JPH07129774A - Image identifying method and image identifying apparatus - Google Patents

Abstract

(57) [Summary] [Purpose] Gain adjustment can be easily performed, and
Gain free maintenance of gain adjustment without providing a special gain adjustment mechanism. In addition, a permissible range for the density (density) of the original image is widened, and a stable image identification method and image identification device that are less susceptible to the density are provided. Furthermore, it is possible to perform image recognition to some extent with an inexpensive device. [Structure] The original image is irradiated with light, the reflected light is received, and photoelectric conversion is performed to generate first image information.
Generating a second image information by delaying the first image information by a predetermined time, comparing the first image information with the second image information,
Pseudo-image information is generated based on the magnitude relation between the first image information and the second image information, and the pseudo-image information is compared with a preset determination pattern, so that the original image is a specific image. Identify whether there is.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image identifying method and an image identifying apparatus, and more specifically, it compares image information generated by optically reading a document image with a preset determination pattern. Thus, the present invention relates to an image identification method and an image identification device for identifying whether or not a document image is a specific image.

[0002]

2. Description of the Related Art In recent years, as the social needs have increased due to improvements in image processing technology, various device technologies, and promotion of automation in commerce and industry, the original image is read optically and Image identification techniques for identifying whether or not a particular image has been widely used.

As a conventional image identification apparatus, for example, an ATM (Automatic T) used in banks and the like is used.
(eller Machine: automatic teller machine), a process of checking printed patterns in a printing factory line, a sorting device for gift certificates, a bar code reading device in a product management system, and a mark reading device in various sorting operations.

In such an image identification device, an original image is irradiated with light, the reflected light is received by an image sensor using a CCD element, photoelectric conversion is performed to generate an analog signal, and an analog signal is generated using an amplifier. The analog signal is amplified to a predetermined range, and then the analog signal is converted into a digital signal by using a threshold value (threshold level) set in advance by the A / D converter to read the image information of the original image, After that, by comparing with a preset determination pattern (digital signal), it is determined (identified) whether or not the original image is a specific image. In other words, the original image is optically read to generate density data (image information), and the original image is discriminated based on the density data.

[0005]

However, according to the above-mentioned conventional image identifying method and image identifying apparatus, an analog signal input by the image sensor is converted into a digital signal by using a preset threshold value. Cage,
Since the analog signal is converted into a digital signal only when the analog signal is larger or smaller than the threshold value, in order to accurately read the original image, the gain adjustment of the amplifier for amplifying the analog signal to a predetermined range is highly accurate. However, there is a problem that the gain adjustment is troublesome.

When the light amount of the light source for irradiating the original image with light changes due to the change over time, the analog signal output from the image sensor becomes small as a whole. In order to read the image accurately, the gain of the amplifier is adjusted manually by hand, or automatically based on the maximum value (white data) and the minimum value (black data) of the analog signal input by the image sensor. There is a problem that it is necessary to provide a gain adjusting mechanism for adjusting the gain.

Further, in the case where the gain adjusting mechanism is not provided, the analog signal input by the image sensor is converted into a digital signal based on whether the analog signal is larger or smaller than a threshold value, and the density of the original image ( There is also a problem that the original images that should be identified as the same image may not be identified because they are easily affected by the density.

Further, the original image is optically read to generate density data (image information), and the original image is identified based on the density data. In other words, the light and shade of the original image is determined by the image information. Since it is generated as, and is used for identifying the original image, even if the images have the same content, if the overall image is dark or light, or if the image density is uneven, the original image There is also a problem in that the original images that should be identified as the same image may not be identified because they are easily affected by the density of the image.

Further, according to the conventional image identifying method and the image identifying apparatus described above, although the image can be recognized with high performance, it is premised that the image sensor using the CCD element is used, so that the apparatus price is low. There is also a problem that it tends to be expensive and is not suitable for an inexpensive system.

There is a demand for some users to perform image identification to some extent with an inexpensive device that does not have such high performance, but an image that can perform some image identification within a practically usable range. Since there is no identification method or image identification device, the conventional image identification method and image identification device described above are also applied to these requirements, and the user needs to purchase a high-priced device with more performance than necessary. There was also a problem.

The present invention has been made in view of the above, and it is an object of the present invention to make it possible to easily perform gain adjustment and to maintain the gain adjustment maintenance-free without providing a special gain adjustment mechanism. To do.

