JP2000182050A - Individual identifying device for animal - Google Patents

Abstract

PROBLEM TO BE SOLVED: To perform highly accurate individual identification of an animal. SOLUTION: Contour data of an iris grain registered in a contour dictionary storage part 3 is geometrically converted into coordinate strings in an inputted image from an eye area image pickup part 1 by a contour geometrical conversion part 2. Edge intensity of the coordinate strings which is geometrically converted by the contour geometrical conversion part 2 is compared with a determined required value and when the edge intensity is larger than the required value, the contour of the iris grain in the inputted image is regarded to be detected by an edge intensity calculating part 4. Average of difference between data of an area in the contour calculated by the edge intensity calculating part 4 and registration data registered in a density dictionary storage part 6 is calculated and a degree of difference between the inputted image and the dictionary is calculated by an average difference calculating part 5. The individual identification is performed by comparing the degree of difference calculated by the average difference calculating part 5 with the required value by an identifying part 7.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to, for example, (race)
TECHNICAL FIELD The present invention relates to an animal individual identification device for registering data for individual identification of animals and performing collation in an institution requiring individual management of animals such as horses and (pastoral) cattle.

[0002]

2. Description of the Related Art Individual identification of thoroughbreds at a racetrack, auction field, or a breeding ranch is carried out under the initiative of a pedigree control agency. The current method of identifying individuals in Japan is based on coat color, vitiligo on the limbs, vitiligo on the head, and curl. In the world, there are other methods based on blood type, branding, tattooing, etc. Reference: "Knowledge of Horses", II-
3. How to identify a horse, P. 153-P. 154).

[0003] In the case of individual identification of cattle, a method using identification nameplates such as collars and ear tags and branding and tattooing has been adopted in general ranches.

[0004] As an individual identification method of an animal that has recently attracted attention, there is a technique using an MC (Micro Chip). This is a chip with a built-in microminiature integrated circuit enclosed in a glass tube, which is implanted in the body of an animal with an injector such as a syringe.In the case of identification, the embedded signal is traced by a non-contact detector to trace the output signal. This is used as information for individual identification.

[0005]

However, the above-mentioned prior art has the following problems to be solved.
In the case of discrimination based on coat color, white mark, vitiligo, and curl, there are many horses with few features and horses with the same feature, so that the identification of the horse is often incomplete. In addition, there is a risk that the branding and tattoos may disappear or be tampered with, and improvements in animal welfare such as pain and local suppuration received by animals have been desired. In addition, although individual identification based on blood type is accurate, there is a disadvantage that a considerable processing time is required for determination and the cost is high.

[0006] The identification nameplate has a risk of being damaged, lost or stolen, and in the case of branding or tattooing, it has a risk of disappearing or being tampered with like a horse. ing. In particular, even if this falsification is revealed, it is difficult to prove where the person originally belonged.

[0007] The MC method can be used semi-permanently once it is embedded, and also has advantages such as high convenience. However, the operability of implantation in living animals, pain given to animals at the time of implantation, local reactivity such as swelling, tenderness, suppuration, etc., motor dysfunction and clinical abnormalities in animals, movement of MC in vivo There are many issues in terms of performance, operability of detector, change in detection sensitivity, stability, and reliability. And above all, from the aspect of animal welfare, there are animal officials who are reluctant to adopt the MC method,
A convenient identification method that replaces the MC identification method has been desired.

[0008]

The present invention employs the following structure to solve the above-mentioned problems. <Configuration 1> A contour geometric transformation unit for geometrically transforming the contour data of the iris granules registered in advance into a coordinate sequence in the input image, and the edge strength of the coordinate sequence geometrically transformed by the contour geometric transformation unit is compared with a predetermined required value. An edge intensity calculation unit that considers that the contour of the iris granule in the input image has been detected if it is larger than a required value; and an area within the contour of the iris granule detected by the edge intensity calculation unit as an iris granule region candidate. An animal individual identification device, comprising: an identification unit configured to identify a candidate with an area of an iris granule registered in advance and identify an individual.

