WO2014075452A1 - 纸币号码识别装置和识别方法 - Google Patents

纸币号码识别装置和识别方法 Download PDF

Info

Publication number
WO2014075452A1
WO2014075452A1 PCT/CN2013/078911 CN2013078911W WO2014075452A1 WO 2014075452 A1 WO2014075452 A1 WO 2014075452A1 CN 2013078911 W CN2013078911 W CN 2013078911W WO 2014075452 A1 WO2014075452 A1 WO 2014075452A1
Authority
WO
WIPO (PCT)
Prior art keywords
banknote
image data
image
light
light source
Prior art date
Application number
PCT/CN2013/078911
Other languages
English (en)
French (fr)
Inventor
陈柱文
刘梦涛
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 AU2013347512A priority Critical patent/AU2013347512B2/en
Priority to US14/426,972 priority patent/US9741193B2/en
Priority to EP13854512.4A priority patent/EP2922037A4/en
Publication of WO2014075452A1 publication Critical patent/WO2014075452A1/zh
Priority to ZA2015/02309A priority patent/ZA201502309B/en

Links

Classifications

    • GPHYSICS
    • G07CHECKING-DEVICES
    • G07DHANDLING OF COINS OR VALUABLE PAPERS, e.g. TESTING, SORTING BY DENOMINATIONS, COUNTING, DISPENSING, CHANGING OR DEPOSITING
    • G07D7/00Testing 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
    • G07D7/06Testing 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
    • G07D7/12Visible light, infrared or ultraviolet radiation
    • GPHYSICS
    • G07CHECKING-DEVICES
    • G07DHANDLING OF COINS OR VALUABLE PAPERS, e.g. TESTING, SORTING BY DENOMINATIONS, COUNTING, DISPENSING, CHANGING OR DEPOSITING
    • G07D7/00Testing 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
    • G07D7/004Testing 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
    • G07D7/0047Testing 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
    • GPHYSICS
    • G07CHECKING-DEVICES
    • G07DHANDLING OF COINS OR VALUABLE PAPERS, e.g. TESTING, SORTING BY DENOMINATIONS, COUNTING, DISPENSING, CHANGING OR DEPOSITING
    • G07D7/00Testing 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
    • G07D7/06Testing 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
    • G07D7/12Visible light, infrared or ultraviolet radiation
    • G07D7/121Apparatus characterised by sensor details

