JP7128555B1 - Data matching method - Google Patents

Abstract

A code generation method, a code generation device, a program, and a data matching method for coding target data such as images and sounds are provided. A method for generating a code according to the content of target data using an information processing device, wherein the target data is divided into a plurality of sampling ranges, S12 is included in each of the plurality of sampling ranges, and each An average value of the element data is obtained for each sampling range for at least one element data among one or more types of element data represented by a numerical value S13, and the average value for each sampling range or a numerical value for a predetermined number of digits from the higher order is obtained. , character string data to generate a reference code corresponding to the target data (S15). [Selection drawing] Fig. 9

Description

The present invention relates to a code generation method, a code generation device, a program, and a data collation method.

There is known a technique for specifying the content of target data by comparing and collating it with reference data prepared in advance. A representative example of this technology is image recognition technology (see, for example, Japanese Patent Application Laid-Open No. 2019-28985). However, in the conventional technology, there is room for improvement in that a huge amount of time is required for learning to obtain reference data, or the processing load required for comparison and collation with target data is large and high-speed processing is difficult. .

JP 2019-28985 A

A specific aspect of the present invention aims to provide a novel technique for encoding target data such as images and sounds with a small amount of data.

[1] A data collation method according to one aspect of the present invention is a method for generating and collating a code according to the content of target data using an information processing device, comprising: (a) a code obtained according to reference data; (b) a second step of generating a specific target code corresponding to specific target data; and (c) comparing the reference code and the specific target code to compare them. (d) said referencing code in said first step comprises: (d1) dividing said referencing data into a plurality of first sampling ranges; (d2 ) For at least one of the first element data included in each of the plurality of first sampling ranges and each represented by a numerical value. (d3) calculating the average value of the first element data for each first sampling range; and (e) the specific target code in the second step is obtained in advance through the step of generating the reference code corresponding to the reference data by concatenating as character string data, ( e1) dividing the specific target data into a plurality of second sampling ranges set in the same manner as each of the first sampling ranges; The second sampling range for at least one type of the second element data of the same type as the first element data among one or more types of the second element data that are element data and are each represented by a numerical value (e3) concatenating the average value of the second element data for each second sampling range or a predetermined number of upper digits thereof as character string data; and (f) the third step includes (f1) each of the corresponding first sampling ranges and Between each of the second sampling ranges, based on the average value for each of the first sampling ranges obtained based on the reference code or a predetermined number of upper digits thereof and the specific target code for each said second sampling range obtained (f2) obtaining a difference between the average value and a predetermined number of digits from the upper value thereof; and determining that the data is collated.
[2] A data collation method according to one aspect of the present invention is a method of generating and collating a code according to the content of target data using an information processing device, comprising: (a) a code obtained according to reference data; (b) a second step of generating a specific target code corresponding to specific target data; and (c) comparing the reference code and the specific target code to compare them. (d) said referencing code in said first step comprises: (d1) dividing said referencing data into a plurality of first sampling ranges; (d2 ) For at least one of the first element data included in each of the plurality of first sampling ranges and each represented by a numerical value. (d3) calculating the average value of the first element data for each first sampling range; and (e) the specific target code in the second step is obtained in advance through the step of generating the reference code corresponding to the reference data by concatenating as character string data, ( e1) dividing the specific target data into a plurality of second sampling ranges set in the same manner as each of the first sampling ranges; The second sampling range for at least one type of the second element data of the same type as the first element data among one or more types of the second element data that are element data and are each represented by a numerical value (e3) concatenating the average value of the second element data for each second sampling range or a predetermined number of upper digits thereof as character string data; and (f) the third step includes (f1) each of the corresponding first sampling ranges and Between each of the second sampling ranges, based on the average value for each of the first sampling ranges obtained based on the reference code or a predetermined number of upper digits thereof and the specific target code For each said second sampling range obtained (f2) a step of determining the average value of the difference from the numerical value of a predetermined number of digits from the upper order or the power of the difference; and determining that the reference code and the specific target code match in the following cases.

According to the above configuration, a novel technique is provided for encoding target data such as images and sounds with a small amount of data.

FIG. 1 is a diagram for explaining the encoding method of the first embodiment. FIG. 2 is a diagram showing an example of database configuration of codes generated by the above method and contents corresponding to each code. FIG. 3 is a diagram conceptually explaining the flow of collation using a database. FIG. 4 is a diagram for explaining a specific method of performing flexible collation. FIG. 5 is a line graph showing the difference corresponding to each of the sampling ranges S1 to S9 in the example shown in FIG. FIG. 6 is a block diagram showing a configuration example of a code generation device for generating code using the method described above. FIG. 7 is a block diagram showing a configuration example of a user device that performs collation of target data using the code generated by the above method. FIG. 8 is a diagram showing a configuration example of a computer that implements the information processing unit in each of the code generation device and the user device. FIG. 9A is a flow chart showing the operation procedure of the code generator. Also, FIG. 9B is a flowchart showing the operation procedure of the user device. FIG. 10 is a diagram for explaining application example 1 of the encoding technique according to the above embodiment. FIG. 11 is a diagram for explaining application example 2 of the encoding technique according to the above embodiment. FIG. 12 is a diagram for explaining application example 3 of the encoding technique according to the above embodiment. FIG. 13 is a diagram for explaining application example 4 of the encoding technique according to the above embodiment. FIG. 14 is a diagram for explaining application example 5 of the encoding technique according to the above embodiment. FIG. 15 is a diagram for explaining an application example 6 of the encoding technique according to the embodiment described above. FIG. 16 is a diagram for explaining an application example 7 of the encoding technique according to the embodiment described above. FIG. 17 is a diagram for explaining the encoding method of the second embodiment. FIG. 18 is a diagram showing an example of a code expressed in hexadecimal character code. FIG. 19 is a diagram for explaining a matching method using codes according to the second embodiment. FIG. 20 is a diagram for explaining the encoding method of the third embodiment. FIG. 21 is a diagram for explaining a matching method using codes according to the third embodiment. FIG. 22 is a conceptual diagram showing how to use each embodiment.

<First embodiment>
FIG. 1 is a diagram for explaining the encoding method of the first embodiment. Here, image data will be described as an example of data to be coded (target data). In this specification, "encoding" refers to generating character string data that can identify target data based on a predetermined numerical value included in the target data.

Image data as shown in FIG. 1A is assumed as target data. The image data may be a color image or a monochrome image. Image data includes values (element data) such as brightness, hue, saturation, and color difference for each pixel. For ease of understanding, luminance values are used as element data for explanation.

As shown in FIG. 1B, image data is divided into a plurality of sampling ranges. Here, the image data is divided into nine sampling ranges S1, S2, S3, . . . S9. For example, each sampling range S1, etc. has a width and height of 40 pixels, and the total number of pixels included in each sampling range S1, etc. is 1600, respectively. It is also assumed that the luminance value (the number of gradations) of each pixel is given by a numerical value in the range of 0-255.