<Structure 2> An area dividing unit that divides an iris granule area into a plurality of areas, and an area geometric transformation unit that geometrically transforms each area for each iris granule area divided by the area dividing unit.
An individual identification device for an animal, comprising: an identification unit that performs individual identification by comparing each region with a value of an iris granule registered in advance for each region geometrically transformed by the region geometric transformation unit.

[0010]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the present invention will be described below in detail with reference to specific examples. The present invention is an individual identification apparatus characterized by registering contour and density information of iris granules and performing collation processing in order to solve a conventional problem in individual identification of an animal. Hereinafter, in each specific example, a case where a target animal is a horse in particular will be described as an example.

FIG. 2 shows a front view of the eyeball of a horse. As shown in FIG. 2, the horse's eye mainly includes a pupil 10, an iris (iris) 11, and an iris granule 12 between upper and lower eyelids 13.

A transparent hemispherically raised cornea is provided in front of the pupil 10 and the iris 11. The major difference between horse eyes and human eyes is that the pupil 10 is elliptical,
It is to have iris granules 12 unique to horses and ruminants.

Light from the outside passes through the pupil 10 and reaches the retina at the back of the pupil. The iris 11 is a muscle surrounding the pupil 10 and has a function of controlling the amount of light incident on the pupil 10 by contracting and expanding. The iris granules 12 are located between the iris 11 and the pupil 10 and have a shape in which hemispherical raised granules are connected. The iris granules 12 have a black color rich in melanin pigment, and are said to have a function of absorbing dazzling daylight even when the iris 11 contracts.

It has not been well known that the iris granules 12 have various sizes and shapes depending on the horse and the right and left eyes of the same horse and have individual differences. In the present invention, an individual is identified by photographing the iris granule 12 with a camera and collating it with data registered in advance.

In order to perform the matching with the registered data with high accuracy, it is effective to perform the matching after extracting the area of the iris granules. For that purpose, important information is a density difference at a boundary (hereinafter referred to as an edge intensity) caused by a difference in shading / organism between an extraction area and a surrounding area. The following documents show examples of pupil extraction using edge strength. Document name "Sakaguchi, Nakano, Yamamoto,"Driver's gaze detection while driving a car ", IEICE PRU Technical Report, PRU95-28"

However, when the eyes are imaged in the bright sunshine during the day, the surroundings are brighter than in the cornea, so that a specular reflection phenomenon occurs in which the surrounding scene is reflected on the corneal surface. When this specular reflection image overlaps the boundary of the extraction area, the density difference becomes dull,
As a result, there is a problem that a region is lost due to an edge detection error and accurate matching cannot be performed.

To solve this problem, the effect of the specular reflection image can be reduced by irradiating stronger illumination. However, there is a limitation in enhancing the illumination from the viewpoint of eye safety, production cost, and the like. It is difficult to erase.

In the present invention, attention is paid to the fact that the outline of a registered image (hereinafter, referred to as a dictionary) is accurately extracted, and this problem is solved by extracting an iris granule region using outline data of the dictionary. . Specifically, the position, size, and inclination of the contour data of the dictionary are geometrically transformed by various transformation parameters, and the iris granule contour of the input image is searched. By calculating the sum of the edge intensities on the coordinates of each of the converted outline dictionaries, a desired area can be detected because the value increases when overlapping with the iris granules of the input image. According to this method, even if there is an area where the edge strength is weak locally due to specular reflection, the value of the sum with the edge having a strong strength is averaged, so that it is possible to stably detect the sum and fix it. By using the shape region, a region without loss can be extracted. Hereinafter, the animal individual identification device of the present invention will be described in detail using specific examples.

<< Specific Example 1 >> In the specific example 1, the contour data of the iris granules are geometrically transformed into coordinates in the input image, and the geometrically transformed edge strength is compared with a predetermined value. The iris granule region candidate is regarded as being detected.