Definitions

  • the present application relates to the field of financial self-service devices, and more particularly to a banknote number identification device and a recognition method. Background technique
  • a known banknote number identifying device acquires an image through a CIS contact image sensor or a CCD image sensor.
  • the image sensor collects the banknote image at a high speed, converts it into image data through an analog-to-digital converter (AD), and the digital signal processor quickly processes the image data, locates the position of the banknote number, and realizes the operation by the recognition algorithm. Identification of banknote numbers.
  • the banknote number is usually obtained by transmission imaging, and only one CIS is installed on the banknote channel, and a set of white LED light sources are installed in front of the CIS, and The fixed light time is set to illuminate the passing banknotes, and the light penetrating the banknotes is photo-imaged on the CIS, and the banknote number can be obtained regardless of whether the banknote number appears on the front side of the CIS or on the back side.
  • the present application provides a new banknote number identification method and device, which can not only correctly identify the new or brand new banknote numbers, but also can The old and new cases of banknotes, the correct identification of the number of old banknotes.
  • the present application provides a banknote number identification device, comprising: a touch sensor for collecting banknote image data by means of transmission imaging; a white light LED transmitting light source plate disposed directly in front of the touch sensor, the contact sensor Collecting banknote image data to provide a transmission light source; an image acquisition processing board connected to the touch sensor through a data line, the image acquisition processing board is integrated with a field editable logic gate array and a digital signal processor, the field can be Editing the logic gate array for controlling the white LED transmission light source panel to light off and driving the CIS to acquire an image, the digital signal processor is configured to process and recognize the banknote image data; and two pairs of position sensors are disposed on the banknote transport channel from the contact type The sensor is at a predetermined distance for detecting whether a banknote enters the banknote number identification device; wherein the white light LED transmission light source plate has two light source modes, and the two light source modes are alternately collected by the contact sensor Banknote image data provides a transmitted light source .
  • the strong and weak light source modes control the white light through the field editable logic gate array
  • the conduction time of the LED transmission light source plate is formed, the conduction time is long, the strong light mode is formed, the conduction time is short, and the low light mode is formed, and the ratio of the on-time of forming the strong light mode to the on-time of forming the low-light mode is 3:1.
  • the image acquisition processing board further comprises a memory
  • the memory comprises a buffer area and a register
  • the buffer area is used for storing banknote image data collected by the touch sensor
  • the register is used to represent the banknote in the buffer area.
  • Type of image data 1 indicates strong light image data, and 0 indicates low light image data.
  • the Field _ Programmable Gate Array (FPGA) is internally provided with a clock MCLK with a period of 0.125 ms and a line counter.
  • the present application also provides a method for identifying a banknote number, comprising: Step 1 : acquiring a strong light image data and a low light image data of the banknote by a transmission imaging method and alternating illumination of the transmitted light source strong light mode and the weak light mode Step 2: performing edge detection on the low light image data, acquiring four edges of the image, and determining the target image; Step 3, processing the target image, identifying the version of the banknote, the face value, facing, and positioning the banknote number area; Step 4 Calling a pattern recognition algorithm for recognition, using a pattern recognition method based on an artificial neural network, comparing the gray level of the banknote image with a threshold value, and identifying the oldness of the banknote, and if it is an old banknote, using the strong light image data for identification; a new banknote, the weak light image data is selected for recognition; and in step 5, the number recognition result is output.
  • step 2 the coordinates of the four vertices of the banknote image are first obtained, and the slope of the straight line where the banknote frame is located is calculated by using the least squares straight line fitting method, and then the four edges of the image are acquired.
  • the specific process of obtaining a strong light image data and a low light image data of the banknote in the step 1 includes: Step 101: detecting a position sensor state; Step 102, determining whether the position sensor is triggered, and determining to be yes, then transferring Step 103, otherwise return to step 101; Step 103, start scanning, obtain banknote image data, the line counter is cleared, and then proceeds to step 104; Step 104, detecting the rising edge of MCLK; Step 105, determining whether the rising edge of MCLK is detected , if yes, go to the step 106, if no, return to step 104; in step 106, the LED switches the strong light illumination; in step 107, the FPGA drives the CIS to collect a line of image data, and then simultaneously proceeds to step 108 and step 120, step 108 and subsequent steps and step 120 And subsequent steps are processed in parallel; Step 108, detecting the rising edge of MCLK again; Step 109, determining whether the rising edge of MCLK is detected, if
  • Step 114 image acquisition is completed, and the scanning ends;
  • Step 120 the DSP enters the interrupt service routine;
  • Step 121 detecting an image type register;
  • Step 122 determining whether the image belongs to the strong light image data, if yes, proceeding to step 123, if not, proceeding to the step 124;
  • step 123 the line data stored in the image buffer light;
  • step 124 the low-light image data into the line buffer.
  • the banknote number identification device and the identification method provided by the present application can obtain two white light transmission images of different brightness of the same banknote at a time, and can judge the old and new state of the banknote, and can effectively select one of the banknote numbers to be clearly identified. Therefore, the number of the old and new banknotes can be effectively recognized, and the accuracy of the identification of the banknote number is improved, the flow is clear, and the structure is simple.
  • FIG. 1 is a schematic diagram of a hardware structure of a banknote number identification device according to a preferred embodiment of the present invention
  • FIG. 2 is a flow chart of a method for identifying a banknote number according to a preferred embodiment of the present application
  • Figure 3 is an exploded flow chart of step 1 of Figure 2.
  • a banknote number identification device provided by a preferred embodiment of the present application includes a touch sensor 10 for collecting banknote image data by means of transmission imaging; a white LED transmitting light source plate 20 disposed on the touch sensor Directly forward, a transmission light source is provided for collecting the banknote image data of the touch sensor; an image acquisition processing board 40 is connected to the touch sensor 10 through a data line, and a field editable logic gate is integrated on the image acquisition processing board 10 Field-Programmable Gate Array (FPGA) and a digital signal processing (DSP), the field editable logic gate array is used to control the white LED transmission light source panel to light off and drive the CIS to acquire an image, the digital signal