Further, the present invention has been made in view of the above, and provides a stable image identification method and image identification apparatus which widen the allowable range for the darkness (density) of a document image and are not easily influenced by the density. The purpose is to provide.

Further, the present invention has been made in view of the above, and it is an object of the present invention to enable image recognition to some extent with an inexpensive device.

[0014]

In order to achieve the above object, the present invention compares the image information generated by optically reading a document image with a preset determination pattern to obtain a document. In an image identification method for identifying whether or not an image is a specific image, first image information is generated by irradiating a document image with light, receiving the reflected light, and performing photoelectric conversion to generate the first image information. The first image information is delayed for a predetermined time to generate the second image information, the first image information and the second image information are compared, and the first image information and the second image information are compared in magnitude. Provided is an image identifying method for identifying whether or not a document image is a specific image by generating pseudo image information based on the relationship and comparing the pseudo image information with a preset determination pattern. Is a thing

The first image information and the second image information are analog signals, and the pseudo image information is "1" when the first image information is larger than the second image information. ", A binary digital signal which becomes a value of" 0 "when the first image information is smaller than the second image information.

In order to achieve the above object, the present invention irradiates the original image with light and receives the reflected light to perform photoelectric conversion to generate the first image information. Generating means, second image information generating means for inputting the first image information and delaying it for a predetermined time to generate second image information, and the first image information and the second image information. In comparison, pseudo image information generating means for generating pseudo image information composed of binary digital signals of "0" and "1" based on the magnitude relation between the first image information and the second image information, A storage unit that stores a binary digital signal for determining a specific image in advance as a determination pattern is compared with the pseudo image information and the determination pattern stored in the storage unit, and the original image is identified as a specific image. Judgment hand to judge whether it is an image There is provided an image identification apparatus equipped with and.

In order to achieve the above-mentioned object, the present invention compares the image information generated by optically reading the original image with a preset judgment pattern to identify the original image. In the image identification method for identifying whether or not it is an image, the original image is irradiated with light,
The reflected light is received and photoelectric conversion is performed to generate first image information, and the first image information is delayed for a predetermined time to generate second image information. The second image information is compared, pseudo image information is generated based on whether the values of the first image information and the second image information match, and the pseudo image information is preset. An image identification method for identifying whether or not a document image is a specific image by comparing it with a determination pattern.

Further, in order to achieve the above object, the present invention irradiates the original image with light and receives the reflected light to perform photoelectric conversion to generate the first image information. Generating means, second image information generating means for inputting the first image information and delaying it for a predetermined time to generate second image information, and the first image information and the second image information. Pseudo-image for comparing and generating pseudo-image information composed of binary digital signals of "0" and "1" based on whether or not the values of the first image information and the second image information match. The information generation means, the storage means that stores a binary digital signal for determining a specific image in advance as a determination pattern, and the pseudo image information and the determination pattern stored in the storage means are compared, Determine whether the original image is a specific image There is provided an image identification apparatus and a judging means for.

[0019]

According to the image identifying method and the image identifying apparatus (claims 1 to 3) of the present invention, the original image is irradiated with light, and the reflected light is received to perform photoelectric conversion. Signal) and delays the first image information for a predetermined time to generate second image information (analog signal).
If the first image information is larger than the second image information based on the magnitude relationship between the first image information and the second image information, "1" is compared with the second image information. Value, the pseudo image information having a value of "0" when the first image information is smaller than the second image information (binary digital signal)
Is generated, and then the pseudo image information is compared with a preset determination pattern to determine (identify) whether or not the original image is a specific image. In other words, the pseudo image information is generated as a binary digital signal without using the threshold value, and this digital signal is compared with the determination pattern (binary digital signal).

The image identifying method and the image identifying apparatus of the present invention (claims 4 and 5) generate the first image information by irradiating the original image with light and receiving the reflected light to perform photoelectric conversion. Then, the first image information is delayed by a predetermined time to generate second image information, the first image information and the second image information are compared, and the first image information and the second image information are compared. Pseudo image information consisting of binary digital signals of "0" and "1" is generated based on whether or not the values match, and then the pseudo image information and the judgment pattern (binary digital signal) are compared. Then, it is determined (identified) whether or not the original image is a specific image. In other words, the pseudo image information is generated as a binary digital signal without using the threshold value, and this digital signal is compared with the determination pattern (binary digital signal).