<Structure> FIG. 1 is a diagram showing the structure of a specific example 1 of the animal individual identification apparatus of the present invention. Specific example 1 includes an eye area imaging unit 1, a contour geometry conversion unit 2, a contour dictionary storage unit 3, an edge strength calculation unit 4, an average difference calculation unit 5, a density dictionary storage unit 6, an identification unit 7, and an identification result output unit 8. Consisting of The specific configuration of each unit will be described below.

The eye area imaging section 1 is composed of an image input section and a light source section, and plays a role of converting an image of a horse eye into an electric signal and inputting it to the apparatus. Here, the image input unit is a lens, C
It is composed of a CD sensor, a video signal output circuit and the like, and performs a process of converting analog optical information of an image of a horse eye into a digital electric signal by a CCD element. The light source unit is illumination composed of a power supply, a lamp, and the like, and is used to catch the horse's eyes with good contrast. In order to prevent the horse from dazzling and moving excessively, light of a wavelength such as red light, near-infrared light, or infrared light that is hardly perceived by animals is used as a light source.

The contour geometric converter 2 is a coordinate converter.
It has a function of converting a two-dimensional coordinate sequence of a contour into a predetermined position, size, and inclination. That is, the contour data (= contour shape) stored in the contour dictionary storage unit 3 is geometrically converted into various shapes, and these coordinates are converted to the coordinates in the input image captured by the eye area imaging unit 1. It has a function to convert.

The outline dictionary storage unit 3 includes a recording medium such as a semiconductor memory, a hard disk, and a magnetic tape;
This is storage means including a control circuit and the like, and a two-dimensional (x, y) coordinate sequence of a contour is registered for each horse ID.

The edge strength calculator 4 comprises means for calculating the average difference between adjacent pixel values of the image representing the edge strength and means for comparing the strength with a required value, and calculates the edge strength on the dictionary contour coordinates in the input image. Has a function of comparing the obtained edge intensity with a required value to determine whether or not to be a candidate for the iris granule region.

The average difference calculator 5 has a dictionary data reading function, a difference calculation function between images, and an accumulation function, and calculates the degree of difference between the input image and the density dictionary for the iris granule area.

The density dictionary storage unit 6, like the contour dictionary storage unit 3, is a storage unit including a recording medium such as a semiconductor memory, a hard disk, and a magnetic tape, a recording head, a control circuit, and the like. Attributes such as the concentration value of granules and the name, age, and sex of the horse are registered.

The discriminating section 7 comprises means for comparing the average difference obtained by the average difference calculating section 5 with a required value, and has a function of determining accept / reject of the input image and a function of determining the end of repetition of the search. ing.

The identification result output unit 8 is a display / output unit such as a display, a printer, and a speaker, and has a function of presenting the identification result and the attributes of the horse to the user.

In the above configuration, the contour geometric conversion unit 2
The identification unit 7 is realized by a microprocessor executing a program corresponding to each function unit. Further, such a program is provided by being recorded on a recording medium.

<Operation> FIG. 3 is an explanatory diagram showing the operation of the first embodiment. Correspondence with FIG. 1 is as follows: Step S1 is the eye area imaging unit 1, Steps S2 and S3 are the contour geometric conversion unit 2, Steps S4 and S5 are the edge strength calculation unit 4, Step S6 is the average difference calculation unit 5, Step S7 and Step S7. S8 corresponds to the identification unit 7, and steps S9 and S10 correspond to the identification result output unit 8, respectively. Hereinafter, each operation step will be described in order.

In step S 1, a user who identifies a horse individually photographs the horse's eyes with the eye area imaging unit 1. The light source unit of the eye area imaging unit 1 illuminates the horse's eyes with illumination using red light, near-infrared light, and infrared light as a light source, and the reflected light from the horse's eyes, that is, the image, is two-dimensional by a CCD sensor. It is converted into an electrical signal that is an array. This is because a two-dimensionally arranged photoelectric conversion element (CCD) exposes reflected light of an imaging target for a certain period of time, and quantizes a voltage impressed by a photoelectric conversion effect at a certain gradation. . The input light is output as a digital electric signal. This two-dimensional array of electric signals is an “image”.