  • the processor is configured to process and recognize the banknote image data; the two pairs of position sensors 30 are disposed on the banknote transport channel at a predetermined distance from the touch sensor, and the predetermined distance is 3 cm in the embodiment to detect whether there is a banknote.
  • the ED transmission light source panel has two light source modes, and the two light source modes alternately provide a transmission light source for the contact sensor to collect banknote image data.
  • the image acquisition processing board further includes a memory, the memory includes two buffer areas and a register for storing banknote image data collected by the touch sensor, and the register is used to represent the banknote image in the buffer area. Type of data: 1 indicates strong light image data, and 0 indicates low light image data.
  • the FPGA drives the CIS to perform image data acquisition in a progressive scan mode.
  • the CIS horizontal resolution 200dpi and the vertical resolution to 100dpi.
  • the banknote passes through the channel at a speed of lm/s and advances at a constant speed.
  • the FPGA needs to collect one line of image data, so the FPGA internally sets a clock MCLK with a period of 0.125ms, whenever the rising edge of the clock is detected. When it arrives, it means that the banknote has advanced by 0.125mm, that is, the scan is started. Process, scanning a line of images.
  • the LED In order to obtain two images with different brightness, the LED needs to generate two different kinds of light.
  • the ratio is about 1: 3 (the two coefficients correspond to the processing of new and old banknotes respectively), the actual test corresponds to 30us and 90us, respectively, both less than the banknote advance 0.125
  • the time required for mm, the extra time, the LED is off.
  • the FPGA detects the rising edge of MCLK
  • the LED's on-time is switched once, and the LED is illuminated.
  • the CIS is sensed to reach the upper limit of the on-time value. The LED is off.
  • the LED is illuminated by strong light, the intensity of CIS induction is large, and the signal output amplitude is large.
  • analog-to-digital converter AD After conversion by analog-to-digital converter AD, a line of strong light image data is formed; LED is irradiated with weak light, CIS-induced light The intensity is small and the output amplitude is small.
  • analog-to-digital converter AD After being converted by the analog-to-digital converter AD, a line of low-light image data is formed. The FPGA needs to acquire this row of data into the image buffer in the slice within 0.125ms.
  • a register is also provided for indicating the type of image data in the image buffer area: 1 indicates strong light image data, and 0 indicates low light image data.
  • the DSP internally sets two buffer areas to store the two types of image data. After the FPGA collects one line of image data, it notifies the DSP to acquire the data by interrupt. The DSP enters the interrupt service routine, detects the image type register of the FPGA, learns the image type of the image, and then starts the Enhanced Direct Memory Access (EDMA) transmission, and stores the data in the space corresponding to the image type in order. .
  • the FPGA scans 120mm (equivalent to collecting 960 lines of data) and ends the scan. When all the image data has been transferred, the two buffers of the DSP will acquire two images, one for each image, one for each image, 480 lines.
  • the banknote number identification method includes: Step 1: acquiring a strong light image data of a banknote and a weak light by a transmission imaging method and alternating illumination of a transmitted light source strong light mode and a low light mode Image data; Step 2, performing edge detection on the low light image data, acquiring four edges of the image, and determining the target image; Step 3, processing the target image, identifying the version of the banknote, the face value, facing, and positioning the banknote number area; Step 4: Calling a pattern recognition algorithm to identify, using a pattern recognition method based on an artificial neural network, comparing the gray level of the banknote image with a threshold value, and identifying the oldness of the banknote, and if the old banknote is used, identifying the strong light image data; If the new banknote
  • step 2 the low-light image data is used for edge detection to obtain the target image.
  • step 3 since the position of the banknote number on the banknote is fixed, the banknote number area can be positioned according to the version, the face value, and the orientation of the banknote.
  • step 2 the coordinates of the four vertices of the banknote image are first obtained, and the slope of the straight line where the banknote frame is located is calculated by using the least squares straight line fitting method, and then the four edges of the image are acquired.
  • the specific process of acquiring a strong light image data and a low light image data of the banknote in the step 1 includes: Step 101: detecting a position sensor state; Step 102, determining whether the position sensor is triggered, determining that is, Then, go to step 103, otherwise go back to step 101; Step 103, start scanning, obtain banknote image data, the line counter is cleared, and then go to step 104; Step 104, detect the rising edge of MCLK; Step 105, determine whether the detected MCLK rising edge, if yes, go to step 106, if no, go back to step 104; Step 106, LED switches strong light illumination; Step 107, FPGA drives CIS to collect one line of image data, and then proceeds to step 108 and steps simultaneously 120, Step 108 and subsequent steps are processed in parallel with step 120 and subsequent steps; Step 108, detecting the rising edge of MCLK again; Step 109, determining whether the rising edge of MCLK is detected, and if yes, proceeding to step 110,
  • Step 112 the FPGA line counter is incremented by 2; Step 113, it is determined whether the FPGA line counter is equal to 960, if yes, go to step 114, if no, go back to step 104; Step 114, image collection is completed, end scanning; Step 120 , the DSP enters the interrupt service program; Step 121, detecting the image type register; Step 122, determining whether the image belongs to the strong light image data, if yes, proceeding to step 123, if not, proceeding to step 124; Step 123, The row data is stored in the highlight image buffer; in step 124, the row data is stored in the low light image buffer.
  • Steps 120 to 124 are image storage processes, that is, one line of image data is stored once per scan, and is stored in a sub-image data type, and the process can be performed in parallel with the image scanning process, for example, step 107 is simultaneously transferred.
  • Steps 108 and 120 may be stored first after each line of image scanning, and then scanned for the next line of images. That is to say, after step 107, the process may first proceed to step 120, and then perform steps 121 and 122 until step 123 or After 124, step 108 is performed.
  • this processing speed is reduced, and may affect the setting of the internal clock MCLK period of the field editable logic gate array, and also affect the setting of the white light LED transmission source on time, which may need to be re-adjusted. Therefore, it is preferable that the image scanning and the reading and storing of each line of image data are performed in parallel.
  • the dual-mode scanning is adopted, and the algorithm uses the clear image in the old banknote image to identify, and uses the dark image in the new banknote image to identify, and the banknote number image is clear, which is suitable for the algorithm to locate the banknote number and correctly identify it.