[0021]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The following is a detailed description with reference to the drawings of [Embodiment 1] and [Embodiment 2] in order of the case where the image identification method and the image identification apparatus of the present invention are applied to a gift certificate classifying apparatus To do.

[Embodiment 1] FIGS. 1A and 1B are block configuration diagrams of a gift certificate classifying apparatus according to a first embodiment, in which a detection target image formed on a gift certificate is input, and described later. Image detection sensors 101 for generating pseudo image information
And a receiver 102 for taking in the pseudo image information generated by the plurality of image detection sensors 101 to the next comparison and determination unit 104, and a binary digital signal for previously determining a gift certificate (specific image). Comparison determination unit 104 that determines whether or not the detection target image is an image of a gift certificate by comparing the determination pattern RAM 103 stored as a use pattern with the pseudo image information input from the image detection sensor 101 and the determination pattern. An operation display unit 105 for performing various operations and displays, and a gift certificate tray (not shown) that feeds the gift certificates one by one from a gift certificate table (not shown). A gift certificate transport unit 106 for transporting to and a ROM used as a work memory for storing various control programs, control numerical values, and work
/ RAM 107 and control unit 108 for controlling the entire apparatus
Composed of and.

As shown in FIG. 2B, the image detection sensor 101 roughly divides the image to be detected by irradiating the image with light and receiving the reflected light to perform photoelectric conversion to obtain the first image information. A first image information generation circuit 109 for generating the first image information, a second image information generation circuit 110 for generating the second image information by inputting the first image information and delaying it by a predetermined time,
Of the second image information and the image information of the second image information, and based on the magnitude relationship between the first image information and the second image information,
And a pseudo image information generation circuit 111 for generating pseudo image information composed of binary digital signals of "0" and "1".

Here, referring to FIG. 2, an example of the circuit configuration of the first image information generation circuit 109, the second image information generation circuit 110, and the pseudo image information generation circuit 111 will be specifically described. The first image information generation circuit 109 includes a light emitting diode 201 for irradiating the detection target image with light, and a photoelectric converter for receiving the reflected light of the light emitted from the light emitting diode 201 to the detection target image and converting it into an electric signal. It is provided with two phototransistors 202 as conversion elements, an amplifier 203 for amplifying an electric signal (analog signal) photoelectrically converted by the phototransistor 202, and a variable resistor VR1 for adjusting the gain of the amplifier 203. . In addition, 204 is a condenser and R1, R2 and R3 are resistors respectively.

Further, the second image information generation circuit 110 is
A time constant circuit including a capacitor 205, diodes 206 to 208, a resistor R4, and a resistor R5. The analog signal output from the amplifier 203 of the first image information generation circuit 109 is input to the capacitor 205 and the resistor R4.
The analog signal delayed by the CR time constant determined by the relationship with is output as the second image information.

Further, the pseudo image information generation circuit 111 is
It comprises a differential amplifier circuit comprising resistors R6, R7 and an amplifier 209, and a knot gate 210. Here, the first image information is input to the negative input terminal of the amplifier 209, and the second image information is input to the positive input terminal. In addition, the knot gate 210 inputs the analog signal that is output after the first image information and the second image information are differentially amplified by the amplifier 209, and the first image information and the second image information match each other. The value (that is, the voltage value “0”) to be used as a threshold value is converted into a binary digital signal of “0” and “1”. In addition, this knot gate 2
The binary digital signal output from 10 becomes pseudo image information.

FIG. 3A shows the image detection sensor 10 described above.
1 shows the structure of No. 1 and each circuit is formed on a control substrate 303 arranged in a case 301 and a glass 302. In addition, 304 is the image detection sensor 101
An output connector 305 for outputting a signal from the receiver to the receiver 102, and a light shielding plate 305.

Further, FIG. 3B shows the image detection sensor 1
In the present embodiment, the light emitted to the detection target image is emitted from the light emitting diode 201 as much as possible.
In order to make only the direct light output from the
6a is provided, and a slit film 306b for narrowing down the detection point is provided at the tip of the cap 306. By changing the size of the slit of the slit film 306b, it is possible to change the detection area (or detection range) of the detection target image detected by the image detection sensor 101. In addition, the light blocking plate 305 cuts off the corners on the phototransistor 202 side in order to block outside light and guide only the reflected light of the light rays of the light emitting diode 201 to the phototransistor 202, and the reflected light from the detection target image. Is formed perpendicularly to the plane portion 305a.