In step S2, a geometric conversion variable is set. The geometric conversion variables are variables (dx, dy), p, and θ representing the amount of conversion of the position, size, and inclination of the contour data registered in the contour dictionary storage unit 3. As described above, since the principle of the present invention is to search the position of the contour of the iris granule in the input image using the outline coordinate sequence of the dictionary, the search is performed by setting this conversion amount to various values.

In step S3, first, the outline dictionary storage unit 3
Read the contour dictionary from. The outline dictionary is an array of coordinate values (x, y) on the image, and variables (dx, d) representing the amount of conversion of the position, size, and inclination set in step S2.
y), p, and θ are used to convert to new coordinate values. The conversion formula uses the coordinate values of the dictionary as (x, y), and the converted coordinate values as (X
t, Yt), it is expressed by the following equation.

Xt = (xcos θ−ysin θ) p + dx (1) Yt = (xsin θ + ycos θ) p + dy (2)

In step S4, the edge strength e of the geometric transformation coordinate sequence of the dictionary of the input image obtained in step S3 is obtained. Various methods can be applied to the present invention to calculate the edge strength e. For example, the absolute value of the difference between the upper and lower adjacent pixels in the y-axis direction (e = | P (Xt, Yt-1) -P
(Xt, Yt + 1) |: P (x, y) is a method of obtaining the edge intensities of all the contour points using coordinates (x, y) pixel values) and the like and calculating the average.

The edge strength e obtained in step S4 should show a large value when the coordinate value of the outline dictionary overlaps with the outline of the iris granules of the input image. Therefore, in step S5, the edge intensity e is compared with a predetermined required value Te. If the intensity e is larger than the required value Te (e> Te), it is determined that the contours of the iris granules overlap and the matching process from step S6 is performed. If not (e ≦ Te), step S
Proceed to the next search step via 8.

In step S6, first, the density dictionary storage unit 6
, The density value of the pixel of the iris granule is read, and the degree of dissimilarity is calculated for a region within the outline of the input image. Although various methods of calculating the degree of difference between images are applicable to this specific example, the average difference d of the density, which is the average of the absolute value of the difference for each corresponding pixel, is used.

The average difference d obtained in step S6 is a measure of the degree of similarity between the area in the input image and the dictionary. If the eyes are of the same horse, the value is small. In this case, the value increases because of individual differences. Therefore, the same eye and other eyes can be distinguished from each other by a predetermined threshold value Td. Therefore, in step S7, if the average difference d is smaller than the required value Td (d <Td), the iris granules proceed to the accepting process in step S10 assuming that the iris granules are eyes of the same horse, and if not (d ≧ Td), the process proceeds to step S10. S8
To the next search step.

In step S8, the number of repetitions of the processing from step S1 to step S8 is counted. If the number of times exceeds a predetermined number, the search processing is terminated, and the flow proceeds to reject processing in step S9. If the number is equal to or less than the predetermined number, the process proceeds to step S2 again to execute a search process.

In step S9, the rejection is output to the display device to notify the user. In step S9, the fact that the horse has been accepted and the ID information of the horse that has been accepted are output to the display device to notify the user. The above is the description of the operation of this specific example.

<Effects> As described above, according to this embodiment, the following effects can be expected. ● Since the iris granule region in the input image is searched using the sum of the edge intensities of the contour dictionary, it is possible to reduce the extraction error due to local edge rounding by the corneal specular reflection image and extract the iris granule region without loss It is possible. Therefore, highly accurate individual identification can be performed.

<< Specific Example 2 >> The iris granules are granular objects located near the iris near the pupil. The iris plays a role of so-called mydriasis and miosis, in which the iris expands and contracts according to the surrounding brightness in order to keep the amount of light incident on the pupil constant.
As a result, mydriasis / miosis causes nonlinear expansion / contraction of the iris margin, which may cause the iris granules to be distorted into different shapes depending on the pupil size. In particular, when there are a plurality of iris granules which are independently connected to the iris margin, the distance between the plurality of iris granules changes due to the expansion and contraction of the iris, which is remarkable.

