JP3674289B2 - Image acquisition method and apparatus for paper sheet media - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an image acquisition method and apparatus for a paper sheet medium, and more particularly to a paper sheet discrimination apparatus such as a banknote for discrimination of an image including a pattern characteristic of a paper sheet medium printed using various colors. The present invention relates to an image acquisition method and apparatus for paper sheet media.
[0002]
[Prior art]
Discrimination of bills and the like is performed by extracting features of a pattern printed on bills and comparing the features with features of genuine bills. For that purpose, a banknote is imaged by an image sensor, and the obtained image data is subjected to image processing. However, the banknote is printed by using various colors to represent features. An image sensor detects the image of a banknote by irradiating light to the conveyed banknote, and detecting the reflected light or transmitted light. A light emitting diode is generally used as a light source for illuminating a banknote, and an image of the banknote illuminated by the light emitting diode is detected by, for example, a CCD (charge coupled device) line sensor.
[0003]
[Problems to be solved by the invention]
However, as the light source, an element that emits light of a color that is not used for printing is used. When the color of ink to be printed becomes diversified, such a light emitting element can obtain a pattern feature as an image. There are cases where it is not possible. For example, when a green light emitting diode is used as a light source, a green pattern cannot be detected. In addition, in some countries, there are some banknotes in which a part of the medium is transparent, and in addition to the fact that multiple colors of ink are used, the characteristics of the banknotes are always expressed using a single light source. There is a problem that it is becoming impossible to detect all images.
[0004]
The present invention has been made in view of the above points, and can reliably detect the characteristic pattern of paper sheets such as banknotes including multicolored prints and transparent portions. An object of the present invention is to provide an image acquisition method and apparatus for a paper sheet medium.
[0005]
[Means for Solving the Problems]
In the present invention, in order to solve the above-described problem, in the image acquisition method of a paper sheet medium for acquiring a discrimination image of the paper sheet medium, the paper sheet medium is switched while switching a plurality of light sources having different wavelengths for each scan line. A plurality of separated images are generated by scanning and sorting the data read for each scan line in synchronization with the switching of the light source, and the separated images are logically combined to generate a single logical combined image. A skew angle is detected for the composite image, a medium area coordinate is obtained for the logical composite image, a skew correction origin is calculated for the logical composite image, and the calculated skew correction origin is the center. An image acquisition method for a paper sheet medium is provided, wherein skew correction is performed on the medium area of each separated image.
[0006]
According to such an image acquisition method for a paper sheet medium, a plurality of images can be simultaneously obtained from a region of one paper sheet medium by scanning the paper sheet medium by switching the light source for each scan line. A plurality of images can be obtained by rearranging each scan line. These images are obtained by scanning almost the same region with a pixel unit shift. Since these are images of the same area, these images can be overlapped to form a single image, which makes it clearer even if the area where the medium exists is unclear. The skew angle common to all images is detected and the skew correction origin is calculated. Using this common skew correction origin, the skew correction of each separated image is performed.
[0007]
Further, according to the present invention, in the image acquisition apparatus for a paper sheet medium for acquiring a discrimination image of the paper sheet medium, a light emitting unit in which a combination of a plurality of light sources having different emission wavelengths is linearly arranged, and the light emitting unit A light receiving unit that sequentially scans and reads the paper sheet medium illuminated by the light source, a light source switching unit that switches and controls a light source of the light emitting unit each time the paper sheet medium is scanned, and a light output from the light receiving unit Analog / digital conversion means for digitizing pixel signals, image separation means for distributing the output from the analog / digital conversion means to image data for each type of the light source in synchronization with the light source switching means, and the image separation means The separated image data storage means for storing the image data separated in step (a) and the plurality of separated image data are logically synthesized into one image. Logically synthesized image data storage means for storing the logically synthesized image data, generation of the logical image data, calculation of the skew angle, medium area coordinates, and skew correction origin from the logically synthesized image data, There is provided an image acquisition apparatus for a paper sheet medium, comprising: a program storage unit that stores a skew correction program that performs skew correction; and a calculation unit that executes the skew correction program. .