At this time, if the pixel values (here, luminance values) of the pixels included in each sampling range S1, etc. are px1, px2, . You are asked to
Sum of pixel values: sumSx=px1+px2+px3+...px1600
Average of pixel values: Ax=sumSx/1600

FIG. 1C shows an example of average values of pixel values in each sampling range S1 and the like. In the illustrated example, the average value A1 of the sampling range S1 is 128, the average value A2 of the sampling range S2 is 86, the average value A3 of the sampling range S3 is 134, the average value A4 of the sampling range S4 is 101, and the average value of the sampling range S5 The value A5 is 68, the average value A6 of the sampling range S1 is 101, the average value A7 of the sampling range S7 is 115, the average value A8 of the sampling range S8 is 91, and the average value A9 of the sampling range S9 is 121.

Next, the relative difference between the average values in each sampling range S1 etc. is obtained as follows.
Sum of average values: sumA=A1+A2+A3+...+A9
=945
Relative difference of mean values: Vx = (100 x Ax)/945

FIG. 1(D) shows an example of the relative difference between the average values in each sampling range S1 and the like. In the illustrated example, the relative difference V1 of the sampling range S1 is 13.54%, the relative difference V2 of the sampling range S2 is 9.10%, the relative difference V3 of the sampling range S3 is 14.17%, and the relative difference of the sampling range S4 is V4 is 10.68%, the relative difference V5 of the sampling range S5 is 7.19%, the relative difference V6 of the sampling range S6 is 10.68%, the relative difference V7 of the sampling range S7 is 12.16%, the sampling range S8 The relative difference V8 is 9.62%, and the relative difference V9 of the sampling range S9 is 12.80%.

Using the relative differences in each of these sampling ranges S1 and the like, numerical values of digits from the upper to the second decimal place of each relative difference (four digits in total) are extracted as character string data. For example, the character string data "1354" is extracted from the sampling range S1. If there is no value in the tens place or no value in the digits below the decimal point, the value is set to 0. For example, 9.1% is "0910".

In this way, the character string data extracted from each sampling range S1 and the like are concatenated. The order of connection is not particularly limited and can be determined arbitrarily. Here, the sampling ranges S1, S2, . . . S9 are connected in this order. As a result, in the above example, the character string data "135409101417106807191068121609621280" is obtained (see FIG. 1(E)). In this embodiment, the character string data generated in this way is used as the code corresponding to the target data.

Note that the relative difference does not necessarily have to be expressed in percentage. The order of concatenation of the character string data corresponding to each sampling range is arbitrary on the premise that a predetermined order is always used. For example, sampling ranges S9, S8, . It is also possible to define to connect in the order of . Also, the number of digits of the relative difference to be used in each sampling range is also arbitrary. The number of divisions of the sampling range is also arbitrary. If the number of divisions of the sampling range is reduced, matching can be performed more easily and at high speed.

FIG. 2 is a diagram showing an example of database configuration of codes generated by the above method and contents corresponding to each code. Here, an image is taken as an example of content. As shown in the figure, codes corresponding to each image 1, 2, . For example, image 1 corresponds to code "2110...", and image 2 corresponds to code "1053...". Here, the images 1, 2, . position information on the physical space, time information, and any other additional information). By constructing such a database in advance, it is possible to use each code in this database as reference data to check which content the specific target data corresponds to.

FIG. 3 is a diagram conceptually explaining the flow of collation using a database. For example, image data obtained by photographing is coded by the above method, and a specific target code ("8841..." in the illustrated example) corresponding to this image data is generated. . This generated specific target code is compared with each reference code in the database, and if a reference code matching the generated specific target code exists in the database, the content corresponding to the photographed image ( "Image 37") can be specified.

For example, it is assumed that image data obtained by photographing an arbitrary space using a terminal device (eg, a smartphone) equipped with a camera is obtained. It is also assumed that the terminal device is pre-installed with an application program capable of executing the above-described encoding, and that a database such as that shown in FIG. 2 is stored. Note that the database may be placed on a server that can communicate with the terminal device via a network. In this case, for example, if the content corresponding to the image of the landscape captured by the user pointing the camera is specified, the information related to the content can be displayed on the screen of the user's terminal device.

According to the study of the inventors of the present application, if the target data is image data obtained by photographing a space, it is practically necessary and sufficient to perform encoding by setting the sampling range to 64 or more. Knowledge has been obtained that matching and judgment can be performed with high accuracy. The data volume of the code generated under this condition is about 192 bytes, which can be said to be a very small data volume.

Here, according to the principle of matching explained with reference to FIG. 3, when the character string data of the codes match, it is determined that the contents corresponding to each code are the same. By setting the permissible range for each case, more flexible collation can be performed.

FIG. 4 is a diagram for explaining a specific method of performing flexible collation. FIG. 5 is a line graph showing the difference corresponding to each of the sampling ranges S1 to S9 in the example shown in FIG.

In the example shown in FIG. 4(A), the character string data of each of the reference code A and the specific target code B is divided into four digits corresponding to the sampling range described above, and each character string data is divided into four digits. , and find the numerical difference (AB) for each corresponding sampling range. For example, the numerical value corresponding to the sampling range S1 is "1354" in the reference code and "1368" in the specific target code, so the difference between the two is "-14". Similarly, the differences for the other sampling ranges S2 to S9 are +15, 0, -12, +19, +17, +15, 0, +18. At this time, if the allowable range of the difference corresponding to each sampling range is predetermined as ±20, for example, in the example shown in FIG. fit. In such a case, it can be determined that the reference code and the specific target code are "identical (matched)". As shown in FIG. 5, when it can be determined that the reference code and the specific target code are the same (A and B), the graphs almost overlap.
On the other hand, in the example shown in FIG. 4B, for the character string data of the reference code A and the specific target code C, the differences for each sampling range S1 to S9 are +74, -101, +27, 0, - 261, +38, +126, -249, +273. Assuming that the allowable range of the difference corresponding to each sampling range is predetermined as ±20, in the example shown in FIG. The corresponding difference is not within the acceptable range. In such a case, it can be determined that the reference code and the specific target code are “different (mismatched)”. In this case, as shown in FIG. 5, when it can be determined that the reference code and the specific target code are different (A and C), the graphs do not overlap.

Note that if the allowable range is not provided, it means that the image capturing conditions (reflected light, ambient light, etc.) are exactly the same. On the other hand, if a permissible range is provided, it becomes possible to derive a "solution" corresponding to a weak environmental change such as the state of light and the user's sense. By using this allowable range (environmental variable), it is possible to realize a judgment (maybe) that is close to the sense of the human eye. The size of the allowable range can be set arbitrarily, but according to the study of the inventors of the present application, it is preferable to set it to about ±20 as described above.

FIG. 6 is a block diagram showing a configuration example of a code generation device for generating code using the method described above. The illustrated code generation device 1 includes an information processing section 10 , a camera 11 , a microphone 12 , a storage device 13 , an operation device 14 and a display device 15 .