As described above, since the iris granules are distorted non-linearly due to pupil change, even if the iris granules having a specific pupil size are used as a dictionary, even if the iris granules are linearly transformed, there is a problem that the collation regions do not match and an erroneous rejection occurs.

On the other hand, in the specific example 2, when there are a plurality of iris granules, or when there are easily deformable parts in the iris granules, they are divided and geometrically transformed to calculate the dissimilarity. The problem of the poor matching is solved by searching for the position where is the smallest.

<Structure> FIG. 4 is a diagram showing the structure of the second embodiment. As shown in FIG. 4, the specific example 2 is different from the eye area
1, a region dividing unit 102, a region geometric transforming unit 103, a dissimilarity calculation unit 104, a storage unit 105, an identification unit 106, and an identification result output unit 107. The specific configuration of each unit will be described below.

The eye area imaging section 101 is composed of an image input section and a light source section, and has a role of converting a horse eye image into an electric signal and inputting it to the apparatus. Here, the image input unit includes a lens, a CCD sensor, a video signal output circuit, and the like, and performs a process of converting analog optical information of an image of a horse eye into a digital electric signal by a CCD element. The light source unit is illumination composed of a power supply, a lamp, and the like, and is used to catch the horse's eyes with good contrast. In order to prevent the horse from dazzling and moving excessively, light of a wavelength such as red light, near-infrared light, or infrared light that is hardly perceived by animals is used as a light source.

The region dividing unit 102 is a means for dividing the iris granule region, and includes an iris granule region extracting function, a histogram creating function, and a region dividing function after obtaining an eye image from the eye region imaging unit 101.

The area geometric conversion unit 103 is a geometric conversion means for each divided area, and has a function of setting a conversion variable and a function of geometric conversion.

The dissimilarity calculation unit 104 is a means for calculating the dissimilarity for each area, and has a function of reading the density dictionary from the storage unit 105 and a function of calculating the dissimilarity between the input image and the image of the dictionary.

The storage unit 105 is a storage means including a recording medium such as a semiconductor memory, a hard disk, a magnetic tape, and the like, a recording head, a control circuit, and the like. Attribute is registered.

The discriminating section 106 comprises means for comparing the degree of difference obtained by the degree of difference calculating section 104 with a required value, and comprises a function of judging accept / reject of an input image and a function of judging the end of repetition of search.

The identification result output unit 107 includes a display,
It is a display / output device such as a printer and a speaker, and has a function of presenting the identification result and the attributes of the horse to the user.

In the above configuration, the area dividing unit 102
The identification unit 106 is realized by a microprocessor executing a program corresponding to each function unit.
Further, such a program is provided by being recorded on a recording medium.

<Operation> FIG. 5 is an explanatory diagram for explaining the operation of the second embodiment. The correspondence with FIG. 4 is that step S101 is the eye region imaging unit 101, steps S102 to S104 are the region division unit 102, steps S105 and S106 are the region geometric conversion unit 103, and step S107 is the dissimilarity calculation unit 10
4. Steps S108 and S109 correspond to the identification unit 106, and steps S110 and S111 correspond to the identification result output unit 107, respectively. The operation steps will be described below in order.

In step S101, a user who individually identifies a horse takes an image of the horse's eyes with the eye area imaging unit 1 of the individual identification apparatus of the present invention. The light source unit of the eye area imaging unit 1 illuminates the horse's eyes with illumination using red light, near-infrared light, and infrared light as a light source, and the reflected light from the horse's eyes, that is, the image, is two-dimensional by a CCD sensor. It is converted into an electrical signal that is an array. This is because a two-dimensionally arranged photoelectric conversion element (CCD) exposes the reflected light of the imaging target for a certain period of time and quantizes the impressed voltage with a certain gradation by the photoelectric conversion effect. The input light is output as a digital electric signal. This two-dimensional array of electric signals is an “image”.

In step S102, the region of the iris granule is extracted. An example of region extraction will be described below, but the present invention is not dependent on this, and various other region extraction methods can be applied to the present invention.