Abstract

一种纸币号码识别装置和方法,该装置包括:一接触式传感器(10),用以采集纸币图像数据;一白光LED透射光源板(20);一图像采集处理板(40),集成有一FPGA和一DSP,该FPGA用以控制白光LED透射光源板亮灭及驱动接触式传感器(10)获取图像,该DSP用以对纸币图像数据进行处理识别;两对位置传感器(30),用以检测是否有纸币进入该纸币识别装置;其中,该白光LED透射光源板(20)具有强弱两种光源模式,且该两种光源模式交替为该接触式传感器(10)采集纸币图像数据提供透射光源。该装置每次可获取同一张纸币的两幅亮度不同的白光透射图像,因此对新旧纸币的号码都能进行有效的识别,提升了对纸币号码识别的准确率。

Description

纸币号码识别装置和识别方法
本申请要求于 2012 年 11 月 15 日提交中国专利局、 申请号为 201210462080.3、 发明名称为"纸币号码识别装置和识别方法"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
本申请涉及金融自助设备领域, 尤其涉及纸币号码识别装置和识别方法。 背景技术
目前, 公知的纸币号码识别装置是通过 CIS接触式图像传感器或 CCD图 像传感器获取图像。 当纸币通过装置的通道, 上述图像传感器高速采集纸币图 像, 通过模数转换器 (AD )转换为图像数据, 数字信号处理器迅速处理图像 数据, 定位纸币号码的位置, 通过识别算法运算, 实现对纸币号码的识别。
由于纸币往往只有一面存在号码, 纸币通过纸币处理装置的方向是任意 的, 所以通常通过透射成像方式获取纸币号码, 只需要在纸币通道上安装一支 CIS , CIS前面安装一组白光 LED光源, 并设置固定的光照时间照射经过的纸 币, 穿透过纸币的光落在 CIS上感光成像, 无论纸币号码出现在 CIS的正面, 或者背面, 都能将纸币号码获取到。
但是,很多在市场上流通了一段时间的纸币,表面或多或少会累积了一层 灰尘, 使透射到 CIS图像传感器的光的能量减弱, 造成图像模糊, 难以辨认, 严重降低了算法识别率。 如果单独加长白光 LED的导通时间, 加强光的能量, 旧纸币成像变清晰, 能被识别, 由于新纸币的透射效果好, 却又使新纸币图像 的轮廓消失, 导致识别算法无法定位新纸币号码位置而造成识别错误。传统的 透射成像方式, 不能对新 1曰纸币同时形成清晰的图像, 给识别算法造成困难, 纸币号码识别容易出现错误。 发明内容
为了克服传统的透射成像处理方式不能同时对新旧纸币形成清晰图像的 不足, 本申请提供一种新的纸币号码识别方法和装置, 不仅能对半新或全新的 纸币号码进行正确识别, 而且能根据纸币的新旧情况, 对旧纸币的号码进行正 确的识别。
本申请提供了一种纸币号码识别装置, 包括: 一接触式传感器, 用以通 过透射成像方式采集纸币图像数据; 一白光 LED透射光源板, 设置在该接触 式传感器正前方, 为该接触式传感器采集纸币图像数据提供一透射光源; 一图 像采集处理板, 其通过数据线与该接触式传感器连接, 该图像采集处理板上集 成有一现场可编辑逻辑门阵列和一数字信号处理器,该现场可编辑逻辑门阵列 用以控制白光 LED透射光源板亮灭及驱动 CIS获取图像, 该数字信号处理器 用以对纸币图像数据进行处理识别; 以及两对位置传感器, 设置在纸币输送通 道上距离该接触式传感器一预定距离处,用以检测是否有紙币进入该紙币号码 识别装置; 其中, 该白光 LED透射光源板具有强弱两种光源模式, 且该两种 光源模式交替为该接触式传感器采集纸币图像数据提供透射光源。
优选的, 该强弱两种光源模式通过该现场可编辑逻辑门阵列控制该白光
LED透射光源板的导通时间形成, 导通时间长, 形成强光模式, 导通时间短, 形成弱光模式, 形成强光模式的导通时间与形成弱光模式的导通时间之比为 3:1。
优选的, 该图像采集处理板上还集成有一存储器, 该存储器包括一緩存区 和一寄存器, 该緩存区用以存储该接触式传感器釆集的纸币图像数据, 该寄存 器用以表示緩存区内纸币图像数据的类型: 1表示强光图像数据, 0表示弱光 图像数据。
优选的, 该现场可编辑逻辑门阵列 ( Field _ Programmable Gate Array, FPGA ) 内部设置有一周期为 0.125ms的时钟 MCLK和一个行计数器。
本申请还提供了一种纸币号码识别方法, 包括: 步骤 1 , 通过透射成像方 法且在透射光源强光模式和弱光模式交替照射情况下获取纸币的一强光图像 数据和一弱光图像数据; 步骤 2, 对该弱光图像数据进行边沿检测, 获取图像 的四个边沿, 确定目标图像; 步骤 3, 对目标图像处理, 辨认出纸币的版本, 面值, 面向, 定位纸币号码区域; 步骤 4, 调用模式识别算法进行识别, 使用 基于人工神经元网络的模式识别方法, 将纸币图像的灰度和阈值比较, 辨认纸 币新旧程度, 如果是旧纸币, 使用该强光图像数据进行识别; 如果是新纸币, 选择该弱光图像数据进行识别; 以及步骤 5, 输出号码识别结果。
优选的, 步骤 2中先获取纸币图像的四个顶点坐标, 采用最小二乘直线拟 合方法, 计算出纸币边框所在的直线斜率, 然后获取图像的四个边沿。
优选的,步骤 1中获取纸币的一强光图像数据和一弱光图像数据的具体过 程包括: 步骤 101 , 检测位置传感器状态; 步骤 102, 判断位置传感器是否被 触发, 判断为是, 则转入步骤 103, 否则回到步骤 101 ; 步骤 103, 开始扫描, 获取纸币图像数据,行计数器清零,然后转入步骤 104; 步骤 104,检测 MCLK 上升沿; 步骤 105, 判断是否检测到该 MCLK上升沿, 如果是, 则转入步骤 106,如果否,则回到步骤 104;步骤 106, LED切换强光照射;步骤 107, FPGA 驱动 CIS采集一行图像数据, 然后同时转入步骤 108和步骤 120, 步骤 108及 其后续步骤与步骤 120及其后续步骤并行处理; 步骤 108, 再次检测 MCLK 上升沿; 步骤 109, 判断是否检测到该 MCLK上升沿, 如杲是, 则转入步骤 110,如果否,则回到步骤 108; 步骤 110, LED切换弱光照射; 步骤 111 , FPGA 驱动 CIS再采集一行数据, 然后同时转入步骤 112和步骤 120, 步骤 112及其 后续步骤与步骤 120及其后续步骤并行处理; 步骤 112, FPGA行计数器加 2; 步骤 113, 判断 FPGA行计数器是否等于 960, 如果是, 转入步骤 114, 如果 否, 回到步骤 104; 步骤 114, 图像采集完成, 结束扫描; 步骤 120, DSP进 入中断服务程序; 步骤 121 , 检测图像类型寄存器; 步骤 122, 判断图像是否 属于强光图像数据, 如果是, 则转入步骤 123, 如果否, 则转入步骤 124; 步 骤 123, 将该行数据存入强光图像緩存区; 步骤 124, 将该行数据存入弱光图 像緩存区。
本申请提供的纸币号码识别装置和识别方法每次可获取同一张纸币的两 幅亮度不同的白光透射图像, 且可判断出纸币新旧状态, 能有效地选择其中一 幅纸币号码清晰的图像进行识别, 因此对新旧纸币的号码都能进行有效的识 另 J , 提升了对纸币号码识别的准确率, 流程清晰, 结构筒单。
附图说明
图 1是本申请一较佳实施例提供的纸币号码识别装置硬件结构示意图; 图 2是本申请一较佳实施例提供的纸币号码识别方法流程图;
图 3是图 2中步骤 1的分解流程图。
具体实施方式 下面结合附图, 对本申请实施例的技术方案进行清楚、 完整地描述。