The information processing unit 10 performs information processing related to code generation. The information processing unit 10 includes a sampling processing unit 20, a feature value calculation unit 21, an encoding processing unit 23, and a code recording processing unit 24 as functional blocks. The information processing unit 10 is realized by using a computer having, for example, a CPU (Central Processing Unit), ROM (Read Only Memory), RAM (Random Access Memory), etc., and causing the computer to execute a predetermined operation program. In this embodiment, the sampling processing unit 20 corresponds to the “dividing unit”, the feature value calculation unit corresponds to the “average value calculation unit” and the “relative difference calculation unit”, and the encoding processing unit 23 corresponds to the “code generation unit”. ” corresponds to

The camera 11 captures a space, an object, or the like that can be a target of code generation, and generates an image thereof. The image here is not necessarily limited to an image in the visible light range, and may be, for example, a thermal image, an infrared image, or the like. The camera 11 outputs image data (or image signals) of the captured image to the information processing section 10 .

The microphone 12 picks up sounds that can be used for code generation and converts them into electric signals. The sound referred to here is not necessarily limited to sound in the audible range, and may be ultrasonic waves, low-frequency sounds, or the like. A sound signal output from the microphone 12 is output to the information processing section 10, where it is converted into digital data and captured.

The storage device 13 stores programs and various data necessary for the configuration and operation of the information processing section 10 and also stores a database of codes generated by the information processing section 10 . The storage device 13 is a non-volatile storage device such as a hard disk drive (HDD) or solid state drive (SSD).

The operation device 14 is used for inputting information necessary for the operation of the information processing section 10, and may include, for example, a keyboard, a mouse, a switch, a touch panel, and the like. The display device 15 is used to display information related to the operation of the information processing section 10, and for example, a liquid crystal display device, an organic EL display device, or the like is used.

The sampling processing unit 20 performs processing for dividing target data for code generation into a plurality of sampling ranges (see FIG. 1B). The target data is, for example, image data or sound data. In the case of sound data, the sound data can be divided into predetermined time intervals and the sound data included in these intervals can be used as data corresponding to each sampling range (see the third embodiment described later).

The feature value calculation unit 21 obtains the average value of predetermined data included in each sampling range divided by the sampling processing unit 20 (see FIG. 1C), and calculates the relative difference between the average values corresponding to each sampling range. (See FIG. 1(D)). The relative difference is expressed as a percentage, for example, as in the example above, but is not limited to this.

The encoding processing unit 23 uses the relative difference (or average value) corresponding to each sampling range calculated by the feature value calculating unit 21 to generate a code (a code obtained by concatenating character string data) corresponding to the target data. .

The code recording processing unit 24 associates the code generated by the encoding processing unit 23 with the information data specifying the content corresponding to this code, records it in the storage device 13, and stores the code database (see FIG. 2). to generate

FIG. 7 is a block diagram showing a configuration example of a user device that performs collation of target data using the code generated by the above method. The illustrated user device 2 includes an information processing section 50 , a camera 51 , a microphone 52 , a storage device 53 , an operation device 54 and a display device 55 . This user device 2 is realized by, for example, installing a predetermined application program in advance in an information processing device such as a smartphone possessed by the user.

The information processing section 50 performs information processing related to code collation. The information processing unit 50 includes a sampling processing unit 60, a feature value calculation unit 61, an encoding processing unit 63, and a matching processing unit 64 as functional blocks. The information processing section 50 is realized by using a computer having, for example, a CPU (Central Processing Unit), ROM (Read Only Memory), RAM (Random Access Memory), etc., and causing the computer to execute a predetermined operation program.

The configurations and operations of the camera 51, the microphone 52, the storage device 53, the operation device 54, and the display device 55 are the same as those of the camera 11, the microphone 12, the storage device 13, the operation device 14, and the display device in the code generation device 1 described above. 15, detailed description is omitted here. Also, the configurations and operations of the sampling processing unit 60, the feature value calculation unit 61, and the encoding processing unit 63 are the same as those of the sampling processing unit 20, the feature value calculation unit 21, and the encoding processing unit 23 in the code generation device 1 described above. Therefore, detailed description is omitted here.

The collation processing unit 64 of the information processing unit 50 compares the specific target code obtained by the respective processes of the sampling processing unit 60, the feature value calculation unit 61, and the encoding processing unit 63 with a database stored in advance in the storage device 53 (Fig. 2) to identify the content corresponding to the specific target code. A specific collation method by the collation processing unit 64 may be a method requiring complete matching between codes, or a flexible method in which an allowable range is provided for differences between codes as described above. good. The identification result is displayed on the display device 55, for example.

FIG. 8 is a diagram showing a configuration example of a computer that implements the information processing unit in each of the code generation device and the user device. The illustrated computer includes a CPU 201, a ROM 202, a RAM 203, a storage device 204, and an external interface (I/F) 205 which are communicatively connected. The CPU 201 operates based on a basic control program read from the ROM 202, reads a program (application program) 206 stored in the storage device 204, and executes the program, thereby realizing the functions of the controller 13 described above. A RAM 203 temporarily stores data to be used when the CPU 201 operates. The storage device 204 is a nonvolatile storage device such as a hard disk or solid state drive, and stores various data such as the program 206 . The external interface 205 is an interface that connects the CPU 201 and an external device, and is used, for example, to connect the camera 11 (51), the microphone 12 (52), the operation device 14 (54), the display device 15 (55) and the CPU 201. .

FIG. 9A is a flow chart showing the operation procedure of the code generator. Also, FIG. 9B is a flowchart showing the operation procedure of the user device. In any operation procedure, the order of processing can be changed as appropriate as long as there is no contradiction or inconsistency in the results of information processing, and other processing not mentioned here may be added. is not excluded.

First, the operation procedure of the code generator will be described.
When target data is input to the code generation device 1 (step S11), the sampling processing unit 20 performs processing for dividing the target data into a plurality of sampling ranges (step S12). Here, the target data may be image data input from the camera 11, sound data corresponding to a signal input from the microphone 12, or image data or sound data previously stored in the storage device 13. good. Alternatively, the code may be transmitted from a server or the like (not shown) via a network and received by the code generation device 1 .

Next, the feature value calculator 21 obtains the average value of predetermined element data included in each sampling range divided by the sampling processor 20 (step S13). Here, the element data may be any data that is included in the target data as described above and forms the contents, features, or characteristics of the target data. For example, as described above, in the case of image data, it may be the brightness of each pixel, and in the case of sound data, it may be the volume of sound for each predetermined sampling time.

Next, the feature value calculator 21 obtains the relative difference between the average values corresponding to each sampling range (step S14). The relative difference is expressed as a percentage, for example, as in the example above, but is not limited to this.

Next, the encoding processing unit 23 uses the relative difference corresponding to each sampling range calculated by the feature value calculating unit 21 to generate a reference code (a code obtained by concatenating character string data) corresponding to the target data. (Step S15).

Next, the code recording processing unit 24 associates the reference code generated by the encoding processing unit 23 with information data specifying the content corresponding to this reference code, and records the data in the storage device 13 for reference. A code database (see FIG. 2) is generated (step S16).