FIG. 6 is an explanatory diagram of the area extracting operation in the second embodiment. An image obtained from the eye area imaging unit 101 includes an eyelid 13, a pupil 10, an iris 11, and an iris granule 1.
Contains 2. The region extraction step S102 aims to extract an iris granule region from these by image processing. Since the pupil 10 is hollow, there is little reflected light due to illumination. Therefore, the pupil 10 is a uniform region having a low density. In contrast, the iris surrounding the pupil 10
Since the iris granules have a higher density, by setting a certain threshold in the density, the pixel value of the image is lower than the threshold,
A series of regions whose area is equal to or larger than a predetermined value
Can be specified as

Next, the contour of the iris granule 12 is determined. Pupil 1
Above 0, there is an iris granule 12, on which an iris 11 is located. The change in density at the boundary between organs is larger than the change inside each organ. Therefore, in the image subjected to the density change detection process represented by the Sobel operator, the pixel value of the boundary has a large value.
Therefore, the center of gravity C of the previously specified pupil region is first obtained, and then the pixel whose pixel value is equal to or more than a predetermined threshold value is searched starting from the center of gravity C and ascending upwards. Is detected, and then a point P2 on the boundary between the iris granules 12 and the iris 11 is detected.

According to the same procedure, a horizontal line L passing through the center of gravity C
For the other points above, if the upward search is performed starting from the point adjacent to the center of gravity C as the starting point, the lower edge and the upper edge are obtained for the entire region of the iris granule. The left and right ends of the upper and lower edges of the iris granule thus obtained are obtained as points E1 and E2 at which the vertical distance between the upper and lower edges is smaller than a required small value. Since the contours of the upper edge and the lower edge of the iris granule 12 can be extracted in this manner, the inside can be specified as the region of the iris granule.

In step S103, step S102
A histogram of the number of pixels and a density histogram are obtained for the region extracted in (2) (see FIG. 7).

FIG. 7 is an explanatory diagram of the area dividing operation. Here, the pixel number histogram is obtained by calculating the number of pixels in the vertical direction of the image for each horizontal pixel, and the value of the histogram increases as the position of the vertical width increases. The density histogram is obtained by calculating the sum of the densities of the area pixels in the vertical direction of the image for each horizontal pixel, and the value of the histogram becomes larger at a position where the vertical width is large and the density is bright.

In step S104, step S103
The area is divided using the pixel number histogram and the density histogram obtained in step (1). Here, the dividing method is to provide a vertical separation line in the iris granule region only when the value of the pixel number histogram is equal to or less than a predetermined required value and the density histogram is also equal to or less than a separately required value. It is divided into left and right. This focuses on the point that the connection portion where a plurality of iris granules are continuous or the position where the single iris granule undergoes the largest deformation is narrow and the density value due to the shadow of illumination is small.

In step S105, a geometric transformation variable is set. The geometric conversion variable is a variable (dx,
dy), p, and θ, which are set for each of the divided areas. As described above, since the principle of the present invention is to search for an iris granule area for each divided area, the search is performed by setting the conversion amount to various values.

In step S106, step S105
Each of the divided areas is geometrically transformed using the transformation variables set in the step (1). The divided area is an array composed of coordinate values (x, y) on the image, and new variables (dx, dy), p, and θ representing the amounts of position, size, and inclination conversion set in step S105. Convert to coordinate values. The conversion formula is the same as that of the first embodiment. The coordinate value of the area is (x, y), and the coordinate value after the conversion is (X
t, Yt), it is expressed by the following equation.

Xt = (xcos θ−ysin θ) p + dx (3) Yt = (xsin θ + ycos θ) p + dy (4)

In step S107, the density dictionary is read from the storage unit 105, and the degree of difference from the divided area obtained in step S106 is individually calculated for each area (d1, d2,..., D1).
dn: n is the number of divisions, and their average (d = (d1 + d
2+,... + Dn) / n). Various methods can be applied to the present invention for calculating the degree of difference.
Similar to step S6 in the operation explanatory diagram, there is a method of averaging the absolute values of the differences between the density values of the corresponding pixels.