本申请一较佳实施例提供的纸币号码识别装置如图 1所示,包括接触式传 感器 10, 用以通过透射成像方式采集纸币图像数据; 一白光 LED透射光源板 20,设置在该接触式传感器正前方, 为该接触式传感器采集纸币图像数据提供 一透射光源;一图像采集处理板 40,其通过数据线与该接触式传感器 10连接, 该图像采集处理板 10 上集成有一现场可编辑逻辑门阵列 ( Field - Programmable Gate Array , FPGA ) 和一数字信号处理器 ( Digital Signal Processing, DSP ),该现场可编辑逻辑门阵列用以控制白光 LED透射光源板亮 灭及驱动 CIS 获取图像, 该数字信号处理器用以对纸币图像数据进行处理识 别; 两对位置传感器 30, 设置在纸币输送通道上距离该接触式传感器一预定 距离处, 本实施例中该预定距离为 3厘米, 用以检测是否有纸币进入该纸币号 码识别装置; 其中, 该白光 LED透射光源板具有强弱两种光源模式, 且该两 种光源模式交替为该接触式传感器采集纸币图像数据提供透射光源。 另外, 该 图像采集处理板上还集成有一存储器, 该存储器包括两个緩存区和一寄存器, 该緩存区用以存储该接触式传感器采集的纸币图像数据,该寄存器用以表示緩 存区内纸币图像数据的类型: 1表示强光图像数据, 0表示弱光图像数据。
当纸币进入传输通道并触发位置传感器 30时, FPGA驱动 CIS采取逐行 扫描方式进行图像数据采集。为了使图像清晰,设置 CIS横向分辨率为 200dpi, 纵向分辨率为 100dpi。 假设纸币经过通道速度为 lm/s, 匀速前进。 为了采集 到两幅纵向分辨率为 100dpi图像, 紙币每前进大约 0.125mm, FPGA就需要 采集完一行图像数据,所以 FPGA内部设置一个周期为 0.125ms的时钟 MCLK, 每当检测到时钟的上升沿到来,就表示纸币已经前进了 0.125mm, 即启动扫描 进程, 扫描一行图像。
为了得到两幅亮度不同的图像,需要 LED产生强弱两种不同的光。将 LED 的导通时间设置为两个不同的值, 比例约为 1 : 3 (两个系数分别对应新纸币 和旧纸币的处理 ),实际测试分别对应为 30us和 90us,都少于纸币前进 0.125mm 所需要的时间, 多出的时间, LED处于熄灭状态。 纸币前进的过程中, FPGA 每次检测到 MCLK 的上升沿到来, 就将 LED 的导通时间切换一次, 并点亮 LED, LED照射的时间段内, 使 CIS感光, 达到导通时间值上限, LED熄灭。 对同一张纸币, LED采用强光照射, CIS感应的光强变大, 信号输出幅度大, 通过模数转换器 AD转换后, 形成一行强光图像数据; LED采用弱光照射, CIS感应的光强变小, 输出幅度小, 通过模数转换器 AD转换后, 形成一行弱 光图像数据。 FPGA需要在 0.125ms内, 将这一行数据获取进片内的图像緩存 区中。
FPGA釆集数据时, 除了图像数据緩存区, 还设置了一个寄存器, 用于表 示图像緩存区内图像数据的类型: 1表示强光图像数据, 0表示弱光图像数据。 DSP内部设置两个緩存区, 分别存放这两种类型的图像数据。 FPGA采集完一 行图像数据后, 通过中断方式通知 DSP获取数据。 DSP进入中断服务程序, 检测 FPGA的图像类型寄存器, 获知此次的图像类型, 然后启动增强型直接 内存存取( Enhanced Direct Memory Access, EDMA )传输, 将数据按顺序存 放到对应图像类型的空间中。 FPGA扫描 120mm (相当于采集 960行数据 )结 束扫描。 当所有图像数据传输完毕, DSP两个緩冲区将分别会获取到一明一暗 两幅图像, 每幅图像 480行。
DSP对明暗两幅图像进行总体识别,判断出图像属于旧纸币, 就采用明图 像进行识别, 反之, 则采用暗图像进行识别, 达到识别新旧纸币号码的目的。 具体的, 如图 2所示, 该纸币号码识别方法包括: 步骤 1 , 通过透射成像 方法且在透射光源强光模式和弱光模式交替照射情况下获取纸币的一强光图 像数据和一弱光图像数据; 步骤 2, 对该弱光图像数据进行边沿检测, 获取图 像的四个边沿, 确定目标图像; 步骤 3, 对目标图像处理, 辨认出纸币的版本, 面值, 面向, 定位纸币号码区域; 步骤 4, 调用模式识别算法进行识别, 使用 基于人工神经元网络的模式识别方法, 将纸币图像的灰度和阈值比较, 辨认纸 币新旧程度, 如果是旧纸币, 使用该强光图像数据进行识别; 如杲是新纸币, 选择该弱光图像数据进行识别; 以及步骤 5, 输出号码识别结果。
由于弱光图像边沿和背景像素值差异大,新旧纸币都能使用其进行边沿检 测, 因此, 步骤 2中采用弱光图像数据进行边沿检测, 以获得目标图像。
步骤 3中,由于纸币号码在纸币上的位置是固定的,因此根据纸币的版本、 面值以及面向, 可定位纸币号码区域。
优选的, 步骤 2中先获取纸币图像的四个顶点坐标, 采用最小二乘直线拟 合方法, 计算出纸币边框所在的直线斜率, 然后获取图像的四个边沿。
如图 3所示,步骤 1中获取纸币的一强光图像数据和一弱光图像数据的具 体过程包括: 步骤 101 , 检测位置传感器状态; 步骤 102, 判断位置传感器是 否被触发, 判断为是, 则转入步骤 103 , 否则回到步骤 101; 步骤 103 , 开始 扫描, 获取纸币图像数据, 行计数器清零, 然后转入步骤 104; 步骤 104, 检 测 MCLK上升沿; 步骤 105 , 判断是否检测到该 MCLK上升沿, 如果是, 则 转入步骤 106, 如果否, 则回到步骤 104; 步骤 106, LED切换强光照射; 步 骤 107, FPGA驱动 CIS采集一行图像数据,然后同时转入步骤 108和步骤 120, 步骤 108及其后续步骤与步骤 120及其后续步骤并行处理; 步骤 108, 再次检 测 MCLK上升沿; 步驟 109, 判断是否检测到该 MCLK上升沿, 如果是, 则 转入步骤 110, 如果否, 则回到步骤 108; 步骤 110, LED切换弱光照射; 步 111 , FPGA驱动 CIS再采集一行数据, 然后同时转入步骤 112和步骤 120, 步骤 112及其后续步骤与步骤 120及其后续步骤并行处理; 步骤 112, FPGA 行计数器加 2; 步骤 113, 判断 FPGA行计数器是否等于 960, 如果是, 转入 步骤 114, 如果否, 回到步骤 104; 步骤 114, 图像釆集完成, 结束扫描; 步 骤 120, DSP进入中断服务程序; 步骤 121 , 检测图像类型寄存器; 步骤 122, 判断图像是否属于强光图像数据, 如果是, 则转入步骤 123, 如果否, 则转入 步骤 124; 步骤 123 , 将该行数据存入强光图像緩存区; 步骤 124, 将该行数 据存入弱光图像緩存区。
其中, 步骤 120至步骤 124为图像的存储过程, 也就是说, 每扫描一行图 像数据存储一次, 而且是分图像数据类型存储, 该过程可以与图像扫描过程并 列进行, 例如步骤 107后同时转入步骤 108和步骤 120, 也可以每扫描完一行 图像即先存储, 然后再扫描下一行图像, 也就是说, 步骤 107后可先转入步骤 120, 依次执行步骤 121、 122直至执行完步骤 123或 124之后, 再执行步骤 108。 当然, 这样处理速度降低, 而且可能影响现场可编辑逻辑门阵列内部时 钟 MCLK的周期的设置, 同时也影响白光 LED透射光源导通时间的设置, 都 可能需要重新调整。 因此优选为图像扫描与每行图像数据的读取存储并列进 行。
如果是单模式扫描, 只采集一幅弱光图像数据, 则只有五成新以上纸币满 足算法识别条件, 而对于旧妙, 表面存在灰尘, 油污, 使纸币透光率降低, 导 致图像灰度值过低, 虽然图像边沿能正确获取, 但图像对比度低, 算法将不能 有效分辨妙票的面额, 面向, 继而不能确定妙票号码在钞票上的位置。 同时图 像对比度低, 区分困难,妙票号码识别正确率大幅度下降,严重影响识别效果。 如果光的强度加大, 只采集一幅强光图像数据, 旧纸币满足算法识别条件, 但 新纸币将不能满足, 因为对于新纸币, 透光率高, 容易出现图像过曝, 边沿图 像缺失, 算法不能有效定位钞票的长度和宽度。钞票号码个別数字也会产生缺 少, 导致识别错误。
本实施例采用双模式扫描, 算法采用旧纸币图像中的明图像进行识别, 采 用新纸币图像中的暗图像进行识别, 其纸币号码图像清晰, 适合算法进行定位 纸币号码位置及正确识别。
以上所述, 仅为本申请的具体实施方式,但本申请的保护范围并不局限于 此,任何熟悉本技术领域的技术人员在本申请揭露的技术范围内, 可轻易想到 变化或替换, 都应涵盖在本申请的保护范围之内。 因此, 本申请的保护范围应 所述以权利要求的保护范围为准。