[0008]
According to the image acquisition apparatus for a paper sheet medium, a light emitting unit in which a plurality of light sources are integrated is used, and the light source switching unit switches the light source every time the paper sheet medium is scanned to emit light. The light receiving means simultaneously obtains the same image when the paper sheet medium is scanned a plurality of times. The image is converted by the analog / digital converting means, and then separated into individual image data by the image separating means. Stored in the data storage means. Next, the separated image data is superposed into one logical image data by the skew correction program. This is possible because a plurality of images are acquired simultaneously, so that they have a common coordinate value. Thereafter, the skew correction program detects the skew angle common to all images and calculates the skew correction origin from the logically synthesized image data, and uses this common skew correction origin to separate each separation. By performing the skew correction of the image data, it is possible to obtain a discrimination image of the paper sheet medium. As a result, only the types of light sources as the light emitting means are increased, and other configurations such as the light receiving means of the image detection unit are the same as in the case of the monochromatic light source, and the installation space is also the same.
[0009]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings, taking as an example a case where the present invention is applied to a banknote discrimination device.
[0010]
FIG. 1 is a schematic diagram showing a configuration example of an image acquisition device in a banknote discrimination device. In FIG. 1, this image acquisition device includes a light emission timing control unit 1, an LED (light emitting diode) array 2, a selfoc lens 3, a CCD image line sensor 4, and an A / D (analog / digital) conversion unit. 5, an image separation unit 6, a central processing unit (CPU) 7, a main memory 8, a work memory 9, a read only memory (ROM) 10, and a bus 11.
[0011]
The LED array 2 is a composite light source diode array in which a green light emitting diode, a red light emitting diode, and an infrared light emitting diode are integrally formed, and the CCD image sensor 4 is configured to detect light transmitted through the banknote 12. ing. When a signal indicating that the banknote 12 has been transported to the image acquisition device is input from the banknote transport unit, the light emission timing control unit 1 sequentially switches the three types of light emitting diodes of the LED array 2 to emit light, and transports them. The banknote 12 that has been used is repeatedly illuminated with different light sources. As a result, the CCD image line sensor 4 scans the banknote 12 with a different color each time, and the image obtained by the green light emitting diode, the image obtained by the red light emitting diode, and the infrared light emitting diode are obtained. The image can be obtained with a single pass of the banknote.
[0012]
Pixel data obtained for each scan is input to the A / D converter 5 where it is converted into digital pixel values. The pixel value is sent to the work memory 9 by adding a code for identifying which light source is used for each scan line in the image separation unit 6. In the work memory 9, three images of a green image (G) 13, a red image (G) 14, and an infrared image (IR) 15 are obtained by sorting based on the light source identification code. . Since these images 13, 14 and 15 are originally obtained by separating one image, they have common coordinates.
[0013]
The banknote 12 is conveyed across the LED array 2 and the CCD image line sensor 4 arranged in the middle of the conveyance path. Since the conveyance width is matched with the banknote width of the largest denomination among banknotes, there is always a margin in the width, and the banknotes are transported in a skewed state slightly inclined with respect to the transport direction. At the time of banknote discrimination, if the acquired image remains in the skew state, the possibility of misclassification increases. Therefore, the images 13, 14, and 15 obtained individually must be skew-corrected. This skew correction is performed by rotating the image by the skew amount. For this purpose, it is necessary to obtain the skew correction origin and rotate the three images 13, 14, and 15 in the same manner. In order to obtain the skew origin, it is necessary to extract the outline of the banknote. In particular, the outline of the infrared image 15 cannot be clearly obtained. However, since the images 13, 14, and 15 have a common origin coordinate, unclear outlines are complemented by superimposing these images.