In step S108, step S107
It is determined whether or not the acceptance is an inspection target using the difference obtained in step (1). The degree of difference is a measure of the degree of similarity between the region in the input image and the dictionary, and the value is small for eyes of the same horse, and there is an individual difference for eyes of another horse. The value increases. Therefore, the same eye can be distinguished from other eyes by a predetermined threshold Td. Therefore, in step S108, when the average difference d is smaller than the required value Td (d <Td), the iris granules proceed to the accepting process in step S111 assuming that the iris granules are eyes of the same horse, and when not (d ≧ Td), the process proceeds to step S111. The process proceeds to the next search step via S109.

In step S109, the number of repetitions of the processing from step S105 to step S108 is counted. If the number of repetitions exceeds a predetermined number, the search processing is terminated, and the flow proceeds to reject processing in step S110. If the number is equal to or less than the predetermined number, the process returns to step S105 to execute the search process.

In step S110, the fact of rejection is output to the display device to notify the user. Step S111
Then, information indicating that the horse has been accepted and the ID information of the horse that has been accepted are output to the display device to notify the user.

<Effects> As described above, according to this example, the following effects can be expected. ● Since the area is divided based on the position where the iris granules are easily deformed, and the geometric transformation is performed individually and collation is performed, it is possible to search for the area following the non-linear deformation due to pupil change, and the collation error is reduced. Can be remedied.

<< Usage Form >> In the description of the first embodiment, the “average” of the difference between adjacent pixels is used to calculate the edge on the dictionary contour coordinates in the input image. Various edge strength calculation methods can be applied to the present invention, such as using together a “standard deviation” indicating the degree of dispersion of values in order to prevent a strong edge from being erroneously detected.

Further, in the above specific examples 1 and 2, the case of a horse was described as an example of an animal, but other animals such as a cow can be similarly applied.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of a specific example 1 of an animal individual identification device of the present invention.

FIG. 2 is an explanatory diagram of an image of a horse's eye.

FIG. 3 is an explanatory diagram showing an operation of a specific example 1 of the animal individual identification device of the present invention.

FIG. 4 is a configuration diagram of a specific example 2 of the animal individual identification device of the present invention.

FIG. 5 is an explanatory diagram showing an operation of a specific example 2 of the animal individual identification device of the present invention.

FIG. 6 is an explanatory diagram of an area extracting operation of a specific example 2 of the animal individual identification device of the present invention.

FIG. 7 is an explanatory diagram of an area dividing operation of the specific example 2 of the animal individual identification device of the present invention.

[Explanation of symbols]

 2 contour geometric conversion unit 3 contour dictionary storage unit 4 edge strength calculation unit 5 average difference calculation unit 6 density dictionary storage unit 7, 106 identification unit 102 region division unit 103 region geometric conversion unit 104 difference degree calculation unit 105 storage unit

Continued on the front page F term (reference) 4C038 VA07 VB04 VC05 5B043 AA09 BA01 DA05 GA02 5L096 AA03 BA03 BA18 CA02 CA14 EA26 FA06 FA32 GA08 HA09 JA03

Claims (2)

[Claims] 1. A contour geometric transformation unit for geometrically transforming contour data of iris granules registered in advance into a coordinate sequence in an input image, and an edge strength of the coordinate sequence geometrically transformed by the contour geometric transformation unit is set to a predetermined required value. Comparing an edge intensity calculation unit that considers that the contour of the iris granule in the input image has been detected when it is larger than a required value; and an area within the contour of the iris granule detected by the edge intensity calculation unit as an iris granule region candidate. An individual identification device for an animal, comprising: an identification unit that identifies the individual by comparing the area candidate with an area of an iris granule registered in advance. 2. An area dividing unit that divides an iris granule area into a plurality of areas; an area geometric transforming unit that geometrically transforms each area for each iris granule area divided by the area dividing unit; An individual identification device for an animal, comprising: an identification unit that performs individual identification by collating each region with a value of an iris granule registered in advance for each region geometrically transformed by the conversion unit.