Claims

1、 一种纸币号码识別装置, 包括: 一接触式传感器, 用以通过透射成像方式采集纸币图像数据; 一白光 LED透射光源板, 设置在该接触式传感器正前方, 为该接触式传 权
感器采集纸币图像数据提供一透射光源;
利 _
一图像采集处理板, 其通过数据线与该接触式传感器连接, 该图像采集处 o要
理板上集成有一现场可编辑逻辑门阵列和一数字信号处理器,该现场可编辑逻 求
辑门阵列用以控制白光 LED透射光源板亮灭及驱动 CIS获取图像, 该数字信 号处理器用以对纸币图像数据进行处理识别; 以及 两对位置传感器,设置在纸币输送通道上距萬该接触式传感器一预定距离 处, 用以检测是否有纸币进入该纸币号码识别装置; 其特征在于, 该白光 LED透射光源板具有强弱两种光源模式, 且该两种 光源模式交替为该接触式传感器采集纸币图像数据提供透射光源。
2、 如权利要求 1所述的紙币号码识别装置, 其特征在于, 该强弱两种光 源模式通过该现场可编辑逻辑门阵列控制该白光 LED透射光源板的导通时间 形成, 导通时间长, 形成强光模式, 导通时间短, 形成弱光模式, 形成强光模 式的导通时间与形成弱光模式的导通时间之比为 3: 1。
3、 如权利要求 1所述的紙币号码识别装置, 其特征在于, 该图像采集处 理板上还集成有一存储器, 该存储器包括两个緩存区和一寄存器, 该緩存区用 以存储该接触式传感器采集的纸币图像数据,该寄存器用以表示緩存区内纸币 图像数据的类型: 1表示强光图像数据, 0表示弱光图像数据。
4、 如权利要求 1所述的纸币号码识别装置, 其特征在于, 该现场可编辑 逻辑门阵列内部设置有一周期为 0.125ms的时钟 MCLK和一个行计数器。
5、 一种纸币号码识别方法, 包括:
步骤 1 , 通过透射成像方法且在透射光源强光模式和弱光模式交替照射情 况下获取纸币的一强光图像数据和一弱光图像数据;
步骤 2, 对该弱光图像数据进行边沿检测, 获取图像的四个边沿, 确定目 标图像;
步骤 3, 对目标图像处理, 辨认出纸币的版本, 面值, 面向, 定位纸币号 码区域;
步骤 4, 调用模式识别算法, 使用基于人工神经元网络的模式识别方法, 将纸币图像的灰度和阈值比较, 辨认纸币新 1日程度, 如果是 1日纸币, 使用该强 光图像数据进行识别; 如果是新纸币, 选择该弱光图像数据进行识别; 以及 步骤 5, 输出号码识别结果。
6、 根据权利要求 5所述的纸币号码识别方法, 其特征在于, 步骤 2中先 获取纸币图像的四个顶点坐标, 采用最小二乘直线拟合方法, 计算出纸币边框 所在的直线斜率, 然后获取图像的四个边沿。
7、 根据权利要求 5所述的纸币号码识别方法, 其特征在于, 步骤 1中获 取纸币的一强光图像数据和一弱光图像数据的具体过程包括:
步骤 101 , 检测位置传感器状态;
步骤 102, 判断位置传感器是否被触发, 判断为是, 则转入步骤 103 , 否 则回到步骤 101;
步骤 103, 开始扫描, 获取纸币图像数据, 行计数器清零, 然后转入步骤 104;
步骤 104, 检测 MCLK上升沿;
步骤 105, 判断是否检测到该 MCLK上升沿, 如果是, 则转入步骤 106, 如果否, 则回到步骤 104;
步骤 106, LED切换强光照射;
步骤 107, FPGA驱动 CIS采集一行图像数据, 然后同时转入步骤 108和 步骤 120, 步骤 108及其后续步骤与步骤 120及其后续步骤并行处理;
步骤 108, 再次检测 MCLK上升沿;
步骤 109, 判断是否检测到该 MCLK上升沿, 如果是, 则转入步骤 110, 如果否, 则回到步骤 108;
步骤 110, LED切换弱光照射;
步骤 111 , FPGA驱动 CIS再采集一行数据, 然后同时转入步骤 112和步 骤 120, 步骤 112及其后续步骤与步骤 120及其后续步骤并行处理;
步骤 112, FPGA行计数器加 2;
步骤 113, 判断 FPGA行计数器是否等于 960, 如果是, 转入步骤 114, 如果否, 回到步骤 104;
步骤 114, 图像采集完成, 结束扫描;
步骤 120, DSP进入中断服务程序;
步骤 121 , 检测图像类型寄存器;
步骤 122, 判断图像是否属于强光图像数据, 如果是, 则转入步骤 123, 如果否, 则转入步骤 124;
步骤 123, 将该行数据存入强光图像緩存区; 步骤 124, 将该行数据存入 弱光图像緩存区。
PCT/CN2013/078911 2012-11-15 2013-07-05 纸币号码识别装置和识别方法 WO2014075452A1 (zh)

