JPH1069559A - Paper money identifying method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To accurately identify paper sheets at high speed without being affected by dirt or distortion by discriminating the truth/false of a paper money on three stages of low-accuracy identifying processing, middle-accuracy identifying processing and highaccuracy identifying processing. SOLUTION: Multilevel gradation data based on a photosensor for paper moneys are fetched (S1). Matching with a reference waveform is executed, and its denomination and the direction of insertion are discriminated (S2). Next, low-accuracy identifying processing is performed for calculating difference in the luminance of image data provided by being irradiated with two beams of different wavelengths (S3). This calculated value is compared with a predetermined threshold value so that the truth / false can be discriminated (S4). Concerning the paper money to be identified discriminated as truth, middle-accuracy identifying processing such as conveyance deviation correcting processing is performed (S5). In this case, concerning the paper money to be identified discriminated as truth (S6), highaccuracy identifying processing is performed based on the discrimination of comparison between matching degree calculating processing and the threshold value of a calculated matching degree (S7) and truth/false is discriminated (S8).

Description

DETAILED DESCRIPTION OF THE INVENTION

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for identifying paper sheets such as banknotes and securities, and more particularly to a method for identifying highly accurate and high-speed judgments by suppressing the influence of various stains on the identified banknotes. About possible identification methods.

2. Description of the Related Art Japanese Patent Laid-Open No.
-215293. In this publication, a bill is divided into a plurality of zones, detection data for each zone is compared with reference data predetermined for each zone, and the bill is identified based on a comparison result in each zone. In the bill discriminating method, a plurality of zones are set in accordance with the front and back of the bill, the orientation, and the displacement at the time of discrimination, and the data of each zone is totaled for one bill, and a ratio value to the total value is calculated. Is stored as reference pattern data, a total value of the detection data is obtained, and a ratio value to the total value is calculated as detection pattern data, and the detection pattern data is within an allowable value range of the reference pattern data. It is determined whether or not the absolute value of the difference between the reference pattern data and the detected pattern data is calculated for each zone and the total is calculated. Bill identifying how total value of the distance calculation to determine whether less than the allowable value and performing the bill validator is disclosed.

However, in the above-mentioned conventional technology, there is a problem that a fake bill is erroneously recognized as a genuine bill because variations in identification due to dirt, distortion and other reasons are allowed. Yes, this is a cause of a decrease in identification accuracy. The present invention has been made in order to solve the problems caused by the conventional method, and a method for accurately and rapidly identifying paper sheets without being affected by dirt, distortion, and the like. The purpose is to provide.

According to the method of the present invention, there is provided a coarse-precision discriminating step of discriminating a genuine bill or a fake bill from images of bills to be discriminated obtained by a plurality of light sources having different wavelengths. A precision determination step of performing processing such as transport deviation and level adjustment on image data of the bill to be identified determined as a genuine bill, predicting an identification error, and using this prediction error to determine whether the bill is genuine or fake; And a high-precision determination step of performing a matching process using a mask on the image data determined to be serious in the fine determination step, and determining a genuine note or a fake note based on the obtained matching degree. Also,
A step of providing a plurality of the masks and repeating the high-accuracy determination step using each of the masks a plurality of times to determine a genuine bill or a fake bill is used.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a perspective view showing a bill discriminating method according to an embodiment of the present invention; FIG. 1 is a flowchart showing the basic algorithm of the present embodiment. In the figure, when the program is started by inserting a bill, multi-value density data by the optical sensor of the bill is taken in (step S1). A part of the captured grayscale data is subjected to matching with three denominations and reference waveforms in four directions, and the denomination and the input direction are determined using the matching degree obtained as a result.
(Step S2). Using the determination result obtained in this way, the grayscale data is converted into data in the front direction and in the erect direction. Next, a coarse accuracy discrimination process in step S3 is performed. This processing irradiates the bill to be identified with light from two light sources having different wavelengths, obtains image data, calculates the difference or ratio of the luminance of these image data, and calculates the calculated value and a predetermined threshold value. This is a process of determining a genuine bill or a counterfeit bill from the comparison with the above. The banknote to be identified that has been roughly determined to be a counterfeit note in step S4 through the above processing is determined as a counterfeit note as it is. On the other hand, the bill to be identified determined as a genuine note in step S4 is
5 enters a medium precision identification processing routine. In this routine, discrimination processing and the like are performed for conveyance deviation correction processing, level adjustment processing, irregular component extraction processing, irregular component prediction processing, prediction error calculation processing, and prediction error threshold determination. This result is passed on to the fineness determination process in step S6, where the authenticity of the bill to be identified previously determined to be genuine is determined. If it is determined to be a counterfeit note, this is determined, and if it is determined to be a genuine note, the process proceeds to a high-precision identification processing routine in step S7. In this routine, matching degree calculation processing using a mask,
High-precision discrimination and the like are performed by comparison with the calculated matching degree and a threshold value. This result is passed on to the high-precision determination processing of step S8, where the authenticity of the identified bill previously determined to be genuine is determined. If it is determined to be a counterfeit note, this is determined, and if it is determined to be genuine, this is determined. Next, details of the flowchart of FIG. 1 will be described. The determination processing for three denominations and four directions is performed, for example, as follows. That is, as shown in the flowchart of FIG. 2 and the conceptual diagram of FIG. 3, when a bill to be identified is inserted into a bill validator (not shown) in step S1, an LED (two light emitting diodes having different wavelengths) and a light receiving element are received. The image data of the bill surface is obtained by the image sensor (line sensor) comprising two waveform data in which the vertical axis represents the luminance value and the horizontal axis represents the position information (point) (FIG. 3a).
reference. Since the number of data points used in the calculation in the denomination / direction determination is only used for the denomination and direction determination, it does not need to be all the points of the bill, but the minimum number of points required for the denomination and direction determination Is fine. In step S21,
The waveform data (banknote data) obtained by the light of wavelength 1 in the obtained waveform data is three denominations and four directions (one denomination and four directions in FIG. 3) previously obtained from the genuine bill by the image sensor. In each case, the data is compared with the direction-specific reference waveform data of the front right A, the front left B, the back right C, and the back left D (step S2).
2) The minimum sum of squares of the difference for each point is calculated (see FIG. 3B). Next, the obtained sums of squared differences (matching degrees) in the respective directions are compared, and in step S23, the direction of the direction-specific reference waveform data having the minimum value is determined as the bill insertion direction (see FIG. 3C). The coarse-precision identification processing in step S3 calculates a difference or a ratio of a luminance value between image data obtained using light of wavelength 2 and image data obtained using light of wavelength 1 as the image sensor, This is a process of comparing the calculated value with a predetermined threshold value and roughly determining a genuine bill or a fake bill (step S4). Further, the medium-precision identification processing in step S5 is roughly divided into a processing for correcting a transport deviation of the bill to be identified and a true / false identification processing. In the transport deviation correcting process, a deviation width K (for calculation) is set in step S51 as shown in FIG. This shift width K is a value between a minimum shift width value at which a transport shift can occur and a maximum shift width, and starts from the minimum value. Then, in step S52, data shifted by K, which is the set input signal of the bill conveyed, is created. Next, in step S53, data at a plurality of target locations is extracted from the data shifted by K created in step S52. In step S54, the absolute value sum of the difference between the extracted data and the corresponding data of the reference waveform is calculated.
If the total value obtained in step S55 is the minimum, the value of K at that time is temporarily recorded as the calculated deviation width. The above operation is repeated from the minimum value to the maximum value of K.
By doing so, every time the value of K changes between the minimum value and the maximum value, the cumulative value of the difference from the reference waveform is obtained, and the minimum cumulative value of K is updated each time. Go. Then, by setting the finally remaining value of K as a target conveyance deviation width, the conveyance deviation value can be accurately obtained. If more accuracy is required, instead of calculating the absolute sum of the differences in step S54,
It is also possible to calculate the sum of squares of the difference (step S541), and determine the one with the smallest value as the target deviation width. The extraction of data at a plurality of locations in step S53 is performed by a method as shown in FIG. That is, in step S531, a basic representative waveform is created from a banknote (closed note) free from dirt and tear. In step S532, a displacement representative value K is set and a displaced basic representative waveform is created in the same manner as in the case of calculating the transport displacement width. Step S5
At 33, the difference value (absolute value) of each position between the shifted basic representative waveform and the original basic representative waveform is recorded. In the above operation, the value of K is changed from the minimum value to the maximum value, and the difference value is sequentially recorded, and in step S534, the minimum value of the difference value at each position is calculated. Finally, a plurality of points are selected in ascending order of the smallest difference value obtained in step S535, and this point is set as a plurality of points from which data should be extracted. In the right half of FIG. 5, a conceptual diagram of a waveform in the middle is shown, and in the left half, a conceptual diagram of a step of selecting a plurality of locations is shown. After the transport deviation correcting process, the level of the captured image of the banknote to be identified is adjusted, and the process proceeds to a process of reducing the identification error due to an irregular component such as a dirt component from the image. This processing is performed using, for example, an auto (AR) regression model. That is, as shown in FIGS. 6 to 8, the present processing method is largely divided into a pre-processing and a processing at the time of bill insertion, and FIGS. 7 and 8 are flowcharts respectively. The pre-processing is processing that is created by learning the fluctuation component estimation model of the sensor input data of the banknote, and as shown in FIG.
At 51, a plurality of genuine bills (new bills) are sensed to obtain sensor signals of brightness and density of images and characters on bills, and a reference waveform (for example, an average value data waveform) as reference data is obtained from each sensor signal. Get. Next, in step S152, a fluctuating component such as dirt or distortion is extracted for each genuine note from the genuine note data using the reference waveform. And step S153
The data of the fluctuation component extracted in is regarded as a periodic time-series signal, and the equation as an autoregressive model is used.

(Equation 1) The coefficients a1, a2, of the equation for contamination at a certain time represented by
... Means to find ap. The learning data in this case is comparable to the time-series signal data of the dirt component when several banknotes are arranged and input. The periodic time series signal of the dirt component created in this way is learned by an autoregressive analysis method in step S4, and as a result of the learning, an estimation model of a dirt variation component of one banknote is created. After the pre-processing is performed in this manner, the flow proceeds to a processing at the time of insertion for performing a true / false determination when a bill is actually inserted. In the bill insertion process, first, a sensor signal from the bill inserted in step 251 is input. Next, in step S252, a difference between the input signal and the reference waveform is obtained to extract a dirt and distortion component as a variation component. In step S253, based on the data of the dirt and distortion components obtained in step S252, dirt estimation using the estimation model is performed by an auto-regression model technique, and a predicted value is calculated. Here, the estimation method will be described. Since the estimation model is obtained by the pre-processing, the dirt component predicted from the estimation model of the autoregressive analysis is calculated based on Equation 1 and the input dirt component data of the banknote. It is calculated (step S253). The waveform of dirt as a fluctuation component of the input bill thus obtained,
From the fluctuation component waveform of the estimation model, the prediction error is calculated in step 254. From this result, if the input bill is within the predetermined prediction error range, it is determined that the bill is genuine. Is determined to be a counterfeit note (step S25).
5). Of the banknotes subjected to the medium-precision identification processing and the fine determination in this way, those that are determined to be counterfeit are determined to be counterfeit, and those that are determined to be genuine are further subjected to high-precision identification processing. move on. In the high-precision identification processing, a reference waveform obtained from a genuine note and a waveform of a bill to be identified are sequentially matched for each point, and a mask is prepared when calculating a matching degree (sum of squared differences). This is a process for performing highly accurate true / false identification of a specific portion (a portion that is a feature of bill identification) of the bill to be identified. In this case, as shown in FIGS. 9 and 10, a plurality of masks for matching an arbitrary specific portion are prepared, and a multi-stage high-precision identification process is performed by using these masks in stages. (Step S
71) is desirable. Then, for the banknote data to be identified which is determined to be genuine at a certain stage, a matching degree is calculated by a mask at the next stage, and the calculation is sequentially advanced (steps S711, S712, S811, S812). The determination of the authenticity by the processing here may be based on a comparison between the value of the matching degree and a predetermined strict threshold value.

As described above, according to the present invention, the authenticity of a banknote is determined in three stages of the coarse-precision identification processing, the medium-precision identification processing, and the high-precision identification processing. Can be expected at a high speed, a medium-precision judgment can be used to identify a counterfeit note having a medium difficulty, and a high-precision judgment can be used to accurately identify a counterfeit note having a high identification difficulty.

[Brief description of the drawings]

FIG. 1 is a flowchart showing an embodiment of a bill discriminating method according to the present invention.

FIG. 2 is a flowchart of a denomination / direction determination process.

FIG. 3 is a conceptual diagram of a denomination / direction determination process.

FIG. 4 is a flowchart illustrating a conveyance deviation width calculation method.

FIG. 5 is a flowchart illustrating a method of selecting data at a plurality of locations from input data.

FIG. 6 is a conceptual diagram of a prediction process using an irregular component such as a stain.

FIG. 7 is a flowchart of a pre-processing.

FIG. 8 is a flowchart of a bill insertion process.

FIG. 9 is a flowchart showing another embodiment of the bill discriminating method of the present invention.

FIG. 10 is a flowchart of a multi-stage high-precision determination and a conceptual diagram of a mask.

Claims (2)

[Claims] 1. A coarse-precision identification step of determining a genuine bill or a counterfeit bill from images of bills to be identified obtained by a plurality of light sources having different wavelengths; A fine determination step of performing processing such as conveyance deviation and level adjustment on the image data of the identification bill, predicting the identification error, and using the prediction error to determine a genuine bill or a fake bill, and a true bill in the fine determination step. And a high-precision determining step of performing a matching process using a mask on the image data determined to be true, and determining a genuine note or a fake note based on the obtained matching degree. 2. A bill discriminating method, wherein a plurality of the masks are provided, and a high-precision determination step is repeated a plurality of times using each of the masks to determine a genuine bill or a fake bill.