In the above configuration, (1) the principle of the image identifying method of the first embodiment and (2) the specific operation will be described in this order.

(1) Principle of Image Identification Method of First Embodiment In the first embodiment, the first image is obtained by irradiating the original image (image to be detected) with light and receiving the reflected light to perform photoelectric conversion. Information is generated, the first image information is delayed by a predetermined time to generate second image information, and then the first image information and the second image information are compared to determine the first image information and the first image information. Based on the magnitude relationship between the two pieces of image information, a value of "1" when the first image information is larger than the second image information, and "0" when the first image information is smaller than the second image information. A binary digital signal that is a value of is generated as pseudo image information and the pseudo image information is compared with a preset determination pattern to identify whether or not the original image is a specific image. It is a thing.

Specifically, in the image detection sensor 101, first, the image to be detected is irradiated from the light emitting diode 201, the reflected light is received by the phototransistor 202, converted into an electric signal, and amplified by the amplifier 203, First image information (see FIG. 2) is obtained.

Next, the second image information generation circuit 110 delays the first image information by the CR time constant to obtain the second image information (see FIG. 2).

The waveforms of the analog signals of the first image information and the second image information thus generated are CR
The waveforms are the same with a time constant difference. For example, as shown in FIG. 4, respective analog signals are obtained such that the waveform of the first image information indicated by the solid line follows the waveform of the second image information indicated by the broken line.

When the first image information and the second image information having such a waveform are input to the negative input terminal and the positive input terminal of the amplifier 209, the output of the amplifier 209 becomes
As shown in FIG. 4, depending on the magnitude relation between the first image information and the second image information, a negative value is obtained when the first image information is larger than the second image information, and the first image information Is smaller than the second image information, a positive value is obtained.

The output of the amplifier 209 is set to "0" by the knot gate 210 with a value at which the first image information and the second image information match (ie, voltage value "0") as a threshold value, It is converted into a binary digital signal of "1", and pseudo image information is obtained as shown in FIG.

The first image information and the second image information, which are original information before generating the pseudo image information, are image data (analog signal in this case) specific to the image to be detected, and the pseudo image information is Since this original information is generated by differentially amplifying and inverting and digitizing it, it naturally becomes image information (pseudo image information) specific to the detection target image. Therefore, when the pseudo image information generated as described above matches, it is possible to determine that the first image information is the same, in other words, the detection target images are the same.

Therefore, the image detection sensor 10
In step 1, an original image to be identified (in this embodiment, an image of a gift certificate) is read to generate pseudo image information, and the pseudo image information is used as a judgment pattern to make a judgment pattern RA.
The image to be detected is read by the image detection sensor 101 to generate pseudo image information. When the pseudo image information is compared with the determination pattern, the detection target image is detected. It can be identified as the original image.

However, since it is rare that the pseudo image information and the judgment pattern are completely coincident with each other due to the image stain of the gift certificate as the detection target image, in reality, the pseudo image information and the judgment pattern are The method of obtaining the similarity and determining that they match when the similarity is equal to or greater than a predetermined value is applied.

As is apparent from the above description, the image identifying method of the present invention does not convert an analog signal into a digital signal by using a fixed threshold value, but rather the first image information and the second image information. Since the analog signal is converted into the digital signal to generate the pseudo image information based on the magnitude relationship with the image information of (3), it is not necessary to accurately adjust the gain of the amplifier 203. In other words, the gain adjustment can be roughly performed because the amplification can be performed to the extent that the circuit after the amplifier 203 can perform signal processing.

Further, in the conventional method, the image to be detected is optically read to generate density data (image information),
Since the original image is discriminated on the basis of the density data, it is easily affected by the density unevenness of the detection target image. However, in the image identification method of the present invention, the first image information and the second image information are combined. Since the analog signal is converted into a digital signal and pseudo image information is generated based on the magnitude relationship of the above, it is less susceptible to the density unevenness of the image to be detected, and the image can be identified in a wider range than before. .