Priority Applications (4)

Application Number Priority Date Filing Date Title
AU2013347512A AU2013347512B2 (en) 2012-11-15 2013-07-05 Paper currency number recognition apparatus and recognition method
US14/426,972 US9741193B2 (en) 2012-11-15 2013-07-05 Paper currency number recognition apparatus and recognition method
EP13854512.4A EP2922037A4 (en) 2012-11-15 2013-07-05 DETECTION DEVICE AND DETECTION PROCEDURE FOR DIRECTORY NUMBERS
ZA2015/02309A ZA201502309B (en) 2012-11-15 2015-04-07 Paper currency number recognition apparatus and recognition method

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201210462080.3A CN102930636B (zh) 2012-11-15 2012-11-15 纸币号码识别装置和识别方法
CN201210462080.3 2012-11-15

Publications (1)

Publication Number Publication Date
WO2014075452A1 true WO2014075452A1 (zh) 2014-05-22

Family

ID=47645424

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2013/078911 WO2014075452A1 (zh) 2012-11-15 2013-07-05 纸币号码识别装置和识别方法

Country Status (7)

Country Link
US (1) US9741193B2 (zh)
EP (1) EP2922037A4 (zh)
CN (1) CN102930636B (zh)
AU (1) AU2013347512B2 (zh)
CL (1) CL2015001060A1 (zh)
WO (1) WO2014075452A1 (zh)
ZA (1) ZA201502309B (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104916037A (zh) * 2015-06-26 2015-09-16 深圳市中钞信达金融科技有限公司 一种纸币冠字号识别装置及方法

Families Citing this family (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102930636B (zh) 2012-11-15 2014-10-22 广州广电运通金融电子股份有限公司 纸币号码识别装置和识别方法
CN103049750B (zh) 2013-01-11 2016-06-15 广州广电运通金融电子股份有限公司 字符识别方法
CN103345798B (zh) * 2013-06-17 2015-09-23 中国人民银行印制科学技术研究所 一种高速实时检测片状材料的系统
CN103310528B (zh) * 2013-07-08 2016-08-17 广州广电运通金融电子股份有限公司 图像补偿修正方法及识别验钞装置
TW201503062A (zh) * 2013-07-09 2015-01-16 Climax Machine Industry Co Ltd 整鈔裝置之系統及方法
CN103400440A (zh) * 2013-08-15 2013-11-20 广州广电运通金融电子股份有限公司 一种磁信号检测方法及装置
CN103413170B (zh) * 2013-08-19 2016-05-11 电子科技大学 用于点钞机纸币边缘的实时定位方法
CN104464077A (zh) * 2014-11-12 2015-03-25 迪堡金融设备有限公司 一种纸币冠字码的识别方法及其装置
CN105447826B (zh) * 2015-11-06 2018-09-07 东方通信股份有限公司 一种纸币图像采集的处理方法
CN115212547A (zh) 2015-11-19 2022-10-21 天使集团股份有限公司 桌面游戏的管理系统、游戏币及检查装置
CN106856021A (zh) * 2016-10-08 2017-06-16 长沙恒强电子科技有限公司 一体化防伪检测传感器模块及具有该模块的点验钞机
CN108022363A (zh) * 2016-11-02 2018-05-11 深圳怡化电脑股份有限公司 一种纸币面向的识别方法及装置
CN107085885A (zh) * 2017-03-22 2017-08-22 深圳怡化电脑股份有限公司 一种验钞数据的采集方法、存储方法及其装置
CN107610319A (zh) * 2017-08-07 2018-01-19 上海灵岳电子设备有限公司 基于单接触式图像传感器的纸币信息识别装置及方法
US11392521B2 (en) * 2020-04-13 2022-07-19 Tcl China Star Optoelectronics Technology Co., Ltd. Heterogeneous computing system and heterogeneous computing method using the system
CN111737169B (zh) * 2020-07-21 2020-11-27 成都智明达电子股份有限公司 一种基于edma的大容量高速行列输出缓存结构的实现方法
CN113538809B (zh) * 2021-06-11 2023-08-04 深圳怡化电脑科技有限公司 一种基于自助设备的数据处理方法和装置
CN114882632A (zh) * 2022-04-02 2022-08-09 温州市质量技术检测科学研究院 一种基于色彩识别防伪的纸币新旧检测点钞机及其方法

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH03191495A (ja) * 1989-12-21 1991-08-21 Shikoku Keisoku Kogyo Kk 紙幣識別装置
KR101014974B1 (ko) * 2008-08-19 2011-02-16 엘지엔시스(주) 지폐 이미지 인식 장치 및 방법
CN102005078A (zh) * 2010-12-23 2011-04-06 北京新岸线软件科技有限公司 一种纸币、票券识别方法和装置
CN102169608A (zh) * 2011-05-27 2011-08-31 徐端全 多光谱纸币验钞机的设计方法
CN102324133A (zh) * 2011-08-10 2012-01-18 深圳市怡化电脑有限公司 用于钞票识别的接触式图像传感器模组
CN102930636A (zh) * 2012-11-15 2013-02-13 广州广电运通金融电子股份有限公司 纸币号码识别装置和识别方法

