WO2016119394A1 - 一种基于Lab色彩空间的人民币变色油墨检测方法和装置 - Google Patents
一种基于Lab色彩空间的人民币变色油墨检测方法和装置 Download PDFInfo
- Publication number
- WO2016119394A1 WO2016119394A1 PCT/CN2015/082616 CN2015082616W WO2016119394A1 WO 2016119394 A1 WO2016119394 A1 WO 2016119394A1 CN 2015082616 W CN2015082616 W CN 2015082616W WO 2016119394 A1 WO2016119394 A1 WO 2016119394A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- gray
- color
- changing ink
- ink character
- lab
- 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.)
- Ceased
Links
Images
Classifications
-
- 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
Definitions
- the invention relates to the field of RMB color changing ink detection, and particularly relates to a RMB color changing ink detecting method based on Lab color space and a RMB color changing ink detecting device based on Lab color space.
- the traditional RMB detection equipment is used for color-changing ink area detection, mainly by collecting RMB image signals, such as single-channel grayscale images or multi-spectral images R, G, B three-channel grayscale images, and counting each color-changing ink character in an independent channel.
- the gray value mean distribution on the image, the gray value difference between adjacent characters or the character area, whether it is within a certain threshold range, and if it exceeds a certain threshold range, it is judged as a color ink counterfeit currency.
- the color changing ink character in the case where the color changing ink character is slightly smeared or worn, the color changing ink character does not change significantly in the gray value or is near the set critical threshold, and the character description of the color changing ink is only passed from Judging the gray value of the three channels of single channel image or multi-spectral image R, G, B, it is difficult to correctly identify the authenticity of the color-changing ink characters that are slightly smeared or worn, and reduce the detection effect of the color-changing ink damaged banknotes. .
- Embodiments of the present invention provide a method and apparatus for detecting a renminbi color ink based on a Lab color space, which are used to quantitatively describe a color attribute of a color smear ink character that is slightly smeared or worn, and improve the detection effect of a color change ink damaged banknote.
- the Lab color feature value of the RMB to be detected is discriminated with the corresponding preset condition to determine whether the preset condition is met. If not, the RMB to be detected is a counterfeit currency, and the Lab color feature value is Includes Gray a, Gray b, and character area for each color-changing ink character.
- Determining the Lab color feature value of the RMB to be detected and the corresponding preset condition one by one includes:
- the embodiment of the invention further provides a RMB color changing ink detecting device based on the Lab color space, and the technical solution comprises:
- Obtaining a module configured to obtain an RGB image of the RMB to be detected
- An identification module configured to identify face value oriented information in the RGB image acquired by the acquiring module
- a positioning module configured to locate a color changing ink character according to the face value information recognized by the identification module
- a calculation module configured to calculate gray values of each color of the red (R), green (G), and blue (B) gray R, Gray G, Gray B;
- a conversion module for converting Gray R, Gray G, and Gray B into gray values Gray a, Gray b in the Lab color space
- the judging module discriminates the Lab color feature value of the RMB to be detected and the corresponding preset condition to determine whether the preset condition is met, and the Lab color feature value includes a Gray of each color ink character. , Gray b, character area;
- the execution module if at least one of the preset conditions is not met, the RMB to be detected is a counterfeit currency.
- the judging module comprises a single color changing ink character discriminating unit, an adjacent color changing ink character discriminating unit, and a single color changing ink character area discriminating unit.
- a first determining subunit determining whether the size of Gray a and Gray b of each color changing ink character, and whether Gray a satisfies a corresponding preset condition
- the first execution subunit if not, is a counterfeit currency.
- the second determining subunit determines whether the size of Gray a of one color changing ink character and the color of another color changing ink character, the size of one color ink character Gray b and the other color changing ink character Gray b satisfy the corresponding preset condition ;
- the second execution subunit if not, is a counterfeit currency.
- a third determining subunit determining whether a character area of each color changing ink character satisfies a corresponding preset condition
- the third execution subunit if not, is a counterfeit currency.
- the acquired RGB image of the renminbi to be detected identifies the face value information of the renminbi to be detected, locates the color changing ink character according to the face value information, and calculates the gradation value of each color ink character in the red, green and blue bands.
- Gray R, Gray G, Gray B then convert the calculated Gray R, Gray G, Gray B into gray values Gray a, Gray b in the Lab color space, since the Lab color space includes the luminance component L, from green to The red a component and the b component from blue to yellow, so Gray a, Gray b quantify the color attribute of the color changing ink character, and then determine the Gray a, Gray b and the corresponding preset conditions one by one, and judge Whether the corresponding preset conditions are satisfied, because the color attributes of the color change inks of the genuine coins and the counterfeit coins are large, by comparing the color attributes of the color changing inks, the color detecting inks have a strong detection effect even when the color changing ink is slightly smeared or rubbed. To increase the acceptance rate of the device for normal banknotes.
- FIG. 1 is a schematic flow chart of an embodiment of a method for detecting a renminbi color changing ink based on a Lab color space according to an embodiment of the present invention
- FIG. 2 is a schematic structural view of a renminbi color changing ink detecting device based on a Lab color space according to an embodiment of the present invention.
- Embodiments of the present invention provide a method and apparatus for detecting a renminbi color ink based on a Lab color space, which are used to quantitatively describe a color attribute of a color smear ink character that is slightly smeared or worn, and improve the detection effect of a color change ink damaged banknote.
- an embodiment of a method for detecting a renminbi color ink based on a Lab color space includes:
- the image of the renminbi to be detected in the RGB spectrum may be acquired by the multi-spectral identification device, and the RGB image may be acquired by other methods, which is not limited herein.
