KR20160092374A - System and method of automatically calculating an amount of coin - Google Patents

Abstract

The present invention relates to a method for distinguishing a coin in a photographed image and calculating the amount of the distinguished coin automatically, and a system thereof. The system for calculating the amount of coins automatically according to the present invention includes a camera photographing the image, a central processing unit, a memory and a storage where an automatic coin amount calculation program is stored. The central processing unit loads an image selected by a user in the images photographed by the camera in the memory by executing the automatic coin amount calculation program, and converts the image loaded in the memory into a gray scale image, and removes noise by filtering the converted gray scale image, and detects an edge of an object included in the image by filtering the image where noise is removed, and detects a circular shape included in the image by performing a circular shape Hough transform of the image whose edge is detected, and detects a coin in the detected circular shape by post-processing the image which is performed by the circular shape Hough transform, and distinguishes the kind of the detected coin, and calculates the sum of the distinguished coins.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a system and method for automatically calculating a coin amount,

The present invention relates to a method for automatically calculating a coin amount and a system thereof, and more particularly, to a method and system for identifying a coin in a photographed image and automatically calculating the amount of the identified coin.

When calculating the amount of a coin, a coin-operated counter can be used when the number of coins to be calculated is large. However, when the number of coins to be calculated is small, it is not practical to use such a coin counter. This is the case when a small amount of coin is used as a payment or change money at a convenience store or a mart, which is a transaction of everyday life. In addition, purchasing and using a separate coin counter to calculate the amount of coins may be a costly burden to the operator.

If the number of coins is small, the operator can directly calculate and calculate the amount of the coins. However, even in this case, if the operator is not skilled in arithmetic, the amount may be calculated incorrectly, and the calculation of the amount may take a considerable amount of time.

Korean Registered Patent No. 10- 0463023 (December 13, 2004), pp. 2 to 3

The present invention provides a method and system for identifying coins in a photographed image and automatically calculating the amount of the identified coins. The present invention also provides a method and system for automatic calculation of a coin amount that can increase coin discrimination power by performing circular Huff transformation and post-processing to identify coins.

A coin amount automatic calculation system according to the present invention includes a camera for capturing an image, a central processing unit, a memory, and a storage for storing a memory and a coin amount automatic calculation program, wherein the central processing unit executes the coin amount automatic calculation program A memory for storing an image selected by a user; a memory for storing the image; a memory for storing the image; a memory for storing the noise; The method includes detecting an edge of an object included in an image by filtering the removed image, detecting a circular shape included in the image by performing circular Hough transform on the edge-detected image, Detecting the detected coin among the detected circles, and detecting the coin of the detected coin Identification, and calculates a summation value by summing the value of the identified coins.

In one embodiment, the central processing unit removes noise from the converted gray-scale image using a Gaussian filter.

In one embodiment, the central processing unit uses a canny filter to detect an edge of an object in the noise-removed image.

In one embodiment, the central processing unit detects a circle included in the image by calculating a center point of a circle including pixel coordinates of the extracted edge through the following equation (1).

[Equation 1]

Where (x, y) is the coordinate value of the pixel, (i, j) is the center point coordinate value of the circle, r is the radius,

In one embodiment, the central processing unit calculates a center point of a circle including the pixel coordinates by changing the angle (?) And the radius (r), wherein the radius (r) is within a radius range of detection target coins The center point is calculated.

In one embodiment, the central processing unit calculates a probability that an actual circle is formed from the center point to the radius based on the radius value and the calculated at least one center point.

In one embodiment, the central processing unit detects a circular shape having a corresponding radius at the corresponding center point as a circular shape included in the image, if the probability is equal to or larger than a set value through the following equation (2).

&Quot; (2) "

Where A (i, j, r) is the probability of achieving a circle with a radius of r at the center point i, j, k is the probability

In one embodiment, the central processing unit compares the reference circle and the detected circle to remove a circle whose difference in position and radius between the reference circle and the center point is equal to or greater than the set value, The center point position and the radius value are arithmetically averaged, and a circle having the arithmetic mean center point and radius is detected as the coin.

In one embodiment, the central processing unit selects, as the reference circle, the circle having the highest probability of forming the actual circle among the detected circle.

In one embodiment, the central processing unit identifies the type of coin by comparing the radius of the detected coin with the radius of each coin type previously stored.