Family Cites Families (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE2204284B2 (de) * 1972-01-29 1974-04-25 National Rejectors Inc. Gmbh, 2150 Buxtehude Vorrichtung zur lichtelektrischen Prüfung von blattförmigem Material, insbesondere von Dokumenten oder Banknoten
US4769532A (en) * 1986-07-11 1988-09-06 Laurel Bank Machines Co., Ltd. Apparatus for adjusting optical sensors with threshold memory
US6278795B1 (en) * 1995-12-15 2001-08-21 Cummins-Allison Corp. Multi-pocket currency discriminator
US8162125B1 (en) * 1996-05-29 2012-04-24 Cummins-Allison Corp. Apparatus and system for imaging currency bills and financial documents and method for using the same
JP3280289B2 (ja) * 1997-09-10 2002-04-30 横浜電子工業株式会社 有価証券識別装置及び方法
US6078683A (en) 1997-11-20 2000-06-20 De La Rue, Inc. Method and system for recognition of currency by denomination
JP2000307819A (ja) * 1999-04-26 2000-11-02 Glory Ltd 多波長光源を用いた画像読取装置及びその制御方法
US8401268B1 (en) * 2007-03-09 2013-03-19 Cummins-Allison Corp. Optical imaging sensor for a document processing device
CA2682467C (en) * 2007-03-29 2016-12-06 Glory Ltd. Paper-sheet recognition apparatus, paper-sheet processing apparatus, and paper-sheet recognition method
US8682056B2 (en) * 2008-06-30 2014-03-25 Ncr Corporation Media identification
US8736587B2 (en) * 2008-07-10 2014-05-27 Semiconductor Energy Laboratory Co., Ltd. Semiconductor device
KR100976245B1 (ko) * 2008-07-25 2010-08-18 엘지엔시스(주) 센서 레벨링 방법 및 장치, 그리고 그를 이용한 자동화기기
JP2010050551A (ja) * 2008-08-19 2010-03-04 Canon Inc 画像形成装置
US8194237B2 (en) 2009-10-15 2012-06-05 Authentix, Inc. Document sensor
CN202306684U (zh) * 2011-08-10 2012-07-04 深圳市怡化电脑有限公司 用于钞票识别的接触式图像传感器模组
US8804101B2 (en) * 2012-03-16 2014-08-12 Advanced Scientific Concepts, Inc. Personal LADAR sensor

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH03191495A (ja) * 1989-12-21 1991-08-21 Shikoku Keisoku Kogyo Kk 紙幣識別装置
KR101014974B1 (ko) * 2008-08-19 2011-02-16 엘지엔시스(주) 지폐 이미지 인식 장치 및 방법
CN102005078A (zh) * 2010-12-23 2011-04-06 北京新岸线软件科技有限公司 一种纸币、票券识别方法和装置
CN102169608A (zh) * 2011-05-27 2011-08-31 徐端全 多光谱纸币验钞机的设计方法
CN102324133A (zh) * 2011-08-10 2012-01-18 深圳市怡化电脑有限公司 用于钞票识别的接触式图像传感器模组
CN102930636A (zh) * 2012-11-15 2013-02-13 广州广电运通金融电子股份有限公司 纸币号码识别装置和识别方法

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104916037A (zh) * 2015-06-26 2015-09-16 深圳市中钞信达金融科技有限公司 一种纸币冠字号识别装置及方法

Also Published As

Publication number Publication date
AU2013347512A1 (en) 2015-04-02
ZA201502309B (en) 2016-01-27
US20150279141A1 (en) 2015-10-01
CN102930636B (zh) 2014-10-22
AU2013347512B2 (en) 2016-06-02
EP2922037A1 (en) 2015-09-23
EP2922037A4 (en) 2015-11-11
CN102930636A (zh) 2013-02-13
CL2015001060A1 (es) 2015-06-26
US9741193B2 (en) 2017-08-22

Similar Documents

Publication Publication Date Title
WO2014075452A1 (zh) 纸币号码识别装置和识别方法
RU2643493C1 (ru) Способ распознавания банкнот на основе накопления пыли в сортировщике и сортировщик
US8483472B2 (en) Paper sheet identifying device and paper sheet identifying method
WO2012051925A1 (zh) 纸币重张的检测方法及装置、自助终端
WO2016078455A1 (zh) 纸币褶皱的识别装置和方法
WO2016184336A1 (zh) 纸币检测方法和纸币处理装置
WO2014173012A1 (zh) 一种金融票据识别模块中积灰检测方法及系统
KR20160084810A (ko) 어종 판별 장치 및 어종 판별 방법
CN101547287A (zh) 图像扫描装置和图像形成设备
JP2008524683A (ja) シート物受入装置
CN103188418B (zh) 图像读取装置及图像形成装置
CN104284053A (zh) 倾斜角度校正装置、图像读取装置、图像形成装置及程序
US8265336B2 (en) Paper identifying apparatus and paper identifying method
EP1601599B1 (en) Optical double feed detection
CN110402386B (zh) 圆筒体表面检查装置及圆筒体表面检查方法
CN102158623A (zh) 图像读取装置
JP5732605B2 (ja) 外観検査装置
JPWO2003054530A1 (ja) 銅箔検査装置、銅箔検査方法、欠陥検査装置、欠陥検査方法
JP2003222597A (ja) 銅箔表面検査装置、および銅箔表面検査方法
JP2003156451A (ja) 欠陥検出装置
CN107218929B (zh) 一种基于机器视觉技术的水松纸折边检测方法
AU2016407875B2 (en) Sheet information collection device and sheet information collection method
US20190228607A1 (en) Sheet recognition device and sheet recognition method
JP2014186030A (ja) 欠点検査装置
RU2305322C2 (ru) Способ и устройство устранения шумов, создаваемых водяными знаками, при определении степени ветхости банкнот в сортирующих машинах

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: 13854512

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 14426972

Country of ref document: US

WWE Wipo information: entry into national phase

Ref document number: 2013854512

Country of ref document: EP

ENP Entry into the national phase

Ref document number: 2013347512

Country of ref document: AU

Date of ref document: 20130705

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE