法对存储器中的图像进行检测识别，最终将识别结果送至上位机控制钞口出钞。
请参阅图14，当有钞票通过CIS图像传感器1351时，其内部的LED光源阵列发出的光线直射到钞票表面，从其表面反射回的光线经自聚焦棒状透镜阵列聚焦，成像在光电传感器阵列上，被转化为电荷存储起来，扫描面不同部位的光强不同，因而不同位置传感器单元(即CIS的像素)接收到的光强不一样。达到积蓄时间后，由移位寄存器控制模拟开关依次打开，将像素的电信号以模拟信号的形式依次输出，从而得到扫描钞票的光反射图像信号。透射光源板1352，安装在CIS图像传感器1351正对面位置。当钞票图像的反射信号接收完成后，透射光源板1352的光源陈列发出光线透过钞票表面由CIS图像传感器1351对其接收，经过上述步骤最终以模拟信号输出即产生透射光谱信号。整个过程瞬间完成，耗时大约几十微秒，反射光谱图像与透射光谱图像几乎同时接收完成，各像素点一一对应，通过透射光谱图像的信号特征进行二次搜索检测，映射到反射光谱图像，能够有效解决边界积灰影响。且每个装置配有两级CIS图像传感器1351和透射光源板1352，目的在于对钞票的正反面图像进行扫描，提高识别效率。
通过利用传感器图像信号前景与背景的灰度差值进行有效区域边界判断，并综合采用多种传感器的信号特征、修改检测方向及二次扫描搜索边缘，最后重新定位图像区域的有效边界的方式，本发明基于清分机积灰条件下的钞票识别方法及清分机能够大大提高清分机的检测率及识别准确率。
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分步骤是可以通过程序来指令相关的硬件完成，其中的程序可以存储于一种计算机可读存储介质中，上述提到的存储介质可以是只读存储器，磁盘或光盘等。
以上对本发明所提供的一种基于清分机积灰条件下的钞票识别方法及清分机进行了详细介绍，对于本领域的一般技术人员，依据本发明实施例的思想，在具体实施方式及应用范围上均会有改变之处，综上所述，本说明书内容不应理解为对本发明的限制。

Claims (9)