[0014]
Therefore, when the scan of one bill is completed, the CPU 7 causes the image synthesis program 17 stored in the ROM 10 to be executed on the main memory 8 so that the three images 13, 14, 15 generated in the work memory 9 are displayed. An OR operation is performed to generate one image 16. Subsequently, with respect to the image 16, the skew angle detection program 18 detects the skew angle, the banknote area coordinate acquisition program 19 obtains the coordinates of the four corners of the banknote, and the skew correction origin calculation program 20 performs a rotation process. The skew correction origin for this is obtained. Based on the skew correction origin, the skew correction program 21 rotates the images 13, 14, and 15 by coordinate conversion to perform skew correction. The three images 13, 14, and 15 that have been skew-corrected in this way are output and used for a device that performs bill direction, denomination identification, and authenticity determination.
[0015]
FIG. 2 is a diagram showing a schematic structure of the LED array. The LED array 2 is configured by arranging a set of a green light emitting diode 31, a red light emitting diode 32, and an infrared light emitting diode 33 on one substrate 30 in one direction. Preferably, in order to eliminate unevenness in the amount of light when the banknote 12 is irradiated, the green light-emitting diode 31, the red light-emitting diode 32, and the infrared light-emitting diode 33 are arranged in two or more rows, and light is emitted in adjacent rows. In addition to staggered arrangement so that the colors are different, the inner wall of the housing for housing the substrate 30 and the light emitting diode mounted thereon is made highly reflective, and a highly light diffusing member is arranged on the light emitting surface It is better to have the structure.
[0016]
FIGS. 3A and 3B are diagrams for explaining separation of the read image, in which FIG. 3A shows a state of scanning of banknotes, and FIG. 3B shows a separation image. When the banknote 12 moves along the transport direction, the CCD line sensor 4 scans in a direction perpendicular to the transport direction of the banknote 12. Since the light emission color of the LED array 2 is switched for each scan line, the CCD line sensor 4 scans the reading area 40 including the banknote 12 in the order of green, red, and infrared. As a result, the CCD line sensor 4 reads G1, R1, IR1, then G2, R2, IR2,. In synchronization with the switching timing of the emission color of the LED array 2, an image 13 is generated by extracting and assembling data for only the green scan line, and an image 14 is generated by extracting and assembling only the data for the red scan line. Is generated, and the image 15 is generated by extracting and assembling data of only the infrared scan line.
[0017]
In the images 13, 14, and 15 generated in this manner, the skew direction of the banknote 12 is detected, and each image of the skewed banknote area is finally corrected. The processing will be described.
[0018]
First, the pixel values at the corresponding coordinate positions of the images 13, 14, and 15 are ORed to be combined into one image 16. The skew angle is detected based on the synthesized image 16.
[0019]
The skew angle is detected by checking a pixel value for each scan line, and the pixel value is a light-transmitting pixel having a very large pixel value that is directly received without passing through the banknote, or is received through the banknote. It is determined whether the pixel value is a small non-transparent pixel, and the arrangement of the non-translucent pixels is found therefrom, and it is determined that the banknote area is crossed by the scan line. This is obtained for the entire image 16 synthesized, and its distribution is examined.
[0020]
By searching for the most frequent short length from the short length distribution, the scan lines crossing the long sides of both sides of the banknote can be found, and the short-length coordinates that appear first on each scan line are connected in the order of scan. Is used to obtain a line representing the skew angle, from which the skew angle is obtained.
[0021]
Next, in order to determine the skew correction target, the bill area is calculated, and the coordinates of the four corners of the bill (that is, the outermost contour of the bill) are obtained. This is done by searching for a scan line having a short length of 1 pixel or more, and setting the short-length coordinate position of the scan line with the smallest number to one end point of the short side of the upper banknote and the shortest of the scan line with the largest number. The coordinate of the long side is one end of the short side of the banknote, and the coordinate of the position where the short side of the scan line with the smallest number among the most frequent short scan lines ends or starts is the upper side of the banknote The coordinates of the position where the short length of the scan line with the largest number in the scan line of the most frequent short length starts or ends is obtained as the other end point of the short side of the bill.