Further, since it is not easily affected by the density unevenness of the image to be detected, when the pseudo image information is generated by inputting the image having the same contents as the judgment pattern, the pseudo image information and the judgment pattern match. If not, it can be easily recognized that stains (image stains) are present at the mismatched portions. Therefore, it is obvious that the image identification method of the present invention can be applied to a device for detecting dirt.

(2) Specific Operation Next, a specific operation of the image identifying apparatus according to the first embodiment will be described. First, before identifying a gift certificate, a determination pattern is registered in the determination pattern RAM 103. To register the judgment pattern, the mode key (not shown) of the operation display unit 105 is operated to specify the judgment pattern registration mode, and the original document (here, a gift certificate) to be identified is detected by the image detection sensor 101. Is performed to generate pseudo image information, and the pseudo image information is stored in the determination pattern RAM 103 as a determination pattern.
It should be noted that the RAM for storing the determination pattern in advance by using a dedicated device may be attached as the determination pattern RAM 103 to omit the registration process of the determination pattern.

FIGS. 5A to 5C show examples of registration of judgment patterns. After designating the determination pattern registration mode, the gift certificate of the original image to be identified is set on a predetermined mounting table (not shown), and the start key (not shown) is pressed. The sheets are fed one by one and, as shown in FIGS. 5A to 5C, are conveyed to an image input position where a plurality of image detection sensors 101 are arranged. At this time, the gift certificate is conveyed while being pressed against the side guide 501 for skew correction, thereby determining the position of the detection target image on the gift certificate read by each image detection sensor 101. Although an example in which five image detection sensors 101 are arranged is described here for simplification of description, the number and arrangement of the image detection sensors 101 are not particularly limited to this.

An image of the conveyed gift certificate is read by each image detection sensor 101 to generate pseudo image information. The pseudo image information is stored in the determination pattern RAM 103 together with the line positions (L1 to L5 in the figure) of the image detection sensor 101 that read the pseudo image information.

The above processing is executed only for the types of gift certificates,
Register each judgment pattern. For example, the pseudo image information of the line positions L1 to L5 of the image of the product exchange ticket of 1000 yen shown in FIG. 9A is generated and registered as the determination pattern. Similarly, the pseudo image information of the line positions L1 to L5 of the image of the product exchange ticket of 500 yen in the same figure (b) is generated as a judgment pattern, and the product discount ticket of 1000 yen in the same figure (c) is generated. The pseudo image information of the image line positions L1 to L5 is generated and registered as a determination pattern.

By storing the pseudo image information generated by the plurality of image detection sensors 101 arranged at different line positions as a judgment pattern, each gift certificate can be surely subjected to the gift certificate identification process described later. Can be identified. For example, an image of L1, L2, and L5 is the same in the product exchange ticket of 1000 yen in FIG. 10A and the product exchange ticket of 500 yen in FIG. ), But L3 and L4 are different, so L
1, L2, L5 can be identified as a product exchange ticket,
Whether L3 or L4 is a 1000-yen ticket or a 500-yen ticket can be reliably identified. In addition, L4 is the same in the 1000 yen merchandise exchange ticket in FIG. 10A and the 1000 yen merchandise discount ticket in FIG. 7C, but L1, L2, L3, L
Since 5 is different, it can be identified reliably.

Next, with reference to FIG. 6 and FIG. 7, a gift certificate identification process will be described. When the gift certificate to be identified is placed on the placing table of the gift certificate transport unit 106 and the start key of the operation display unit 105 is pressed, the control unit 108 executes the gift certificate identification process shown in the flowchart of FIG. 6. First,
By controlling the gift certificate transport unit 106, the gift certificate is transferred from the mounting table to 1
Sheets are fed and conveyed to an image input position where a plurality of image detection sensors 101 are arranged (S601).

Next, with a plurality of image detection sensors 101,
The detection target image on the gift certificate conveyed to the image input position is input, pseudo image information is generated (S602), and the comparison / determination unit 104 compares each pseudo image information with the determination pattern (S603).

Then, based on the judgment result of the comparison judgment unit 104, the kind of the gift certificate is identified from the combination of the pseudo image information in which the positions L1 to L5 (see FIG. 5) coincide (S6).
04), based on the identified result, the gift certificate transport unit 106
Is controlled to sort the gift certificate into the discharge destination tray of the corresponding classification destination (S605).

FIG. 7 shows steps S602 and S60 of FIG.
3 is a flowchart showing 3 more specifically. The image detection sensor 101 illuminates a gift certificate with a light emitting diode 201 and reflects the reflected light on the phototransistor 20.
The image to be detected is optically read by converting it into an analog signal in step 2, and the read analog signal is amplified by the amplifier 203 to generate the first image information (S70).
1).

Next, the second image information generation circuit 110 (time constant circuit) delays the first image information by a predetermined time to generate the second image information (S702).

Subsequently, in the pseudo image information generation circuit 111, the first image information and the second image information are differentially amplified by using the amplifier 209 and the knot gate 210,
Pseudo image information is generated by inverting and converting it into a binary digital signal. In other words, the first image information and the second image information
The pseudo image information is generated based on the size relationship with the image information of (S703).

The comparison / determination unit 104 inputs pseudo image information from each image detection sensor 101 via the receiver 102, and the determination pattern corresponding to the position (L1 to L5) of each image detection sensor 101 from the determination pattern RAM 103. Is read out and compared to obtain the similarity (S70).
4). Here, the method of obtaining the degree of similarity can use a general pattern matching method, and thus detailed description thereof will be omitted.

Next, the calculated similarity is compared with a reference value (similarity as a threshold value), and if the calculated similarity ≧ reference value, it is determined that the pseudo image information and the determination pattern match. Then, the determination result is output to the control unit 108 (S70).
5).

As shown in the above processing, by comparing the pseudo image information with the determination pattern, it can be determined whether or not the detection target image matches the original image to be identified. Furthermore, by identifying the type of the gift certificate based on these determination results, it is possible to reliably identify the gift certificate.

As described above, by generating the pseudo image information and comparing the pseudo image information with the judgment pattern,
Since it is determined whether or not the image to be detected matches the original image to be identified, gain adjustment can be easily performed, and maintenance of gain adjustment can be achieved without providing a special gain adjustment mechanism. it can. Further, since the image detection sensor can be configured without using a CCD element, an inexpensive device can be provided, and since the original image can be specified by the pseudo image information, the image can be identified to some extent by the inexpensive device.

Further, since a fixed threshold value is not required to convert the image information into a digital signal, even if the light amount of the light source (light emitting diode 201) changes due to a change with time, the image can be identified. Can be executed stably.

[Embodiment 2] The image identifying method and the image identifying apparatus according to Embodiment 2 generate the first image information by irradiating the original image with light and receiving the reflected light to perform photoelectric conversion. , The first image information is delayed by a predetermined time to generate second image information, the first image information and the second image information are compared, and the values of the first image information and the second image information are compared. Pseudo image information consisting of binary digital signals of "0" and "1" is generated based on whether or not match, and then the pseudo image information and the judgment pattern (binary digital signal) are compared. Then, it is determined whether or not the original image is a specific image. In other words, without using a threshold
The pseudo image information is generated as a binary digital signal, and this digital signal is compared with the judgment pattern.

8A and 8B are block configuration diagrams of a gift certificate classifying apparatus to which the image identifying method and the image identifying apparatus of the second embodiment are applied. The image detecting sensor 1 of the first embodiment is shown in FIGS.
Instead of 01, an image detection sensor 801 is arranged. Since the other configurations are the same as those in the first embodiment, only different parts will be described here.

As shown in FIG. 6B, the image detection sensor 801 roughly divides and irradiates the detection target image with light, receives the reflected light, and performs photoelectric conversion to generate the first image information. A first image information generating circuit 109, a second image information generating circuit 110 for inputting the first image information and delaying the first image information for a predetermined time, and generating the second image information; 2 image information, and based on whether the values of the first image information and the second image information match,
And a pseudo image information generation circuit 802 which generates pseudo image information composed of binary digital signals of "0" and "1".

FIG. 9 shows the first image information generation circuit 109,
A circuit configuration example of the second image information generation circuit 110 and the pseudo image information generation circuit 802 is shown. Here, the first image information generation circuit 109 and the second image information generation circuit 110
Since it is the same as the first embodiment, the pseudo image information generation circuit 802 will be described. Pseudo image information generation circuit 802
Inputs the first image information and the second image information, compares them, and outputs the digital signal "1" only when the first image information and the second image information have the same value. It is composed of a comparator 901 that outputs a digital signal "0" and outputs a digital signal "0" otherwise, and a resistor R6.

Here, the waveforms of the analog signals of the first image information and the second image information input to the comparator 901 are the same waveforms shifted by the CR time constant. For example, as shown in FIG. 10, respective analog signals are obtained such that the waveform of the first image information indicated by the solid line follows the waveform of the second image information indicated by the broken line.

The comparator 901 inputs the first image information and the second image information having such a waveform, and as shown in FIG. 10, the first image information and the second image information have the same value. In the case of, the digital signal "1" is output. The output of the comparator 901 is output from the pseudo image information generation circuit 802 as pseudo image information.

In the second embodiment as well, similar to the first embodiment, the first information which is the original information before the pseudo image information is generated.
Image information and the second image information are image data unique to the detection target image, and the pseudo image information is generated by digitizing this original information by a predetermined method, and thus is naturally unique to the detection target image. Image information (pseudo image information). Therefore, if the pseudo image information generated as described above matches, the first
It is possible to determine that the image information is the same, that is, the detection target images are the same.

Therefore, the image detection sensor 80
In 1, the original image to be identified (the image of the gift certificate) is read to generate pseudo image information, the pseudo image information is stored in the determination pattern RAM 103 as a determination pattern, and then the detection target image is image-detected. When the sensor 801 reads the pseudo image information to generate pseudo image information, and the pseudo image information is compared with the determination pattern, the detection target image can be identified as the original image.

However, since it is rare that the pseudo image information and the judgment pattern are completely coincident with each other due to the image stain on the gift certificate as the detection target image, in reality, the pseudo image information and the judgment pattern are The method of obtaining the similarity and determining that they match when the similarity is equal to or greater than a predetermined value is applied.

As is clear from the above description, the image identifying method of the second embodiment does not convert the analog signal into the digital signal by using the fixed threshold value, but rather the first
Since the analog signal is converted into the digital signal to generate the pseudo image information based on whether or not the image information of 1 and the second image information match, it is not necessary to accurately adjust the gain of the amplifier 203. In other words, the gain adjustment can be roughly performed because the amplification can be performed to the extent that the circuit after the amplifier 203 can perform signal processing.

Further, in the conventional method, the detection target image is optically read to generate density data (image information),
Since the original image is discriminated based on this density data, it was easily affected by the uneven density of the detection target image. However, in the image identification method of the second embodiment, the first image information and the second image information are used. Since the analog signal is converted into a digital signal and pseudo image information is generated based on whether or not coincides with, it is less susceptible to the density unevenness of the detection target image, and the image can be identified in a wider range than before. You will be able to do it.

Further, since the influence of the density unevenness of the image to be detected is less likely to occur, when the pseudo image information is generated by inputting the image having the same contents as the judgment pattern, the pseudo image information and the judgment pattern match. If not, it can be easily recognized that stains (image stains) are present at the mismatched portions. Therefore, it is obvious that the image identification method of the second embodiment can be applied to the dirt detection device.

[0070]

As described above, according to the image identifying method and the image identifying apparatus of the present invention, the first image information is obtained by irradiating the original image with light and receiving the reflected light to perform photoelectric conversion. The first image information is generated by delaying the first image information for a predetermined time to generate the second image information, the first image information and the second image information are compared, and the first image information and the second image information are generated. The pseudo image information is generated on the basis of the magnitude relation between the two, and the pseudo image information is compared with a preset judgment pattern to identify whether the original image is a specific image or not. Can be easily performed, and maintenance of gain adjustment can be achieved without providing a special gain adjustment mechanism. Further, the permissible range for the darkness (density) of the original image can be widened, and To be affected There can be provided a stable image identification method and an image identification apparatus, further, the equipment is inexpensive and can perform some degree of image identification.

Further, the image identifying method and the image identifying apparatus of the present invention compare the image information generated by optically reading the original image with the preset determination pattern to identify the original image. In an image identifying method for identifying whether an image is an image or not, first image information is generated by irradiating a document image with light, receiving the reflected light, and performing photoelectric conversion to generate the first image information. The second image information is generated with a delay of a predetermined time, and the first image information and the second image information are generated.
Of the first image information and the second image information, and generates pseudo image information based on whether or not the values of the first image information and the second image information match, and the pseudo image information is compared with the preset determination pattern. By comparing, it is possible to identify whether the original image is a specific image or not, so that the gain adjustment can be easily performed, and the gain adjustment is maintenance-free without providing a special gain adjustment mechanism. Further, it is possible to provide a stable image identification method and an image identification device that widen the tolerance range for the density (density) of the original image and are not easily affected by the density.
It is possible to perform image identification to some extent with an inexpensive device.