  1. 一种基于清分机积灰条件下的钞票识别方法，其特征在于，包括：
    S1、采集钞票的反射光谱图像及透射光谱图像；
    S2、定位所述反射光谱图像的四个边缘并判断是否定位成功，若是则得到定位图像并执行步骤S3及S4，否则执行步骤S5；
    S3、对所述定位图像进行角度旋转映射，得到所述反射光谱图像的正向图像；
    S4、判断所述反射光谱图像的正向图像是否正常，若是执行步骤S7，否则执行步骤S5；
    S5、定位所述透射光谱图像的四个边缘，若成功则执行步骤S6及S7，否则执行步骤S8；
    S6、将所述透射光谱图像的四个边缘映射至所述反射光谱图像并进行角度旋转映射，得到所述反射光谱图像的正向图像；
    S7、对所述钞票进行识别；
    S8、退出所述钞票。
  2. 根据权利要求1所述的基于清分机积灰条件下的钞票识别方法，其特征在于，所述步骤S2包括：
    S21、定位所述反射光谱图像的四个边缘；
    S22、判断所述反射光谱图像的四个边缘是否定位成功，若是则得到定位图像并执行步骤S3及S4，否则执行步骤S5。
  3. [根据细则26改正16.10.2015]　
    根据权利要求2所述的基于清分机积灰条件下的钞票识别方法，其特征在于，所述步骤S21包括：
    所述四个边缘包括左边缘、右边缘、上边缘及下边缘；
    从所述反射光谱图像的左侧开始搜索，若像素点满足
    Figure PCTCN2015087901-appb-100001
    则停止搜索，并标记所述像素点的坐标，得到一系列标记像素点坐标，将所述像素点进行直线拟合，完成对左边缘的定位；
    使用对所述左边缘定位的方法对所述右边缘、所述上边缘及所述下边缘进行定位；
    其中notegray(i,j)为所述反射光谱图像第i行、第j列像素点的灰度值，H为所述反射光谱图像的高度，W为所述反射光谱图像的宽度，Threshold为边缘检测判断阈值。
  4. 根据权利要求2所述的基于清分机积灰条件下的钞票识别方法，其特征在于，所述步骤S22包括：
    若所述四个边缘组成的图像满足
    Figure PCTCN2015087901-appb-100004
    则确定所述反射光谱图像的四个边缘定位成功，得到定位图像并执行步骤S3及S4，否则执行步骤S5；
    其中pixgray(i,j)为积灰线位置像素灰度值，notegray(i,j)为钞票前景灰度值，backgray(i,j)为钞票背景灰度值，Threshold为边缘检测阈值。
  5. 根据权利要求2所述的基于清分机积灰条件下的钞票识别方法，其特征在于，所述步骤S22包括：
    若所述四个边缘组成的图像满足
    Figure PCTCN2015087901-appb-100005
    则确定所述反射光谱图像的四个边缘定位成功，得到定位图像并执行步骤S3及S4，否则执行步骤S5；
    其中pixgray(i,j)为积灰位置像素灰度值，notegray(i,j)为钞票前景灰度值，backgray(i,j)为钞票背景灰度值，Threshold为边缘检测阈值。
  6. [根据细则26改正16.10.2015]　
    根据权利要求1所述的基于清分机积灰条件下的钞票识别方法，其特征在于，所述步骤S4包括：
    若所述反射光谱图像的正向图像中的像素点满足
    notegray(i,W-j)-notegray(i,j)>Threshold(0<i<H,0<j<1/5W)，
    或notegray(i,j)-notegray(i,W-j)>Threshold(0<i<H,0<j<1/5W)；
    则sum(j)累加，sum(j)满足(0<j<1/5W)；
    若sum(j)>T，则所述正向图像为积灰图像，统计变量SUM累加；
    若SUM>T1，则判定所述正向图像为异常检边图像并执行步骤S5，否则执行步骤S7；
    其中notegray(i,j)为反射光谱图像第i行、第j列像素灰度值，H为反射光谱图像高度，W为反射光谱图像宽度，Threshold为设定阈值，T为单列积灰点数阈值，T1为积灰列数阈值。
  7. 根据权利要求1所述的基于清分机积灰条件下的钞票识别方法，其特征在于，所述步骤S7包括：
    对所述钞票进行钞票面值、面向、真伪鉴别及清分功能识别。
  8. 一种清分机，其特征在于，包括：
    采集模块，用于采集钞票的反射光谱图像及透射光谱图像；
    定位判断模块，用于定位所述反射光谱图像的四个边缘并判断是否定位成功；
    第一旋转映射模块，用于对所述定位图像进行角度旋转映射，得到所述反射光谱图像的正向图像；
    第二判断模块，用于判断所述反射光谱图像的正向图像是否正常；
    定位模块，用于定位所述透射光谱图像的四个边缘；
    第二旋转映射模块，用于将所述透射光谱图像的四个边缘映射至所述反射光谱图像并进行角度旋转映射，得到所述反射光谱图像的正向图像；
    识别模块，用于对所述钞票进行识别；
    退出模块，用于退出所述钞票。
  9. 一种清分机，包括入钞口、出钞口、退钞口、传送轨道及识别模块，其特征在于，所述识别模块包括：两组相对设置的CIS图像传感器及透射光源板、存储模块、检测模块及显示模块；
    两组所述CIS图像传感器分别设置在两个侧面上；
    两组所述透射光源板分别设置在两个侧面上；
    所述CIS图像传感器用于产生并接收反射光谱图像；
    所述CIS图像传感器及所述透射光源板配合用于产生并接收所述透射光谱图像；
    所述存储模块用于存储所述反射光谱图像及所述透射光谱图像。
PCT/CN2015/087901 2014-09-11 2015-08-24 一种基于清分机积灰条件下的钞票识别方法及清分机 WO2016037523A1 (zh)

Priority Applications (4)

Application Number Priority Date Filing Date Title
US15/328,814 US9928677B2 (en) 2014-09-11 2015-08-24 Banknote recognition method based on sorter dust accumulation and sorter
EP15840300.6A EP3193313A4 (en) 2014-09-11 2015-08-24 Banknote recognition method based on sorter dust accumulation and sorter
RU2017106097A RU2643493C1 (ru) 2014-09-11 2015-08-24 Способ распознавания банкнот на основе накопления пыли в сортировщике и сортировщик
ZA2017/01209A ZA201701209B (en) 2014-09-11 2017-02-17 Banknote recognition method based on sorter dust accumulation and sorter

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201410460813.9 2014-09-11
CN201410460813.9A CN104200566B (zh) 2014-09-11 2014-09-11 一种基于清分机积灰条件下的钞票识别方法及清分机

Publications (1)

Publication Number Publication Date
WO2016037523A1 true WO2016037523A1 (zh) 2016-03-17

Family

ID=52085851

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2015/087901 WO2016037523A1 (zh) 2014-09-11 2015-08-24 一种基于清分机积灰条件下的钞票识别方法及清分机

Country Status (6)

Country Link
US (1) US9928677B2 (zh)
EP (1) EP3193313A4 (zh)
CN (1) CN104200566B (zh)
RU (1) RU2643493C1 (zh)
WO (1) WO2016037523A1 (zh)
ZA (1) ZA201701209B (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106023430A (zh) * 2016-05-06 2016-10-12 广州大学 一种硬币纸币分类整理方法及装置

Families Citing this family (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104200566B (zh) * 2014-09-11 2018-04-20 广州广电运通金融电子股份有限公司 一种基于清分机积灰条件下的钞票识别方法及清分机
CN106991752A (zh) * 2017-03-31 2017-07-28 深圳怡化电脑股份有限公司 一种验钞方法及装置
CN108961200A (zh) * 2017-05-17 2018-12-07 深圳怡化电脑股份有限公司 一种灰尘检测方法及装置
CN107369241A (zh) * 2017-07-13 2017-11-21 深圳怡化电脑股份有限公司 一种票据处理装置及方法
CN108366187B (zh) * 2018-02-12 2021-05-07 杭州创恒电子技术开发有限公司 基于cis的斜视成像装置及方法
CN111415450B (zh) * 2019-01-08 2022-03-08 深圳怡化电脑股份有限公司 票据涂改的检测方法及其检测装置、计算机存储介质
CN113283417A (zh) * 2020-12-31 2021-08-20 深圳怡化电脑股份有限公司 一种图像数据处理方法、装置、电子设备和介质
CN114018982B (zh) * 2021-10-14 2023-11-07 国网江西省电力有限公司电力科学研究院 一种空预器积灰可视化监测方法
CN114120518B (zh) * 2021-11-26 2024-02-02 深圳怡化电脑股份有限公司 纸币连张检测方法、装置、电子设备及存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4334913B2 (ja) * 2003-05-28 2009-09-30 ローレル精機株式会社 紙幣画像検出装置
CN101794476A (zh) * 2009-01-16 2010-08-04 劳雷尔机械株式会社 纸币处理机
CN103456075A (zh) * 2013-09-06 2013-12-18 广州广电运通金融电子股份有限公司 一种纸币处理方法及装置
CN103903329A (zh) * 2014-04-14 2014-07-02 华中科技大学 接触式图像传感器cis前置的金融票据鉴伪识别装置
CN104200566A (zh) * 2014-09-11 2014-12-10 广州广电运通金融电子股份有限公司 一种基于清分机积灰条件下的钞票识别方法及清分机

Family Cites Families (17)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2641275B2 (ja) * 1988-11-28 1997-08-13 株式会社日立製作所 紙葉類堆積・繰り出し装置
JP3064739B2 (ja) * 1993-04-16 2000-07-12 富士通株式会社 画像処理装置
EP1344193B1 (en) * 2000-12-21 2007-07-04 De La Rue International Limited Optical sensor device and method for spectral analysis
JP2004280188A (ja) * 2003-03-12 2004-10-07 Omron Corp 価値媒体処理装置
WO2004081887A1 (ja) 2003-03-14 2004-09-23 Fujitsu Limited 紙葉類識別方法及び紙葉類識別装置
JP4222546B2 (ja) 2003-03-31 2009-02-12 株式会社日本コンラックス 紙葉類識別装置および方法
JP2004326624A (ja) * 2003-04-25 2004-11-18 Aruze Corp 識別センサ
JP4334911B2 (ja) * 2003-05-28 2009-09-30 ローレル精機株式会社 紙幣画像検出装置
JP2007179323A (ja) * 2005-12-28 2007-07-12 Nippon Conlux Co Ltd 紙葉類識別装置および方法
GB0612856D0 (en) * 2006-06-28 2006-08-09 Rue De Int Ltd Document handling apparatus
JPWO2008096430A1 (ja) * 2007-02-08 2010-05-20 グローリー株式会社 紙幣処理装置
JP5005760B2 (ja) 2007-03-29 2012-08-22 グローリー株式会社 紙葉類識別装置および紙葉類識別方法
WO2009031242A1 (ja) * 2007-09-07 2009-03-12 Glory Ltd. 紙葉類識別装置及び紙葉類識別方法
JP2011123301A (ja) * 2009-12-10 2011-06-23 Oki Electric Industry Co Ltd 表示入力装置の汚れ検出方法、表示入力装置および自動取引装置
JP5244952B2 (ja) * 2010-12-21 2013-07-24 キヤノン・コンポーネンツ株式会社 イメージセンサユニット、及び、画像読取装置
CN102779275B (zh) * 2012-07-04 2015-06-17 广州广电运通金融电子股份有限公司 一种纸类字符识别方法及相关装置
CN103279735B (zh) * 2013-04-24 2017-07-25 广州广电运通金融电子股份有限公司 一种金融票据识别模块中积灰检测方法及系统

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4334913B2 (ja) * 2003-05-28 2009-09-30 ローレル精機株式会社 紙幣画像検出装置
CN101794476A (zh) * 2009-01-16 2010-08-04 劳雷尔机械株式会社 纸币处理机
CN103456075A (zh) * 2013-09-06 2013-12-18 广州广电运通金融电子股份有限公司 一种纸币处理方法及装置
CN103903329A (zh) * 2014-04-14 2014-07-02 华中科技大学 接触式图像传感器cis前置的金融票据鉴伪识别装置
CN104200566A (zh) * 2014-09-11 2014-12-10 广州广电运通金融电子股份有限公司 一种基于清分机积灰条件下的钞票识别方法及清分机

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106023430A (zh) * 2016-05-06 2016-10-12 广州大学 一种硬币纸币分类整理方法及装置
CN106023430B (zh) * 2016-05-06 2019-08-09 广州大学 一种硬币纸币分类整理方法及装置

Also Published As

Publication number Publication date
EP3193313A4 (en) 2017-10-04
US9928677B2 (en) 2018-03-27
CN104200566B (zh) 2018-04-20
RU2643493C1 (ru) 2018-02-01
CN104200566A (zh) 2014-12-10
US20170213409A1 (en) 2017-07-27
EP3193313A1 (en) 2017-07-19
ZA201701209B (en) 2018-05-30

Similar Documents

Publication Publication Date Title
WO2016037523A1 (zh) 一种基于清分机积灰条件下的钞票识别方法及清分机
CN105957238B (zh) 一种纸币管理方法及其系统
EP2889846B1 (en) Paper money identification method and device
CN106056751B (zh) 冠字号码的识别方法及系统
CN103279735B (zh) 一种金融票据识别模块中积灰检测方法及系统
CN104091388B (zh) 一种基于磁图像的纸币鉴伪方法及装置
CN103377509B (zh) 介质验证器和对缺损进行分类的方法
WO2015032187A1 (zh) 一种纸币处理方法及装置
WO2014075452A1 (zh) 纸币号码识别装置和识别方法
KR102007685B1 (ko) 하이브리드 위폐 감별 장치 및 시스템
CN103021069B (zh) 高速票据图像采集处理系统及其采集处理方法
CN105046808A (zh) 一种纸币多光谱高分辨率图像采集系统及采集方法
US9047723B2 (en) Defect categorization
CN101297328B (zh) 用于钞票鉴定的光传感器阵列
WO2015172304A1 (zh) 一种自动选择纸币不同特征点判断的方法
KR102070002B1 (ko) 위폐 감별 효율 향상을 위한 하이브리드 위폐 감별 장치
UA73732C2 (en) Method for determining authenticity, nominal value, and wear and tear of bank notes; device for sorting and counting bank notes
CN103390308A (zh) 基于部分图像子块的高速纸币判别系统及其判别方法
WO2013127186A1 (zh) 纸页类介质识别装置及识别方法
JPH09245216A (ja) 紙葉類識別装置
JP2015001759A (ja) 紙葉類処理装置、及び紙葉類処理方法
KR20170044502A (ko) 빛을 모아주는 집광판이 구비된 지폐 감별기

Legal Events

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

Ref document number: 15840300

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 15328814

Country of ref document: US

REEP Request for entry into the european phase

Ref document number: 2015840300

Country of ref document: EP

WWE Wipo information: entry into national phase

Ref document number: 2015840300

Country of ref document: EP

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 2017106097

Country of ref document: RU

Kind code of ref document: A