By repeating the above series of information processing a necessary number of times, reference codes corresponding to each of a plurality of contents are generated, and a database containing such information is obtained. Note that the generated database may be transmitted to and stored in a server device or the like (not shown) communicably connected via a network. As a result, the database can be shared by multiple users via the network.

Next, the operating procedure of the user device will be described.
When specific target data is input to the user device 2 (step S21), the sampling processing section 60 performs processing to divide the specific target data into a plurality of sampling ranges (step S22). The specific target data here may also be image data input from the camera 11, sound data corresponding to a signal input from the microphone 12, or image data or sound data previously stored in the storage device 13. It's okay.

Next, the feature value calculator 61 obtains the average value of predetermined element data included in each sampling range divided by the sampling processor 60 (step S23). The element data here may also be data that is included in the target data as described above and forms the contents, features, or characteristics of the target data.

Next, the feature value calculator 61 obtains the relative difference between the average values corresponding to each sampling range (step S24).

Next, the encoding processing unit 63 uses the relative difference corresponding to each sampling range calculated by the feature value calculating unit 61 to generate a specific target code (a code obtained by concatenating character string data) corresponding to the specific target data. (step S25).

Next, the collation processing unit 64 refers to the database stored in advance in the storage device 53 to identify the content corresponding to the identification target code generated in step S25 (step S26). The identification result is displayed on the display device 55 as appropriate. The database stored in the storage device 53 is, for example, transmitted from a server or the like (not shown) via a network and received by the user device 2 .

The above is the basic contents of the code generation and collation of the first embodiment, and some application examples thereof will be described next. Both code generation and verification according to the following application examples are executed by the code generation device 1 and the user device 2 (or other device).

FIG. 10 is a diagram for explaining application example 1 of the encoding technique according to the above embodiment. Here, an example of application to face recognition technology is shown. The face recognition technology here means a technology for specifying a person corresponding to a face included in an image. A certain image may be a still image captured by a camera, or may be a moving image. In the case of moving images, still images obtained for each frame can be used. Assume that face image data 300 of a certain person from the front is given as shown in FIG. 10(A). For this face image data 300, a known technique is used to automatically generate face image data corresponding to when the person is viewed from above, below, left, and right. Of course, instead of automatic generation, the image may be taken directly from the person. Here, a total of 15 face image data including front face image data 300 and face image data corresponding to each direction of up, down, left, and right automatically generated (or photographed) around this are shown.

These face image data are coded by the code generation device 1 as shown in FIG. 10B to obtain reference codes. Here, code 1 corresponding to the front face image data 300, and codes 2 to 15 corresponding to the face image data in the other vertical and horizontal directions are obtained. These 15 codes 1 to 15 are linked as a set of face authentication data. For example, identification information such as "xxxxx_CEC" can be attached to each code as identification information (FaceID) and linked to them. In addition, attribute information (for example, name, age, affiliation, etc.) related to a person to be identified is associated with the identification information.

By generating and storing such face authentication data in advance, for example, the user device 2 (or another device) can perform face authentication based on image data obtained from a camera in real time. Various specific matching methods are conceivable. For example, if any one of the codes 1 to 15 matches, the person may be identified, or the person may be identified under the condition that a plurality of codes match. good. The code obtained by the encoding technology of this embodiment is irreversible character string data (text data), and the face image data itself is not used. Even if information is leaked, privacy can be protected. Also, since the data size of one code is several hundred bytes, the data volume required for face recognition can be dramatically reduced. For example, if the data size of one face image data is 100 kilobytes, the data capacity of 15 face image data is 1500 kilobytes, while the data size of one code is, for example, about 192 bytes. Therefore, the data capacity of 15 codes is 2880 bytes, and the difference in data capacity is obvious.

FIG. 11 is a diagram for explaining application example 2 of the encoding technique according to the above embodiment. Here, an application example is shown in which image data such as space is divided into a plurality of partial images, codes are generated for each partial image, and image recognition (space recognition) is performed. Specifically, as shown in FIG. 11A, image data 400 as target data is divided into a plurality of (here, 12) partial images A to L by the code generation device 1 described above. Then, the above coding is performed for each of the partial images A to L. A code generated corresponding to the partial image A is represented as A_code. The same applies to codes corresponding to other partial images. The 12 codes such as the A_code generated in this way are linked and recorded as one reference code group.

Image data as specific target data is similarly divided into a plurality of partial images by the user device 2 (or another device), and codes corresponding to each partial image are generated. For example, code generation is performed for a still image captured in real time by a user's camera or for one frame of a moving image. At this time, for example, as shown in FIG. 11B, it is assumed that the imaging range of the image data 400 corresponding to the reference data group partially overlaps with the image data 410 as the specified target data by the user. Here, it is assumed that portions corresponding to partial images E, F, I, and J overlap.

In this state, assuming that the code group generated from the image data 410, which is the specific target data, is referred to for the reference code group in a brute-force manner, the matching partial image for the total number of partial images N=12 , the number of partial images T=4, and the number of unmatched partial images F=8 (see FIG. 11(C)). If the matching rate (recognition rate) is defined as T/N and the non-matching rate is defined as F/N, then in this example the matching rate is about 33% (T/N=4/12), and the non-matching rate is about 66% (F/N=8/12). The percentage above which the matching rate is determined to match the specific target data and the reference data differs depending on the application, and may be set as appropriate. Also, the match rate can be used to judge similarity. For example, when the matching rate is about 33% as described above, it can be determined that there is a high possibility that the images are similar, or that the images are taken in a close place, but the images are high. By using such a technique, any frame-in can be dealt with. Further, the recognition accuracy can be improved as the total number N of partial images is increased. Even in this case, since search, collation, and determination are basically based on text data, high-speed processing equivalent to commonly known text search can be performed.

FIG. 12 is a diagram for explaining application example 3 of the encoding technique according to the above embodiment. Here, an application example for determining the vertical direction of an object is shown. Any object can be used as long as it has an asymmetrical shape in the vertical direction. For example, there is a case of judging the vertical direction in order to crack pistachios with an industrial machine.

As shown in FIG. 12(A), an object 500 is divided into two parts a and b by focusing on its vertical asymmetry. As an example, part a is considered a triangle and part b is considered a trapezoid. Assume that the base of part a and the bottom of part b are equal, and that the heights of both are equal. When the portions a and b are applied to regions 510 and 512 (rectangular regions indicated by dotted lines in the figure) of the same size and the remaining portions other than the portions a and b are compared, the rectangular region 510 corresponding to the portion a is larger than the portion b. The residual portion is large compared to the corresponding rectangular area 512 . Based on such knowledge, coding technology is applied.

Specifically, as shown on the left side of FIG. 12(B), first, a base area is photographed without the object 500, and the image data 520 is divided into a plurality of (for example, several hundred) partial images. Then, the code for reference corresponding to each is generated, and the reference code group is recorded.

Next, as shown in the center of FIG. 12B, the edge of the object 500 is detected based on the image data 522 obtained by photographing the object 500 placed in the base region. Then, an effective area 524 (a rectangular area indicated by a thick line in the figure) in which the object 500 exists is extracted. As shown in FIG. 12B, the inside of the extracted rectangular region 524 is divided so that each of the upper and lower halves has the same number of partial images, and each partial image is coded. The partial image is divided in the same way as when the reference code group is generated.

Next, when the specific target code group generated in the state where the target object 500 is placed is collated with the reference code group, each code corresponding to the area where the target object exists does not match, and the other areas will match. In FIG. 12(B), regions where the codes match are indicated by diagonal lines. As shown in FIG. 12(C), the number of areas where these codes match is divided into an upper half area 524a and a lower half area 524b of the effective area and counted. In the illustrated example, 24 codes match in the upper half area 524a of the valid area, and 13 codes match in the lower half area 524b. In this case, since the upper half area 524a of the effective area corresponds to the above-described portion a, it is determined to be "above", and since the lower half area 524b of the effective area corresponds to the above-described portion b, it is determined to be "below". can be done.

FIG. 13 is a diagram for explaining application example 4 of the encoding technique according to the above embodiment. Here, an application example is shown in which an object (an example of an animal) in a video (image) is coded in advance, and the reference code group is used to recognize the object existing in the video. ing. Specifically, first, as shown in FIG. 13(A), an object 600 (in this case, an animal) in the image is subjected to edge detection by a known technique, and as shown in FIG. , extract the edge 610 of . Next, as shown in FIG. 13(C), a plurality of rectangular regions are mapped to the extracted edge 610 and its internal region 612 . The number of rectangular regions can be set arbitrarily, and the recognition accuracy can be improved as the number increases. Then, for the partial images of each mapped rectangular area, reference codes corresponding to the respective partial images are generated by the method described above, and the reference code group is recorded. Similarly, a reference code group corresponding to each imaging direction of the target object is generated and recorded. Information related to the object (for example, attribute information such as the name of an animal) is recorded in association with these reference code groups.

Next, for example, from the video captured in real time, edge detection of the object and mapping of the rectangular area are performed in the same manner as described above, and the partial image of each rectangular area is coded. Generates a specific target code group corresponding to . By collating this specific target code group with a pre-recorded reference code group, the target included in the video can be recognized, and attributes such as presence or absence and name of the target can be specified.

By combining the existing edge detection technology and the encoding technology of this embodiment in this way, it becomes possible to perform immediate registration and collation of the code, thereby realizing an improvement in recognition accuracy. The same applies when the object is space or electronic data. The encoding technology of the present embodiment can achieve real-time processing from generation to registration, comparison, collation, and determination, so that recognition accuracy and recognition speed can be improved.

FIG. 14 is a diagram for explaining application example 5 of the encoding technique according to the above embodiment. Here, for example, an example of application to a visual inspection process during manufacture of industrial products is shown. Specifically, first, as shown in FIG. 14A, image data 700 is obtained by photographing the appearance of the product to be inspected. Next, as shown in FIG. 14B, the image data 700 is rotated 360° by arbitrary angles (for example, by 1°) to obtain image data 701, 702, etc. of respective angles. Then, encoding is performed based on the image data of each angle, and each code is recorded as a reference code. Next, as shown in FIG. 14(C), in the actual appearance inspection process, for example, codes are generated from image data 710 to 713 obtained by photographing each product flowing on the belt, and these codes are prepared in advance. Check against the reference code you recorded. As a result, even if the angle of the product arrangement is not constant, if the shape and color tone match, it is judged to be an acceptable product (OK product). (NG product) can be determined. In the illustrated example, the products corresponding to the image data 711 and 713 can be judged as acceptable products, and the products corresponding to the image data 710 and 713 can be judged as defective products.

When generating a reference code in advance in this way, if one good product (product with no defects) is photographed in one arrangement state (angle), multiple images can be identified even if the direction or angle changes by processing in memory. can automatically generate a reference code for In addition, by embedding color information in the code, it is possible to determine not only the shape but also the color tone.

FIG. 15 is a diagram for explaining an application example 6 of the encoding technique according to the embodiment described above. For example, if it is determined in the product visual inspection process that the product will flow in a fixed arrangement, it is not necessary to generate a reference code with a different angle. In this case, one reference code corresponding to non-defective products placed in a fixed arrangement can be used. Here, as an example, let us consider the case of detecting defects such as missing keyboards for personal computers, poor printing, and scratches.

First, as shown in 15A, the external appearance of the product to be inspected is photographed to obtain image data 800, which is coded and recorded as a reference code. At this time, as shown in FIG. 15B, the image data 800 of the target product is divided into a plurality of rectangular areas to obtain partial images 801 corresponding to each of them, and the code for reference corresponding to each partial image 801 is generated. It is preferable to obtain a collection of reference codes. Next, in the actual appearance inspection process, for example, specific target codes (group) are generated from image data obtained by photographing each product flowing on the belt, and these are prerecorded reference codes ( group). As a result, defects such as missing keyboards, poor printing, scratches, etc., can be detected. In this case, as shown in FIG. 15(C), it is possible to specify the region 810 where the defect exists based on the mismatched reference code.

FIG. 16 is a diagram for explaining an application example 7 of the encoding technique according to the embodiment described above. Here, a method for obtaining the shape and area of the object will be described. It is assumed that the object is placed in a fixed imaging area and photographed. In this case, first, the image data of the photographing area is obtained in a state where the target object is not arranged, and this is divided into a plurality of areas, and the partial images of each area are coded. As a result, a reference code group consisting of a plurality of (81 in the illustrated example) codes Cd11, Cd12, . . . Cd99 is obtained as shown in FIG. back.

Next, the image data of the photographing area where the object is arranged is obtained, this is divided into a plurality of areas, and the partial images of each area are coded. As a result, a specific target code group is obtained. Check this against the reference code set above. Then, as shown in FIG. 16(B), an area 901 where the code matches (a non-patterned rectangular area in the figure) and areas 902 and 903 where the code does not match (a patterned area in the figure). rectangular area) is obtained. Regions 902 and 903 where the codes do not match are regions where the object exists. Here, the concordance rate is classified in 10% increments. By connecting regions 902 with a matching rate of, for example, 30% to 40% in the illustrated example, among the regions where the codes do not match, the edge of the object, that is, the shape can be obtained. Further, by obtaining the area within the edge, it is possible to obtain the planar view area of the object. Further, it is also possible to estimate the unevenness of the object based on the portion of the region 903 with different matching rate (here, for example, a matching rate of 60 to 70%) within the edge. There are various ways to obtain the rate of matching between codes, but as a simple example, the number of matching characters is counted by comparing the codes one by one in the order in which they are arranged, and that number is calculated as the total number of characters. A method of dividing by can be used. For example, if the code contains 384 characters and 370 of them match, the match rate is about 96%.

<Second embodiment>
FIG. 17 is a diagram for explaining the encoding method of the second embodiment. Here, image data will be described as an example of data to be coded (target data). The configurations of the code generation device and the user device are the same as those of the first embodiment (see FIGS. 6 and 7), and only the information processing method for code generation and collation is different. Here, the description of the device configuration is omitted, and the information processing method for code generation and collation will be described in detail.

As shown in FIG. 17A, image data as target data (reference data) is divided into a plurality of sampling ranges. In the illustrated example, it is divided into 64 sampling ranges S11, S12, S13, . The total number of pixels included in each sampling range S11-S88 is, for example, 1000-2000. It is assumed that each pixel contains color information in YUV format according to a known format. Here, Y is the luminance value, U(Cb) is the difference between the luminance value and the blue component, and V(Cr) is the difference between the luminance value and the red component. These values are hereinafter referred to as Y value, U value, and V value. This processing is executed by the sampling processing units 20 and 60 .

At this time, the average values of the Y value, U value, and V value are obtained for each of the sampling ranges S11 to S88. The method of obtaining the average value is the same as in the above-described first embodiment. , is obtained by dividing the sum by the total number of pixels in the sampling range S11. The same applies to other sampling ranges. This process is executed by the feature value calculators 21 and 61 .

FIG. 17B shows an example of each average value of the Y value, U value, and V value in each sampling range. Here, the average of the Y values in the sampling range S11 is Y11, the average of the U values is U11, and the average of the V values is V11. represents a value. For example, in sampling range S12, the average values are Y12=119, U12=128, and V12=142.

A code is generated using the average values of the Y, U, and V values in each sampling range obtained as described above. As an example is shown in FIG. 17C, for example, each average value is Y11, Y12, Y13, . . . Y21, Y22, . U12, U13, . . . U21, U22, . . . U87, U88, V11, V12, V13, . Note that this is just an example, and the order of combining the Y value, U value, and V value can be set as appropriate. In this way, a reference code corresponding to image data as target data (reference data) is generated. This processing is executed by the encoding processing units 23 and 63 .

Here, if each value of Y11 to Y88, U11 to U88, and V11 to V88 is represented by a hexadecimal character string, for example, when the sampling range is set to 64 divisions of 8 rows and 8 columns, The data size of the code obtained in 1 is 192 bytes (=64×3). This corresponds to 384 characters of text data. FIG. 18 shows an example of a code expressed in hexadecimal character code. In hexadecimal characters, since the data amount per character is 4 bits, the number of characters is 384, but the data amount when recording is 192 bytes.

According to the studies of the inventors of the present application, this 192-byte data amount is the practical minimum value for real-time comparison, collation, judgment, and recognition without using complicated calculations. That is, the minimum value of an effective code that can achieve a recognition rate of about 98% for visualized object data such as space, landscape, object, heat source video, video, and image is 192 bytes. This data size does not depend on the area or data size of the original video or image.

For example, consider the amount of data when 24 codes are generated per second corresponding to a moving image with a frame rate of 24 fps. Assuming that the data size of the code corresponding to one image is 192 bytes as described above, the number of codes corresponding to a 30-minute moving image is 43200, and the data amount is approximately 7.9 MB. Similarly, the number of codes corresponding to a 1-hour movie is 86400 and the amount of data is about 15.8MB, the number of codes corresponding to a 2-hour movie is 172800 and the amount of data is about 31.6MB, and the number of codes corresponding to a 1-day movie is 172800. The number of corresponding codes is 2073600 and the amount of data is about 379.7 MB. In addition, for each code, related information such as coordinates (longitude, latitude, altitude: 24 bytes), time (year, month, day, hour, minute, second: 8 bytes), and comments for 100 double-byte characters (200 bytes) are provided. Even if they are added, the amount of data corresponding to a 30-minute moving image is about 17.5MB, the amount of data corresponding to a 1-hour moving image is about 34.9MB, and the amount of data corresponding to a 2-hour moving image is about 69MB. 9 MB, the amount of data corresponding to one day's video is about 838.5 MB. Note that the amount of data when one code is generated corresponding to one frame per second is 1/24 of the above. In this case, the amount of data for one day's moving image is very small, 15.8 MB (without related information) or 34.9 MB (with related information).

Also, since this code includes YUV color information, it is possible to obtain RGB format color information by the following conversion formula using this color information.
R = (1.164 x (Y value - 16)) + (1.596 x (V value - 128))
G = (1.164 x (Y value -16)) - (0.391 x (U value -128)) - (0.813 x (V value -128))
B = (1.164 x (Y value - 16)) + (2.018 x (U value - 128))

FIG. 19 is a diagram for explaining a matching method using codes according to the second embodiment. Here, for the sake of simplification, consider a case where a sampling range of 3 rows and 3 columns is set. In the drawing, the reference code generated and recorded in advance is shown on the left side, and the specific target code obtained from the specific target data is shown on the right side. Each code is actually a character string as shown in FIG. 17C, but is expressed here so that the correspondence between sampling ranges can be easily understood. As a matching method, for example, a simple comparison method, a relative comparison method, and the like are conceivable. These processes are executed by the collation processing unit 64 .

First, a simple comparison method will be described. For each sampling range, the difference between the Y value, U value, and V value of the reference code and the specific target code is obtained, and the difference is squared. For example, in the illustrated example, the square of the Y value difference in the sampling range of 1 row and 1 column is (202-195) ² =49. Similarly, the square of the Y value difference in the sampling range of 1 row and 2 columns is (195-212) ² =289. Similarly, the square of the Y value difference in the sampling range of, for example, 3 rows and 3 columns is (149−190) ² =1681. For the other sampling ranges, the squares of the differences of the Y values are obtained in the same manner, and the total value thereof is obtained. In the illustrated example, summing up the squares of the differences of the Y values in each sampling range gives a total value of 15114 (=49+289+...+1681). Likewise, for the U value and the V value, summing up the squares of the differences between the values yields 8694 and 3072, respectively.

When the total values corresponding to the Y, U, and V values are obtained in this way, they are compared with preset reference values, and if each value is equal to or less than the reference value, a reference code and Determine that the specific target codes match. The reference value may be set in consideration of the degree of similarity at which "matching" is to be made, for example, a reference value corresponding to a matching rate of 97% or more.

The reason why the difference is further squared for each of the Y value, U value, and V value is to make it easier to detect the difference between the reference code and the specific target code. Depending on the application, squaring may not be necessary, and on the contrary, the exponent may be increased further, such as by cubing.

Next, a method for relative comparison will be described. First, for the reference code, a value Ya12 is obtained by squaring the difference in Y value between the sampling range of 1st row, 1st column and the sampling range of 1st row, 2nd column. In the illustrated example, Ya12=(202-195) ² =49. Similarly, values Ya13, Ya14, .

Next, a value Ya23 is obtained by squaring the difference in Y value between the sampling range of 1st row, 2nd column and the sampling range of 1st row, 3rd column. In the illustrated example, Ya23=(195-213) ² =324. Similarly, values Ya24, Ya25, . Similarly, values Y34, Y35, .