- the face value of the RMB to be detected includes 100, 50, 20, etc., and the faces can be divided into four types: face up, face down, back up and back down.
- the face value of the RMB to be detected is identified by the difference of the features of the banknotes of different denominations.
- the area where the color-changing ink character is located is located according to the recognized face value-oriented information.
- the color changing ink characters are divided, and the gray values of the color (R), green (G), and blue (B) gray R, Gray G, and Gray B are calculated for each color ink character. That is, the gray values Gray R, Gray G, and Gray B of each of the red (R), green (G), and blue (B) channels of each color changing ink character are calculated.
- Gray R, Gray G, and Gray B of each color-changing ink character are converted into gray values Gray a, Gray b in the Lab color space according to a conversion formula.
- step 107 Determine the Lab color feature value of the RMB to be detected and the corresponding preset condition to determine whether the preset condition is met. If not, execute step 107.
- the Lab color feature value includes Gray a, Gray b, and the character area of each color ink character; and the Lab color feature value of the RMB to be detected is discriminated with the corresponding preset condition, if at least one is not satisfied. If the condition is preset, step 107 is performed.
- the RMB to be tested is a counterfeit currency.
- the one-to-one discrimination refers to whether the Lab color feature value of the RMB to be tested satisfies the corresponding preset condition, and if one of the values is not satisfied, the RMB to be detected is a counterfeit currency.
- the face value information of the RMB to be detected is identified, and the color changing ink character is located according to the face value information, and each color ink character is calculated in red, green, and blue.
- the gray value of the band, Gray R, Gray G, and Gray B then converts the calculated Gray R, Gray G, and Gray B into gray values Gray a, Gray b in the Lab color space, since the Lab color space includes the luminance component.
- Gray a, Gray b quantify the color properties of the color changing ink characters, and then Gray a, Gray b and the corresponding preset conditions Judging one by one, judging whether the corresponding pre-conditions are satisfied, due to the difference in color attributes of the color-changing inks of genuine coins and counterfeit coins Don't be big, by comparing the color properties of the color-changing ink, if the color-changing ink is slightly smeared or rubbed, it still has a strong detection effect, and improves the acceptance rate of the device for normal genuine banknotes.
- the discrimination between the Lab color feature value of the RMB to be detected and the corresponding preset condition may include a single color ink character discrimination, adjacent color ink character discrimination, and a single color ink character area discrimination, which are respectively described below. :
- the above method is used for discriminating one by one, and the detection effect of the damaged banknote of the color changing ink is improved, and the flexibility of operation is improved.
- an embodiment of a RMB color changing ink detecting apparatus based on Lab color space includes:
- the obtaining module 201 is configured to obtain an RGB image of the RMB to be detected
- the identification module 202 is configured to identify face value oriented information in the RGB image acquired by the obtaining module 201;
- the positioning module 203 is configured to locate the color changing ink character according to the face value information recognized by the identification module 202;
- the calculation module 204 is configured to calculate gray values of the color (R), green (G), and blue (B) of each color changing ink character, Gray R, Gray G, Gray B;
- the conversion module 205 is configured to convert the Gray R, the Gray G, and the Gray B into the Lab color space. Gray value under the gray a, Gray b;
- the judging module 206 discriminates the Lab color feature value of the RMB to be detected and the corresponding preset condition to determine whether the preset condition is met, and the Lab color feature value includes Gray a, Gray b, and characters of each color changing ink character. area.
- the execution module 207 if at least one of the preset conditions is not met, the RMB to be detected is a counterfeit currency.
- the determining module 206 may further include: a single color changing ink character discriminating unit, an adjacent color changing ink character discriminating unit, and a single color changing ink character area discriminating unit.
- a first determining subunit determining whether the size of Gray a and Gray b of each color changing ink character, and whether Gray a satisfies a corresponding preset condition
- the first execution subunit if not, is a counterfeit currency.
- the second determining subunit determines whether the size of Gray a of one color changing ink character and the color of another color changing ink character, the size of one color ink character Gray b and the other color changing ink character Gray b satisfy the corresponding preset condition ;
- the second execution subunit if not, is a counterfeit currency.
- a third determining subunit determining whether a character area of each color changing ink character satisfies a corresponding preset condition
- the third execution subunit if not, is a counterfeit currency.
- the obtaining module 201 acquires the RGB image of the RMB to be detected, and then the recognition module 202 recognizes the face value oriented information, the positioning module 203 realizes the positioning and segmentation of the color changing ink characters, and the calculation module 204 and the conversion module 205 count the statistics.
- the Lab color feature value of each color changing ink character realizes the quantification and feature statistics of the color information.
- the judging module 206 performs various discriminations on the quantized color values, and the execution module 207 realizes the detection of the color changing ink to improve the color changing ink.
- the judging module 206 can discriminate in a plurality of ways, thereby improving the flexibility of operation. Sex.
- the Lab color space is a color space that is closer to human vision, where L (Luminosity) represents brightness and a and b represent opposite dimensions of two colors.
- L Luminosity
- a represents the range from red to green
- the value of a is from +127 to -128
- b represents the range from blue to yellow
- b is taken.
- the value is +127 to -128, which is warm and negative.
- Lab's color gamut is wide and wide. It not only contains all the gamuts of RGB and CMYK, but also the colors they can't express. The color that the human eye can perceive can be expressed by the Lab model.
- the Lab color model can compensate for the unequal color distribution of the RGB color model because the RGB model has too many transition colors between blue and green, and yellow and other colors between green and red.
- the 100 yuan RMB color-changing ink character observes the color from the front as green.
- the green principal component is in the negative direction of the a-axis, and the closer to the negative direction, the greener the color.
- the specific steps include:
- Multispectral image acquisition An image of a 100-yuan denomination banknote in the RGB spectrum is acquired using a multi-spectral identification device.
- the conversion formula is:
- Gray a 1.4749*(0.2213*Gray R-0.3390*Gray G+0.1177*Gray B)+128;
- Gray b 0.6245*(0.1949*Gray R+0.6057*Gray G-0.8006*Gray B)+128;
- the "100" three-character Lab mean distribution is counted as “1" (Gray L1, Gray a1, Gray b1), "0" (Gray L0, Gray a0, Gray b0), “0” (Gray L00). , Gray a00, Gray b00); Secondly, the character area that satisfies the color distribution of the normal color changing ink Lab is distributed as “1” (Lab Area1), "0" (Lab Area0), and "0" (Lab Area00).
- the Lab color feature values of the RMB to be detected are discriminated one by one with the corresponding preset conditions.
- the discrimination of the color-changing ink zone is achieved based on the comparison of the color-changing ink Lab color characteristic value calculated for each banknote with the normal banknote.
- Discrimination condition 1 single color change ink character Lab color feature value discrimination.
- the three color-changing ink characters are respectively subjected to the following logical discrimination, and all the conditions are true coins, otherwise they are counterfeit coins:
- Discrimination condition 2 Discriminant of adjacent color-changing ink characters Lab color feature values. According to the characteristic that the Lab color value fluctuation between adjacent characters of the normal color changing ink is small, if the fluctuation is large, the color change ink may be damaged characters, and the following logical discrimination is respectively performed, and all the conditions are true coins, otherwise it is a counterfeit currency:
- Judgment condition 3 single color change ink character area discrimination.
- the number of the pixel points satisfying the discrimination condition 1 is calculated for each character, Lab Area1, Lab Area0, and Lab Area00, and the following logical determination is performed, and all the conditions are true coins, otherwise it is a counterfeit currency:
- the banknotes satisfying the judgment condition 1, the judgment condition 2, and the judgment condition 3 are normal banknotes, otherwise the color-changing ink damages the counterfeit banknotes.
Landscapes
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Toxicology (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Facsimile Image Signal Circuits (AREA)
- Inspection Of Paper Currency And Valuable Securities (AREA)
- Color Image Communication Systems (AREA)
- Image Processing (AREA)
Abstract
一种基于Lab色彩空间的人民币变色油墨检测方法及装置。该方法包括获取待检测人民币的RGB图像(101);根据RGB图像识别待检测人民币的面值面向信息(102);根据面值面向信息定位出变色油墨字符(103);计算出每个变色油墨字符在红(R)、绿(G)、蓝(B)三个波段的灰度值Gray R、Gray G、Gray B(104);将每个变色油墨字符的Gray R、Gray G、Gray B转化为Lab色彩空间下的灰度值Gray a、Gray b(105);将待检测人民币的Lab色彩特征值与对应的预设条件进行一一判别,判断是否均满足预设条件(106),若至少一个不满足预设条件,则待检测人民币为假币(107)。Lab色彩特征值包括每个变色油墨字符的Gray a、Gray b、字符面积。
Description
本申请要求于2015年1月30日提交中国专利局、申请号为201510054056.X、发明名称为“一种基于Lab色彩空间的人民币变色油墨检测方法和装置”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。
本发明实施例涉及人民币变色油墨检测领域,具体涉及一种基于Lab色彩空间的人民币变色油墨检测方法和一种基于Lab色彩空间的人民币变色油墨检测装置。
目前国内大部分银行要求能准确识别出变色油墨区受到涂污、摩擦或拼接等类型的人民币,并当成伪钞处理。但是由于市场上变色油墨区域受损伪钞在特征上具有多样性、不确定性,比如钞票的纸质产生脱色、变色、变形;或钞票票面有印油污渍、墨水污渍;或钞票票面被涂写字迹等,其检测也一直是钞票识别设备的难点。
传统人民币检测设备针对变色油墨区域检测,主要是通过采集人民币图像信号,如单通道灰度图像或多光谱图像R、G、B三个通道灰度图像,统计出每个变色油墨字符在独立通道图像上的灰度均值分布、相邻字符之间灰度值差异性或字符面积,判断其是否在一定的阈值范围内,如果超出一定阈值范围则判为变色油墨假币。
现有技术中,对于变色油墨字符受到稍微涂污或磨损的情况下,变色油墨字符在灰度值上变化并不明显或处于设定的临界阈值附近,对变色油墨字符特征的描述仅通过从单通道图像或多光谱图像R、G、B三个通道的灰度值进行判断,难以正确识别变色油墨字符受到稍微涂污或磨损的钞票的真伪,降低了变色油墨受损钞票的检测效果。
发明内容
本发明实施例提供一种基于Lab色彩空间的人民币变色油墨检测方法和装置,用于对受到稍微涂污或磨损的变色油墨字符的颜色属性进行量化描述,提高变色油墨受损钞票的检测效果。
本发明实施例一种基于Lab色彩空间的人民币变色油墨检测方法的技术方案包括:
获取待检测人民币的RGB图像;
根据所述RGB图像识别所述待检测人民币的面值面向信息;
根据所述面值面向信息定位出变色油墨字符;
计算出每个变色油墨字符在红(R)、绿(G)、蓝(B)三个波段的灰度值Gray R、Gray G、Gray B;
将所述每个变色油墨字符的Gray R、Gray G、Gray B转化为Lab色彩空间下的灰度值Gray a、Gray b;
将所述待检测人民币的Lab色彩特征值与对应的预设条件进行一一判别,判断是否均满足所述预设条件,若否,则所述待检测人民币为假币,所述Lab色彩特征值包括每个变色油墨字符的Gray a、Gray b、字符面积。
优选的,
将所述待检测人民币的Lab色彩特征值与对应的预设条件进行一一判别包括:
单个变色油墨字符判别、相邻变色油墨字符判别、单个变色油墨字符面积判别。
优选的,
所述单个变色油墨字符判别包括:
判断所述每个变色油墨字符的Gray a与Gray b的大小、Gray a是否满足对应的预设条件,若否,则为假币。
优选的,
所述相邻变色油墨字符判别包括:
判断一个变色油墨字符的Gray a与另一个变色油墨字符的Gray a的大小、一个变色油墨字符的Gray b与另一个变色油墨字符Gray b的大小是否满足对应的预设条件,若否,则为假币。
优选的,
所述单个变色油墨字符面积判别包括:
判断每个变色油墨字符的字符面积是否满足对应的预设条件,若否,则为假币。
本发明实施例还提供一种基于Lab色彩空间的人民币变色油墨检测装置,技术方案包括:
获取模块,用于获取待检测人民币的RGB图像;
识别模块,用于识别所述获取模块获取的RGB图像中的面值面向信息;
定位模块,用于根据识别模块识别出的面值面向信息定位出变色油墨字符;
计算模块,用于计算出每个变色油墨字符在红(R)、绿(G)、蓝(B)三个波段的灰度值Gray R、Gray G、Gray B;
转化模块,用于将Gray R、Gray G、Gray B转化为Lab色彩空间下的灰度值Gray a、Gray b;
判断模块,将所述待检测人民币的Lab色彩特征值与对应的预设条件进行一一判别,判断是否均满足所述预设条件,所述Lab色彩特征值包括每个变色油墨字符的Gray a、Gray b、字符面积;
执行模块,若至少一个不满足所述预设条件,则所述待检测人民币为假币。
优选的,
所述判断模块包括单个变色油墨字符判别单元、相邻变色油墨字符判别单元、单个变色油墨字符面积判别单元。
优选的,
所述单个变色油墨字符判别单元包括:
第一判断子单元,判断所述每个变色油墨字符的Gray a与Gray b的大小、Gray a是否满足对应的预设条件;
第一执行子单元,若否,则为假币。
优选的,
所述相邻变色油墨字符判别单元包括:
第二判断子单元,判断一个变色油墨字符的Gray a与另一个变色油墨字符的Gray a的大小、一个变色油墨字符的Gray b与另一个变色油墨字符Gray b的大小是否满足对应的预设条件;
第二执行子单元,若否,则为假币。
优选的,
所述单个变色油墨字符面积判别单元包括:
第三判断子单元,判断每个变色油墨字符的字符面积是否满足对应的预设条件;
第三执行子单元,若否,则为假币。
本发明实施例的有益效果是:
通过获取的待检测人民币的RGB图像,识别出待检测人民币的面值面向信息,根据面值面向信息定位出变色油墨字符,计算出每个变色油墨字符在红、绿、蓝三个波段的灰度值Gray R、Gray G、Gray B,然后将计算出的Gray R、Gray G、Gray B转化为Lab色彩空间下的灰度值Gray a、Gray b,由于Lab色彩空间包括亮度分量L、从绿到红的a分量和从蓝到黄的b分量,所以Gray a、Gray b对变色油墨字符的颜色属性进行了量化描述,然后将Gray a、Gray b与对应的预设条件进行一一判别,判断是否均满足对应的预设条件,由于真币和假币的变色油墨的颜色属性差别大,通过对比变色油墨的颜色属性,对于变色油墨稍微涂污或摩擦的情况下,依然具有较强的检测效果,提高设备对正常真钞的接受率。
为了更清楚地说明本发明实施例或现有技术中的技术方案,下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图仅仅是本发明的实施例,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据提供的附图获得其他的附图。
图1是本发明实施例一种基于Lab色彩空间的人民币变色油墨检测方法的一个实施例流程示意图;
图2是本发明实施例一种基于Lab色彩空间的人民币变色油墨检测装置的结构示意图。
本发明实施例提供一种基于Lab色彩空间的人民币变色油墨检测方法和装置,用于对受到稍微涂污或磨损的变色油墨字符的颜色属性进行量化描述,提高变色油墨受损钞票的检测效果。
下面将结合本发明实施例中的附图,对本发明实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例仅仅是本发明一部分实施例,而不是全部的实施例。基于本发明中的实施例,本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例,都属于本发明保护的范围。
请参阅图1,本发明实施例一种基于Lab色彩空间的人民币变色油墨检测方法的一个实施例包括:
101、获取待检测人民币的RGB图像;
本实施例中,可以通过多光谱识别装置采集待检测人民币在RGB光谱下的图像,也可以通过其它方式获取RGB图像,此处具体不做限定。
102、根据RGB图像识别待检测人民币的面值面向信息;
本实施例中,待检测人民币的面值包括100、50、20等,面向可以分为四种:正面向上,正面向下,背面向上和背面向下。通过不同面额钞票图案特征的差异性实现待检测人民币的面值面向识别。
103、根据面值面向信息定位出变色油墨字符;
本实施例中,根据识别出的面值面向信息,定位出变色油墨字符所在的区域。
104、计算出每个变色油墨字符在红(R)、绿(G)、蓝(B)三个波段的灰度值Gray R、Gray G、Gray B;
本实施例中,对变色油墨字符进行分割,计算出每个变色油墨字符在红(R)、绿(G)、蓝(B)三个波段的灰度值Gray R、Gray G、Gray B,也就是计算出每个变色油墨字符在红(R)、绿(G)、蓝(B)三个通道的灰度值Gray R、Gray G、Gray B。
105、将每个变色油墨字符的Gray R、Gray G、Gray B转化为Lab色彩空间下的灰度值Gray a、Gray b;
本实施例中,将每个变色油墨字符的Gray R、Gray G、Gray B按照转化公式转化为Lab色彩空间下的灰度值Gray a、Gray b。
106、将待检测人民币的Lab色彩特征值与对应的预设条件进行一一判别,判断是否均满足预设条件,若否,则执行步骤107。
本实施例中,Lab色彩特征值包括每个变色油墨字符的Gray a、Gray b、字符面积;将待检测人民币的Lab色彩特征值与对应的预设条件进行一一判别,若至少一个不满足预设条件,则执行步骤107。
107、待检测人民币为假币。
本实施例中,一一判别是指待检测人民币的Lab色彩特征值是否满足对应的预设条件,若其中一个不满足,则待检测人民币为假币。
本实施例中,通过获取的待检测人民币的RGB图像,识别出待检测人民币的面值面向信息,根据面值面向信息定位出变色油墨字符,计算出每个变色油墨字符在红、绿、蓝三个波段的灰度值Gray R、Gray G、Gray B,然后将计算出的Gray R、Gray G、Gray B转化为Lab色彩空间下的灰度值Gray a、Gray b,由于Lab色彩空间包括亮度分量L、从绿到红的a分量和从蓝到黄的b分量,所以Gray a、Gray b对变色油墨字符的颜色属性进行了量化描述,然后将Gray a、Gray b与对应的预设条件进行一一判别,判断是否均满足对应的预设条件,由于真币和假币的变色油墨的颜色属性差
别大,通过对比变色油墨的颜色属性,对于变色油墨稍微涂污或摩擦的情况下,依然具有较强的检测效果,提高设备对正常真钞的接受率。
需要说明的是,将待检测人民币的Lab色彩特征值与对应的预设条件进行一一判别可以包括单个变色油墨字符判别、相邻变色油墨字符判别、单个变色油墨字符面积判别,下面分别进行说明:
(1)单个变色油墨字符判别
判断每个变色油墨字符的Gray a与Gray b的大小、Gray a是否满足对应的预设条件,若否,则为假币。
(2)相邻变色油墨字符判别
判断一个变色油墨字符的Gray a与另一个变色油墨字符的Gray a的大小、一个变色油墨字符的Gray b与另一个变色油墨字符Gray b的大小是否满足对应的预设条件,若否,则为假币。
(3)单个变色油墨字符面积判别
判断每个变色油墨字符的字符面积是否满足对应的预设条件,若否,则为假币。
采用上述方式进行一一判别,提高变色油墨受损钞票的检测效果的同时,提高了操作的灵活性。
请参阅图2,本发明实施例一种基于Lab色彩空间的人民币变色油墨检测装置的一个实施例包括:
获取模块201,用于获取待检测人民币的RGB图像;
识别模块202,用于识别获取模块201获取的RGB图像中的面值面向信息;
定位模块203,用于根据识别模块202识别出的面值面向信息定位出变色油墨字符;
计算模块204,用于计算出每个变色油墨字符在红(R)、绿(G)、蓝(B)三个波段的灰度值Gray R、Gray G、Gray B;
转化模块205,用于将Gray R、Gray G、Gray B转化为Lab色彩空间
下的灰度值Gray a、Gray b;
判断模块206,将待检测人民币的Lab色彩特征值与对应的预设条件进行一一判别,判断是否均满足预设条件,Lab色彩特征值包括每个变色油墨字符的Gray a、Gray b、字符面积。
执行模块207,若至少一个不满足预设条件,则待检测人民币为假币。
本实施例中,判断模块206可以进一步包括:单个变色油墨字符判别单元、相邻变色油墨字符判别单元、单个变色油墨字符面积判别单元。
单个变色油墨字符判别单元可以包括:
第一判断子单元,判断每个变色油墨字符的Gray a与Gray b的大小、Gray a是否满足对应的预设条件;
第一执行子单元,若否,则为假币。
相邻变色油墨字符判别单元可以包括:
第二判断子单元,判断一个变色油墨字符的Gray a与另一个变色油墨字符的Gray a的大小、一个变色油墨字符的Gray b与另一个变色油墨字符Gray b的大小是否满足对应的预设条件;
第二执行子单元,若否,则为假币。
单个变色油墨字符面积判别单元可以包括:
第三判断子单元,判断每个变色油墨字符的字符面积是否满足对应的预设条件;
第三执行子单元,若否,则为假币。
本实施例中,首先,获取模块201获取到待检测人民币的RGB图像,然后识别模块202识别出面值面向信息,定位模块203实现变色油墨字符的定位和分割,计算模块204和转化模块205统计出每个变色油墨字符的Lab色彩特征值,实现颜色信息的量化和特征统计,最后判断模块206对量化后的颜色值进行各项一一判别,执行模块207实现对变色油墨的检测,提高变色油墨受损钞票的检测效果;
其次,判断模块206可以采用多种方式一一判别,提高了操作的灵活
性。
下面以100元面值人民币的变色油墨为例进行详细说明。
Lab色彩空间是比较接近于人类视觉的一种颜色空间,其中L(Luminosity)表示亮度,a和b表示两种颜色的对立维度。其中L的值域由0到100,表示从黑到纯白,a表示从红色至绿色的范围,a的取值为+127至-128,b表示从蓝色至黄色的范围,b的取值为+127至-128,正为暖色、负为冷色。Lab色域宽阔很广,它不仅包含了RGB、CMYK的所有色域,还能表现它们不能表现的色彩。人的肉眼能感知的色彩,都能通过Lab模型表现出来。另外,Lab色彩模型还能弥补RGB色彩模型色彩分布不均的不足,因为RGB模型在蓝色到绿色之间的过渡色彩过多,而在绿色到红色之间又缺少黄色和其他色彩。
根据人眼感知的颜色,100元人民币变色油墨字符从正面观测颜色为绿色,根据Lab颜色空间模型,绿色的主分量在a轴的负方向上,a越趋近于负方向表示颜色越绿,并且|a|>|b|。
具体步骤包括:
(1)多光谱图像获取。利用多光谱识别装置获取100元面额钞票在RGB光谱下的图像。
(2)根据100元面额的RGB图像识别对应的面值面向信息;
(3)根据100元的面值面向信息定位出变色油墨字符所在区域,并对单个变色油墨字符进行分割;
(4)计算出每个变色油墨字符在红(R)、绿(G)、蓝(B)三个波段的灰度值Gray R、Gray G、Gray B;
(5)将每个变色油墨字符的Gray R、Gray G、Gray B转化为Lab色彩空间下的灰度值Gray a、Gray b;
转换公式为:
Gray L=0.2126*Gray R+0.7152*Gray G+0.0722*Gray B
Gray a=1.4749*(0.2213*Gray R-0.3390*Gray G+0.1177*Gray B)+128;
Gray b=0.6245*(0.1949*Gray R+0.6057*Gray G-0.8006*Gray B)+128;
*意思为乘积。
(6)统计Lab色彩特征值;
首先统计出“100”三个字符Lab均值分布,分别记为“1”(Gray L1,Gray a1,Gray b1)、“0”(Gray L0,Gray a0,Gray b0)、“0”(Gray L00,Gray a00,Gray b00);其次统计满足正常变色油墨Lab色彩分布的字符面积,分布记为“1”(Lab Area1)、“0”(Lab Area0)、“0”(Lab Area00)。
以正常100元人民币为例,统计三个变色油墨字符颜色,分别为:
“1”(Gray L1=100,Gray a1=-9,Gray b1=8)、
“0”(Gray L0=100,Gray a0=-9,Gray b0=8)、
“0”(Gray L00=100,Gray a00=-9,Gray b00=8);
满足Lab颜色特征的三个字符的面积分布为“1”(Lab Area1=44)、“0”(Lab Area0=80)、“0”(Lab Area00=80)。
当其中一个变色油墨字符“1”被涂污后,三个字符的Lab颜色变为:
“1”(Gray L1=98,Gray a1=-1,Gray b1=8)、
“0”(Gray L0=100,Gray a0=-9,Gray b0=8)、
“0”(Gray L00=100,Gray a00=-9,Gray b00=8);
满足Lab颜色特征的三个字符的面积分布为“1”(Lab Area1=10)、“0”(Lab Area0=80)、“0”(Lab Area00=80)。
可以看出Gray a1、Lab Area1两个特征值发生大的波动,而相对Gray L1、Gray L0、Gray L00三个值波动相对较小,因此不做判断。
(7)将待检测人民币的Lab色彩特征值与对应的预设条件进行一一判别。
根据对每个钞票统计出的变色油墨Lab色彩特征值和正常钞票的比较,实现对变色油墨区的判别。
判别条件1,单个变色油墨字符Lab色彩特征值判别。三个变色油墨字符分别进行以下逻辑判别,满足所有条件为真币,否则为假币:
|Gray a1|>|Gray b1|
|Gray a1-(-9)|<3
|Gray a0|>|Gray b0|
|Gray a0-(-9)|<3
|Gray a00|>|Gray b00|
|Gray a00-(-9)|<3
判别条件2、相邻变色油墨字符Lab色彩特征值判别。根据正常变色油墨相邻字符间的Lab色彩值波动较小的特征,如果波动较大可能为变色油墨受损字符,分别进行以下逻辑判别,满足所有条件为真币,否则为假币:
|Gray a1-Gray a0|<3
|Gray b1-Gray b0|<3
|Gray a1-Gray a00|<3
|Gray b1-Gray b00|<3
|Gray a0-Gray a00|<3
|Gray b0-Gray b00|<3
判别条件3、单个变色油墨字符面积判别。分别对每个字符统计满足判别条件1的像素点的个数Lab Area1、Lab Area0、Lab Area00,并进行以下逻辑判别,满足所有条件为真币,否则为假币:
|Lab Area1-44|<10
|Lab Area0-80|<10
|Lab Area00-80|<10
同时满足判决条件1、判决条件2、判决条件3的钞票为正常钞票,否则为变色油墨受损伪钞。
以上所述,以上实施例仅用以说明本发明的技术方案,而非对其限制;尽管参照前述实施例对本发明进行了详细的说明,本领域的普通技术人员应当理解:其依然可以对前述各实施例所记载的技术方案进行修改,或者对其中部分技术特征进行等同替换;而这些修改或者替换,并不使相应技
术方案的本质脱离本发明各实施例技术方案的精神和范围。
Claims (10)
- 一种基于Lab色彩空间的人民币变色油墨检测方法,其特征在于,包括:获取待检测人民币的RGB图像;根据所述RGB图像识别所述待检测人民币的面值面向信息;根据所述面值面向信息定位出变色油墨字符;计算出每个变色油墨字符在红(R)、绿(G)、蓝(B)三个波段的灰度值Gray R、Gray G、Gray B;将所述每个变色油墨字符的Gray R、Gray G、Gray B转化为Lab色彩空间下的灰度值Gray a、Gray b;将所述待检测人民币的Lab色彩特征值与对应的预设条件进行一一判别,判断是否均满足所述预设条件,若否,则所述待检测人民币为假币,所述Lab色彩特征值包括每个变色油墨字符的Gray a、Gray b、字符面积。
- 根据权利要求1所述的一种基于Lab色彩空间的人民币变色油墨检测方法,其特征在于,将所述待检测人民币的Lab色彩特征值与对应的预设条件进行一一判别包括:单个变色油墨字符判别、相邻变色油墨字符判别、单个变色油墨字符面积判别。
- 根据权利要求2所述的一种基于Lab色彩空间的人民币变色油墨检测方法,其特征在于,所述单个变色油墨字符判别包括:判断所述每个变色油墨字符的Gray a与Gray b的大小、Gray a是否满足对应的预设条件,若否,则为假币。
- 根据权利要求2所述的一种基于Lab色彩空间的人民币变色油墨检测方法,其特征在于,所述相邻变色油墨字符判别包括:判断一个变色油墨字符的Gray a与另一个变色油墨字符的Gray a的大小、一个变色油墨字符的Gray b与另一个变色油墨字符Gray b的大小是否满足对应的预设条件,若否,则为假币。
- 根据权利要求2所述的一种基于Lab色彩空间的人民币变色油墨检测方法,其特征在于,所述单个变色油墨字符面积判别包括:判断每个变色油墨字符的字符面积是否满足对应的预设条件,若否,则为假币。
- 一种基于Lab色彩空间的人民币变色油墨检测装置,其特征在于,包括:获取模块,用于获取待检测人民币的RGB图像;识别模块,用于识别所述获取模块获取的RGB图像中的面值面向信息;定位模块,用于根据识别模块识别出的面值面向信息定位出变色油墨字符;计算模块,用于计算出每个变色油墨字符在红(R)、绿(G)、蓝(B)三个波段的灰度值Gray R、Gray G、Gray B;转化模块,用于将Gray R、Gray G、Gray B转化为Lab色彩空间下的灰度值Gray a、Gray b;判断模块,将所述待检测人民币的Lab色彩特征值与对应的预设条件进行一一判别,判断是否均满足所述预设条件,所述Lab色彩特征值包括每个变色油墨字符的Gray a、Gray b、字符面积;执行模块,若至少一个不满足所述预设条件,则所述待检测人民币为假币。
- 根据权利要求6所述的一种基于Lab色彩空间的人民币变色油墨检测装置,其特征在于,所述判断模块包括单个变色油墨字符判别单元、相邻变色油墨字符判别单元、单个变色油墨字符面积判别单元。
- 根据权利要求7所述的一种基于Lab色彩空间的人民币变色油墨检测装置,其特征在于,所述单个变色油墨字符判别单元包括:第一判断子单元,判断所述每个变色油墨字符的Gray a与Gray b的大小、Gray a是否满足对应的预设条件;第一执行子单元,若否,则为假币。
- 根据权利要求7所述的一种基于Lab色彩空间的人民币变色油墨检测 装置,其特征在于,所述相邻变色油墨字符判别单元包括:第二判断子单元,判断一个变色油墨字符的Gray a与另一个变色油墨字符的Gray a的大小、一个变色油墨字符的Gray b与另一个变色油墨字符Gray b的大小是否满足对应的预设条件;第二执行子单元,若否,则为假币。
- 根据权利要求7所述的一种基于Lab色彩空间的人民币变色油墨检测装置,其特征在于,所述单个变色油墨字符面积判别单元包括:第三判断子单元,判断每个变色油墨字符的字符面积是否满足对应的预设条件;第三执行子单元,若否,则为假币。
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510054056.X | 2015-01-30 | ||
| CN201510054056.XA CN104574642B (zh) | 2015-01-30 | 2015-01-30 | 一种基于Lab色彩空间的人民币变色油墨检测方法和装置 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2016119394A1 true WO2016119394A1 (zh) | 2016-08-04 |
Family
ID=53090596
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2015/082616 Ceased WO2016119394A1 (zh) | 2015-01-30 | 2015-06-29 | 一种基于Lab色彩空间的人民币变色油墨检测方法和装置 |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN104574642B (zh) |
| WO (1) | WO2016119394A1 (zh) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110570209A (zh) * | 2019-07-30 | 2019-12-13 | 平安科技(深圳)有限公司 | 证件真伪验证方法、装置、计算机设备及存储介质 |
| CN119049170A (zh) * | 2024-10-30 | 2024-11-29 | 日照财金路演有限公司 | 基于大数据的纸币异常检测方法及系统 |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104574642B (zh) * | 2015-01-30 | 2017-04-26 | 广州广电运通金融电子股份有限公司 | 一种基于Lab色彩空间的人民币变色油墨检测方法和装置 |
| CN104952143A (zh) * | 2015-06-18 | 2015-09-30 | 辰通智能设备(深圳)有限公司 | 一种票据图像红水线变色检测方法及系统 |
| CN105069899B (zh) * | 2015-08-03 | 2017-08-29 | 深圳怡化电脑股份有限公司 | 一种验钞的方法及验钞装置 |
| CN106952391B (zh) * | 2017-02-27 | 2019-06-07 | 深圳怡化电脑股份有限公司 | 一种污损纸币识别方法及装置 |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004265036A (ja) * | 2003-02-28 | 2004-09-24 | Nippon Conlux Co Ltd | 紙葉類識別装置および方法 |
| CN101901511A (zh) * | 2009-05-27 | 2010-12-01 | 株式会社东芝 | 纸张确定设备 |
| CN103155008A (zh) * | 2010-10-08 | 2013-06-12 | 德国捷德有限公司 | 用于检查值文件的光学安全特征的方法 |
| CN104200561A (zh) * | 2014-06-16 | 2014-12-10 | 华中科技大学 | 一种基于纹理特征识别具有字迹污损人民币的方法 |
| CN104574642A (zh) * | 2015-01-30 | 2015-04-29 | 广州广电运通金融电子股份有限公司 | 一种基于Lab色彩空间的人民币变色油墨检测方法和装置 |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP4003428B2 (ja) * | 2001-10-10 | 2007-11-07 | セイコーエプソン株式会社 | 小切手等処理装置及び小切手等処理方法 |
| DE10314071B3 (de) * | 2003-03-28 | 2004-09-30 | Koenig & Bauer Ag | Verfahren zur qualitativen Beurteilung eines Materials mit mindestens einem Erkennungsmerkmal |
| CN101751713A (zh) * | 2008-12-03 | 2010-06-23 | 北京金储自动化技术研究所 | 一种基于光谱影像分析的银行票据鉴伪系统 |
| CN101908241B (zh) * | 2010-08-03 | 2012-05-16 | 广州广电运通金融电子股份有限公司 | 有价文件识别方法及其识别系统 |
-
2015
- 2015-01-30 CN CN201510054056.XA patent/CN104574642B/zh active Active
- 2015-06-29 WO PCT/CN2015/082616 patent/WO2016119394A1/zh not_active Ceased
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004265036A (ja) * | 2003-02-28 | 2004-09-24 | Nippon Conlux Co Ltd | 紙葉類識別装置および方法 |
| CN101901511A (zh) * | 2009-05-27 | 2010-12-01 | 株式会社东芝 | 纸张确定设备 |
| CN103155008A (zh) * | 2010-10-08 | 2013-06-12 | 德国捷德有限公司 | 用于检查值文件的光学安全特征的方法 |
| CN104200561A (zh) * | 2014-06-16 | 2014-12-10 | 华中科技大学 | 一种基于纹理特征识别具有字迹污损人民币的方法 |
| CN104574642A (zh) * | 2015-01-30 | 2015-04-29 | 广州广电运通金融电子股份有限公司 | 一种基于Lab色彩空间的人民币变色油墨检测方法和装置 |
Non-Patent Citations (1)
| Title |
|---|
| JIN, YE ET AL.: "A Defect Detection Algorithm for Paper Currency Sorting System", ACTA AUTOMATICA SINICA, vol. 33, no. 9, 30 September 2007 (2007-09-30), pages 972 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110570209A (zh) * | 2019-07-30 | 2019-12-13 | 平安科技(深圳)有限公司 | 证件真伪验证方法、装置、计算机设备及存储介质 |
| CN119049170A (zh) * | 2024-10-30 | 2024-11-29 | 日照财金路演有限公司 | 基于大数据的纸币异常检测方法及系统 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104574642B (zh) | 2017-04-26 |
| CN104574642A (zh) | 2015-04-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2016119394A1 (zh) | 一种基于Lab色彩空间的人民币变色油墨检测方法和装置 | |
| WO2016115829A1 (zh) | 一种基于Lab色彩空间的钞票分类鉴别方法和装置 | |
| CN105243731B (zh) | 一种光谱自适应的多国纸币图像识别装置、系统及方法 | |
| CN105894655B (zh) | 基于rgb-d相机的复杂环境下纸币检测和识别方法 | |
| CN107730493A (zh) | 产品颜色差异检测方法、装置、介质以及计算机设备 | |
| WO2019072249A1 (zh) | 利用皮肤图像分析皮肤水分的方法及装置 | |
| CN111986378B (zh) | 一种票据彩色纤维丝检测方法和系统 | |
| KR100881230B1 (ko) | 스테레오 영상을 이용한 고 정밀 위폐 감별 시스템 | |
| CN101882223B (zh) | 人体肤色的测评方法 | |
| CN107437293A (zh) | 一种基于票据全局特征的票据防伪鉴别方法 | |
| CN106934915A (zh) | 纸币处理装置 | |
| CN108596196A (zh) | 一种基于绝缘子图像特征词典的污秽状态评估方法 | |
| CN104464078A (zh) | 通过光变油墨识别损伤钞的方法及系统 | |
| CN117132544A (zh) | 一种基于深度学习和颜色空间变换的图像偏色检测方法 | |
| WO2017045403A1 (zh) | 一种票据重张检测方法和装置 | |
| CN104732644B (zh) | 纸币鉴别的质量控制方法及其系统 | |
| CN106875543B (zh) | 一种基于rgb-d相机的视障人士纸币识别系统及识别方法 | |
| Pawar et al. | Recognition of Indian currency note based on HSV parameters | |
| CN106600812A (zh) | 一种纸币识别方法及装置 | |
| CN108537856A (zh) | 一种色觉分类检测图的制作方法 | |
| TWI586954B (zh) | 檢測細胞受人類乳突病毒(hpv)感染的裝置及其檢測方法 | |
| CN115761013A (zh) | 一种基于有无纹理分类的布匹色差检测方法 | |
| TW200834476A (en) | A fire image detection method | |
| TW202208826A (zh) | 以連續角度色彩檢測印刷油墨反光現象之方法 | |
| AU2016100492A4 (en) | A method and system for identifying and measuring a defect that reduces transparency in a substrate for a security document |
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: 15879597 Country of ref document: EP Kind code of ref document: A1 |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 15879597 Country of ref document: EP Kind code of ref document: A1 |