[0022]
An intersection of a scan line including one end point of the banknote upper short side and a line representing the skew angle is obtained, and this is set as a skew correction origin. As a result, since the skew correction origin is set outside the banknote area, when coordinate conversion for rotating the banknote area is performed, the area including the banknote area is reliably tilted without partial data loss. It becomes possible to correct the line.
[0023]
As a result, the skew angle and the skew correction origin are obtained. Finally, coordinate transformation (affine transformation) is performed on each of the images 13, 14, 15 before synthesis. The skew-corrected images 13, 14, and 15 are used for banknote discrimination, for example, by taking a mutual color difference.
[0024]
In the above embodiment, three types of light sources, green, red, and infrared, are used as the light source, but two colors of green and red, or other visible light colors, and further a light source in the ultraviolet region are used in combination. You can also.
[0025]
【The invention's effect】
As described above, in the present invention, a plurality of images are simultaneously read from one reading region by a composite sensor using a plurality of light sources, and the plurality of images are superimposed to form one composite image. The skew angle of the medium is detected by using the outline (edge). Thus, even if the edge of the medium image data cannot be clearly obtained by a certain light source sensor, the images obtained by other light source sensors are superimposed, so that the edge becomes clear and the angle detection accuracy can be improved.
[0026]
In addition, after detecting the skew angle, the same origin of the composite image is used to correct the skew of each image before composition, so there is no deviation of each image and the accuracy of the origin of each image is improved. It is possible to improve.
[Brief description of the drawings]
FIG. 1 is a schematic diagram showing a configuration example of an image acquisition device in a bill validating device.
FIG. 2 is a diagram showing a schematic structure of an LED array.
FIGS. 3A and 3B are diagrams for explaining separation of a read image, where FIG. 3A shows a state of a banknote scan, and FIG. 3B shows a separated image;
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 Light emission timing control part 2 LED (light emitting diode) array 3 Selfoc lens 4 CCD image line sensor 5 A / D (analog / digital) conversion part 6 Image separation part 7 Central processing unit (CPU)
8 Main memory 9 Work memory 10 Read-only memory (ROM)
11 Bus 12 Banknote

Claims (4)

In the image acquisition method of a paper sheet medium for acquiring a discrimination image of the paper sheet medium,
Scan paper media while switching multiple light sources with different wavelengths for each scan line,
Generate multiple separated images by distributing the data read for each scan line in synchronization with the switching of the light source,
Logically synthesizing the separated images to generate one logical composite image;
Detecting a skew angle with respect to the logical composite image;
Obtain media area coordinates for the logic composite image;
Skew correction origin calculation for the logical composite image,
Performing skew correction on the medium area of each separated image centered on the calculated skew correction origin;
An image acquisition method for a paper sheet medium. In an image acquisition apparatus for a paper sheet medium for acquiring a discrimination image of a paper sheet medium,
A light emitting means in which a combination of a plurality of light sources having different emission wavelengths is linearly arranged; a light receiving means for sequentially scanning and reading the paper sheet medium illuminated by the light emitting means;
A light source switching means for switching and controlling a light source of the light emitting means each time the paper sheet medium is scanned;
Analog / digital conversion means for digitizing the pixel signal output from the light receiving means;
Image separation means for distributing the output from the analog / digital conversion means to image data for each type of the light source in synchronization with the light source switching means;
Separated image data storage means for storing the image data separated by the image separation means;
Logically synthesized image data storage means for storing logically synthesized image data obtained by logically synthesizing the plurality of separated image data into one image;
Program storage means for storing the skew correction program for generating the logical image data, calculating the skew angle, the medium area coordinates, and the skew correction origin from the logical composite image data, and performing skew correction of each separated image data When,
Arithmetic means for executing the skew feeding correction program;
An image acquisition apparatus for paper sheet media, comprising: 3. The paper sheet medium image acquiring apparatus according to claim 2, wherein the light emitting means includes a light source in at least a visible light region. 3. The image acquisition apparatus for a paper sheet medium according to claim 2, wherein the light receiving means detects transmitted light of the paper sheet medium.

Images