Similarly below, the sampling range of 2 rows and 1 column and the subsequent sampling range, the sampling range of 2 rows and 2 columns and the subsequent sampling range, the sampling range of 2 rows and 3 columns and the subsequent sampling range, 3 rows and 1 A value obtained by squaring the difference in Y values between the column sampling range and the subsequent sampling range, and the 3-row, 2-column sampling range and the subsequent sampling range is obtained.

Then, all the values obtained by squaring the obtained differences of the Y values are totaled. In the illustrated example, the total value is 47,502. Further, when the U value and the V value are similarly calculated, the total value for the U value is 73256 and the total value for the V value is 0.

Next, when the specific target code is calculated in the same manner as above, the total value of the Y values is 70,898, the total value of the V values is 58,104, and the total value of the U values is 18,272.

Then, when the differences in the total values of YUV are obtained between the reference code and the specific target code, ΔY=23396, ΔU=15152, and ΔV=18272.

When the difference of the total values corresponding to each of the Y value, U value, and V value is obtained in this way, they are compared with a preset reference value, and if each value is equal to or less than the reference value, it is used for reference. Judge that the code and the specific target code match. The reference value may be set in consideration of the degree of similarity at which "matching" is to be made, for example, a reference value corresponding to a matching rate of 97% or more.

As in the simple comparison, the difference between the Y value, the U value, and the V value is further squared so that the difference between the reference code and the specific target code can be detected more easily. Depending on the application, squaring may not be necessary, and on the contrary, the exponent may be increased further, such as by cubing. Also, in the case of relative comparison, in the above example, the difference between a certain sampling range and the subsequent sampling range was obtained, but for each sampling range, the difference is obtained by round-robining all other sampling ranges. may

The code according to the second embodiment as described above can be applied to the application examples variously described in the first embodiment. That is, although there is a difference in the code generation method between the first embodiment and the second embodiment, each application example relates to a method of utilizing the generated code and is not affected by the code generation method. is.

<Third Embodiment>
FIG. 20 is a diagram for explaining the encoding method of the third embodiment. Here, sound data will be described as an example of data to be coded (target data). The configurations of the code generation device and the user device are the same as those of the first embodiment (see FIGS. 6 and 7), and only the information processing method for code generation and collation is different. Here, the description of the device configuration is omitted, and the information processing method for code generation and collation will be described in detail.

In general, the three elements of sound are loudness, pitch, and timbre. Coding of sound data is not a technique for reproducing the sound itself, but a technique for analyzing the sound data in real time to recognize and identify the sound. In this embodiment, among the three elements of sound, coding is performed by paying attention to the pitch.

FIG. 20A is an example of a waveform diagram obtained by performing a discrete Fourier transform on a sound picked up by a microphone and finding a change in frequency over time. Note that the fundamental frequency may be calculated by cepstrum analysis as necessary. For such a waveform diagram, the sampling interval Q is set to 0.2 seconds, for example, and the frequency is extracted for each sampling interval. The extracted frequency values are then coded and recorded. At this time, related data such as the sampling interval Q, the sampling time (2.0 seconds in the illustrated example), and the title of the song are added and recorded. The sound data to be coded may be obtained in real time by picking up sound with a microphone, or may be previously picked up and stored. When the format of the stored sound data is compressed format such as mp3, it is appropriately converted into a non-compressed format such as WAVE format and encoded.

In the above information processing, the sampling processing unit 20 performs data division for each sampling interval, the feature value calculation unit 21 performs frequency calculation, and the encoding processing unit 23 performs encoding. Recording of the code and related information is performed by the code recording processing unit 24 .

FIG. 20B is a diagram showing an example of encoding. 20(A) shows an example in which coding is performed every 0.2 seconds based on the waveform of the sound data shown in FIG. 20(A). For example, when the acquisition timing is 0.2 seconds, the frequency is extracted as 1046.50 Hz, and the corresponding code (denoted as "CEC" in the figure) is obtained as 0416. This note corresponds to "do/C6" in terms of scale names, and corresponds to 64 in terms of keyboard symbols (in the case of an 88-key keyboard). The same applies to other acquisition timings. In this example, the audible range is assumed to be 20 Hz to 20000 Hz at the time of frequency extraction, and decimal points are discarded at the time of encoding. For example, in the case of 20.2 Hz, the value after truncating the decimal point is "20", which is "14" when coded in hexadecimal. Similarly, in the case of 261.62 Hz, the value after truncating the decimal point is "261", which is "105" when coded in hexadecimal. Similarly, in the case of 20000.0 Hz, the value after truncating the decimal point is "20000", which is "4e20" when coded in hexadecimal. Therefore, the code generated in this embodiment can be represented by a maximum of 4 hexadecimal characters, so the maximum data size is 2 bytes.

FIG. 21 is a diagram for explaining a matching method using codes according to the third embodiment. As shown in FIG. 21(A), a code corresponding to a part (2000 ms) of a musical piece (Twilight Message) is obtained in advance by setting the sampling interval Q to 0.2 seconds. Recorded as a group. To simplify the explanation, the codes are represented by letters A to G in the figure, and the codes represented by the same letters are assumed to have the same contents. On the other hand, the specific target data of each piece of music given in real time (or sound data) is coded and compared with the reference data group. In the illustrated example, since there is no part in which the reference data group and the code arrangement match each of the song (1) and the song (3), it can be determined that the song is not specified by the reference data group. On the other hand, since song (2) has a portion whose code arrangement matches that of the reference data group, it can be determined that song (2) is the song (Twilight Message) corresponding to the reference data group. Regarding the match/mismatch between the codes at each sampling interval Q, it may be judged that the code is completely the same as "match", or within a certain error range (for example, an error range corresponding to ±10 Hz). ), it may be judged as “match”.

In the above information processing, data division at each sampling interval is performed by the sampling processing unit 60, frequency calculation is performed by the feature value calculation unit 61, and encoding is performed by the encoding processing unit 63. A collation processing unit 64 collates the code with the reference data group.

Here, when encoding the sound data, it is more preferable to set the sampling interval in consideration of the tempo, especially when the object is the sound data of a song. For example, in the case of a piece of music with a tempo (BPM) of 120, 120 quarter notes (beats) are included in one minute. In this case, the interval of one beat, which is two beats per second, is 0.5 seconds. In this case, by setting (adjusting) the sampling interval in accordance with the tempo of the sound data, it is possible to perform coding corresponding to the timing of the notes. Specifically, let Q be the sampling interval, B be the tempo, and X be the beat length of the target note. Here, the beat length of a note is set such that X=4 for a quarter note, and X=8 for an eighth note, for example. At this time, as shown in FIG. 21B, the sampling interval Q can be expressed as Q=(1/((B/60)*(X/4)))*1000. By setting the sampling interval Q using this relational expression and encoding the sound data, the accuracy of determining the similarity of music can be further improved. The lower column of FIG. 21(B) exemplifies specific values of the preferred sampling interval Q. FIG.