[Brief description of drawings]

FIG. 1 is a block configuration diagram of a gift certificate classifying apparatus to which an image identifying method and an image identifying apparatus according to a first embodiment are applied.

FIG. 2 is a circuit configuration example of a first image information generation circuit, a second image information generation circuit, and a pseudo image information generation circuit of the first embodiment.

FIG. 3A is an explanatory diagram showing a structure of an image detection sensor, and FIG. 3B is an explanatory diagram showing optical characteristics of the image detection sensor.

FIG. 4 is an explanatory diagram showing the principle of the image identification method of the first embodiment.

FIG. 5 is an explanatory diagram showing an example of registration of a determination pattern.

FIG. 6 is a flowchart of a gift certificate identifying process.

FIG. 7 is a flowchart showing steps S602 and S603 of FIG. 6 more specifically.

FIG. 8 is a block configuration diagram of a gift certificate classifying apparatus to which the image identifying method and the image identifying apparatus according to the second embodiment are applied.

FIG. 9 is a circuit configuration example of a first image information generation circuit, a second image information generation circuit, and a pseudo image information generation circuit of the second embodiment.

FIG. 10 is an explanatory diagram showing the principle of the image identifying method according to the second embodiment.

[Explanation of symbols]

 Reference Signs List 101 image detection sensor 103 determination pattern RAM 104 comparison determination unit 108 control unit 109 first image information generation circuit 110 second image information generation circuit 111 pseudo image information generation circuit 801 image detection sensor 802 pseudo image information generation circuit

Claims (5)

[Claims] 1. An image identification method for identifying whether an original image is a specific image by comparing image information generated by optically reading the original image with a preset determination pattern. In, the first image information is generated by irradiating the original image with light and receiving the reflected light to perform photoelectric conversion, and delays the first image information for a predetermined time to generate the second image information. The first image information and the second image information are generated, pseudo image information is generated based on the magnitude relation between the first image information and the second image information, and the pseudo image information is generated. An image identification method characterized by identifying whether or not a document image is a specific image by comparing it with a preset determination pattern. 2. The first image information and the second image information are analog signals, and the pseudo image information is "1" when the first image information is larger than the second image information. 2. The image identification method according to claim 1, wherein the image signal is a binary digital signal having a value of "0" and a value of "0" when the first image information is smaller than the second image information. 3. A first image information generating apparatus for illuminating a document image with light, receiving the reflected light and performing photoelectric conversion to generate first image information.
Image information generating means, second image information generating means for generating the second image information by inputting the first image information and delaying it for a predetermined time, the first image information and the second image Pseudo-image information generation for comparing the information and generating pseudo-image information composed of binary digital signals of "0" and "1" based on the magnitude relation between the first image information and the second image information. Means and 2 for judging specific image beforehand
A determination is made by comparing the pseudo image information and the determination pattern stored in the storage with storage means that stores a digital signal of a value as a determination pattern to determine whether the original image is a specific image. An image identification device comprising: 4. An image identification method for identifying whether or not a document image is a specific image by comparing image information generated by optically reading a document image with a preset determination pattern. In, the first image information is generated by irradiating the original image with light and receiving the reflected light to perform photoelectric conversion, and delays the first image information for a predetermined time to generate the second image information. Generating, comparing the first image information with the second image information, and generating pseudo image information based on whether or not the values of the first image information and the second image information match, An image identifying method characterized by identifying whether or not a document image is a specific image by comparing the pseudo image information with a preset determination pattern. 5. A first image information generating apparatus for illuminating a document image with light and receiving the reflected light to perform photoelectric conversion to generate first image information.
Image information generating means, second image information generating means for generating the second image information by inputting the first image information and delaying it for a predetermined time, the first image information and the second image Pseudo-image information composed of binary digital signals of "0" and "1" is generated by comparing with information and based on whether the values of the first image information and the second image information match. The pseudo image information generating means, the storage means that stores a binary digital signal for determining a specific image in advance as a determination pattern, and the pseudo image information and the determination pattern stored in the storage means are compared. Then, the image identifying apparatus is provided with a determining means for determining whether or not the original image is a specific image.