By encoding sound data in this way, for example, the user device 2 can display the title of a song played in space, or can display AR content (augmented reality content), moving images, or still images matching the song. Such control can be realized. In addition, regarding the so-called music plagiarism problem, it is possible to present the similarity as an objective judgment rather than a human subjective judgment. In addition, given the insight that, in general, all living things have different heart sounds, and even twins have different heart sounds, it is possible to perform person authentication based on heart sounds (heart sound authentication) by coding heart sounds. . For example, it can create unprecedented use cases, such as enabling person authentication at gates such as airports.

According to each of the above-described embodiments, a novel technique for encoding target data such as images and sounds with a small amount of data is provided. As a result, a new technology is provided that enables matching between specific target data and reference data with a low load and a short processing time.

That is, according to each embodiment, for example, an image captured by a camera is divided into an arbitrary number, and each block (sampling range) divided into a plurality of blocks is assigned a unique code based on information such as the light source, luminance, and color. is generated and recorded, and this code can be used for high-speed (or real-time) comparison, verification, and authentication. By using this peculiar code, it is also possible to realize transmission of invisible information such as heat source, sound, and X-rays and information of the five senses with a smaller data size. As shown in the conceptual diagram of FIG. 22, a unique code according to each embodiment is added to what is currently established and recognized as data such as spatial information, moving images, still images, temperature and sound. This not only creates a connection with the real world, but also creates new ways of using data, and it is possible to add further value and use methods to existing data that has already been recorded. For example, in addition to GPS information (positional information), direction, angle, temperature, pressure, etc., there are already various types of accompanying information such as sources and URLs on big data. However, when comparing and collating with the real world in real time, the "situation", "state", and "spatial data" necessary for judgment are missing, and the current situation is that the necessary linkage is not possible. On the other hand, by using the code according to each embodiment to create a hand that connects with reality called "Earth Data", it will play a role as a "key" and a trigger of the link between the real world and the system. can be done. Since the data size of the code according to each embodiment is very small, a new concept of "byte chain" is provided.

It should be noted that the present invention is not limited to the content of each embodiment described above, and various modifications can be made within the scope of the gist of the present invention. The utilization method and application method of the code generated by each embodiment are not limited to the contents described above. Also, the numerical values and the like given in each embodiment are examples and are not limited to these. In addition, in the first embodiment, the relative difference between each sampling range is obtained, but this may be omitted and the average luminance value of each sampling range may be concatenated as character string data.

In each of the above-described embodiments, character string data obtained from average values or relative differences are directly concatenated to generate a code, but the aspect of "concatenation" in the present invention is not limited to this. For example, each average value or each relative difference may be multiplied by a constant coefficient, and may be concatenated as character string data. Further, each character string data may be concatenated in such a manner that specific data (a delimiter character or the like) is interposed between the character string data corresponding to each average value or the like. Alternatively, character string data corresponding to each average value or the like may be converted to other character string data under a certain rule and concatenated. That is, any mode of connection is included in the concept of "connection" in the present invention as long as the mutual relative relationships such as average values do not collapse or can be restored.

10, 50: information processing unit 11, 51: camera 12, 52: microphone 13, 53: storage device 14, 54: operation device 15, 55: display device 20, 60: sampling processing unit 21 , 61: feature value calculation unit, 23, 63: encoding processing unit, 24: code recording processing unit, 64: matching processing unit

Claims (3)

A method of generating and collating a code according to the contents of target data using an information processing device,
a first step of obtaining a reference code obtained according to the reference data;
a second step of generating specific target code corresponding to specific target data;
a third step of comparing said reference code and said specific target code to determine their match/mismatch;
including
The reference code in the first step is
dividing the reference data into a plurality of first sampling ranges;
For at least one type of the first element data out of one or a plurality of types of the first element data, which are the first element data included in each of the plurality of first sampling ranges and each is represented by a numerical value obtaining an average value of the first element data for each first sampling range;
generating the reference code corresponding to the reference data by concatenating, as character string data, the average value of the first element data for each of the first sampling ranges or numerical values corresponding to a predetermined number of upper digits thereof; When,
is obtained in advance through
The specific target code in the second step is
dividing the specific target data into a plurality of second sampling ranges set in the same manner as each of the first sampling ranges;
At least one of one or a plurality of types of second element data included in each of the plurality of second sampling ranges, each of which is represented by a numerical value, and is the first element data A step of obtaining the average value of the second element data for each second sampling range for the second element data of the same type as
A step of generating the specific target code corresponding to the specific target data by concatenating the average value of the second element data for each of the second sampling ranges or a predetermined number of upper digits thereof as character string data. When,
is generated through
The third step is
Between each of the first sampling ranges and each of the second sampling ranges, the average value for each of the first sampling ranges obtained based on the reference code or a numerical value of a predetermined number of digits from the higher order a step of obtaining a difference between an average value for each of the second sampling ranges obtained based on the specific target code or a numerical value corresponding to a predetermined number of higher digits;
determining that the reference code and the specific target code match when each of the differences falls within a predetermined allowable range;
Data matching methods, including A method of generating and collating a code according to the contents of target data using an information processing device,
a first step of obtaining a reference code obtained according to the reference data;
a second step of generating specific target code corresponding to specific target data;
a third step of comparing said reference code and said specific target code to determine their match/mismatch;
including
The reference code in the first step is
dividing the reference data into a plurality of first sampling ranges;
For at least one type of the first element data out of one or a plurality of types of the first element data, which are the first element data included in each of the plurality of first sampling ranges and each is represented by a numerical value obtaining an average value of the first element data for each first sampling range;
generating the reference code corresponding to the reference data by concatenating, as character string data, the average value of the first element data for each of the first sampling ranges or numerical values corresponding to a predetermined number of upper digits thereof; When,
is obtained in advance through
The specific target code in the second step is
dividing the specific target data into a plurality of second sampling ranges set in the same manner as each of the first sampling ranges;
At least one of one or a plurality of types of second element data included in each of the plurality of second sampling ranges, each of which is represented by a numerical value, and is the first element data A step of obtaining the average value of the second element data for each second sampling range for the second element data of the same type as
A step of generating the specific target code corresponding to the specific target data by concatenating the average value of the second element data for each of the second sampling ranges or a predetermined number of upper digits thereof as character string data. When,
is generated through
The third step is
Between each of the first sampling ranges and each of the second sampling ranges, the average value for each of the first sampling ranges obtained based on the reference code or a numerical value of a predetermined number of digits from the higher order a step of obtaining a difference or a power of the difference from an average value for each of the second sampling ranges obtained based on the specific target code or a numerical value corresponding to a predetermined number of upper digits thereof;
a step of determining that the reference code and the specific target code match when the total value of the differences or the total value of the powers of the differences are all equal to or less than a predetermined reference value;
Data matching methods, including each said first sampling range and each said second sampling range are divided by the same width and height or the same interval;
3. The data matching method according to claim 1 or 2.

Images