A method for automatically calculating a coin amount according to the present invention includes the steps of executing a program for automatically calculating a coin amount by a central processing unit, loading an image selected by a user among images photographed by the camera into a memory, Converting the image loaded into the image into a gray-scale image, filtering the noise-removed image by filtering the converted gray-scale image by the central processing unit, filtering the noise-removed image, Detecting the edge of the circular Huff transform, the central processing unit performing a circular Huff transform on the detected image to detect the circular contained in the image, the central processing unit postprocessing the image subjected to the circular Huff transform Detecting the detected coin among the circles; and detecting the coin among the detected coin Identification, and a step of calculating a summation value by summing the value of the identified coins.

In one embodiment, the step of performing the circular Hough transform to detect a circular shape contained in the image may include calculating a center point of a circle including pixel coordinates of the extracted edge through Equation (3) Detect the prototype.

&Quot; (3) "

Where (x, y) is the coordinate value of the pixel, (i, j) is the center point coordinate value of the circle, r is the radius,

In one embodiment, the post-processing of the image subjected to the circular Huff transform may include comparing the reference circle with the detected circle, and removing a circle having a difference between the position difference between the reference circle and the center point and a difference between the radius and the reference circle, And arithmetically averaging the center point position and the radius value with respect to at least one of the remaining circles, and detecting the circle having the arithmetic mean center point and the radius as the coin.

In one embodiment, the step of identifying the type of the detected coin identifies the type of coin by comparing the radius of the detected coin with the radius of the stored coin type.

As described above, the method for automatically calculating the amount of coins according to the present invention and the system thereof can improve the coin discrimination power by performing the circular Hough transform and the post-processing to identify coins.

The method and the system for automatically calculating the amount of coins according to the present invention can reduce errors in calculating the amount of coins accurately by accurately identifying the coins.

Further, the method and system for automatically calculating the amount of coins according to the present invention can be implemented through a system having a camera and a computing device, or a portable terminal such as a smart phone, a tablet PC, and a laptop PC, It is easy to implement.

1 is a diagram showing a system for automatically calculating a coin amount according to an embodiment of the present invention;
2 is a diagram showing the configuration of the computing device of Fig. 1
3 (a) to 3 (d) are diagrams showing a process in which a coin amount automatic calculation process is performed
4 is a flowchart illustrating a method for automatically calculating a coin amount according to an embodiment of the present invention.
5 is a view showing an identified coin
6 is a view showing a result of coin identification according to the arrangement type of coins
7 is a view for comparing coin identification results of a conventional system with coin identification results of the present invention

Hereinafter, a method for automatically calculating the amount of coins according to the present invention and a system thereof will be described in detail.

1 is a view showing a system for automatically calculating a coin amount according to an embodiment of the present invention.

Referring to FIG. 1, a coin amount automatic calculation system includes a computing device 110 and a camera 120. The system for automatically calculating the amount of money can be implemented not only with a system having a camera and a computing device but also through a portable terminal such as a smart phone, a tablet PC, and a laptop PC. Hereinafter, the present invention will be described by way of example of a system having a camera 120 and a computing device 110 for convenience of explanation.

The computing device 110 stores a coin amount automatic calculation program and executes a coin amount automatic calculation program at the request of the user. The computing device 110 can automatically calculate the amount of the coins included in the image through the executed coin amount automatic calculation program. In one embodiment, the computing device 110 may analyze the selected image from the images stored in the storage and calculate the amount of the coins included in the image. In another embodiment, the computing device 110 can turn on the camera 120 at the request of the user to take an image, analyze the taken image, and calculate the amount of the coins included in the image.

The camera 120 transmits the photographed image of the coin 130 to the computing device 110.

2 is a diagram showing the configuration of the computing device of FIG.

2, the computing device includes an input / output interface 210, a central processing unit 220, a memory 230, and a storage 240.

The input / output interface 210 interfaces inputs and outputs of commands or data. For example, the input / output interface 210 may include a command input interface for receiving a command from a user, an external device connection interface for transmitting / receiving data to / from an external device such as the camera 120, and the like.

The central processing unit 220 controls each component of the computing device, performs input / output control of data, and performs data operations. For example, the central processing unit 220 may include a central processing unit (CPU), a graphic processing unit (GPU), and an application processor (AP) of a mobile terminal.

The memory 230 temporarily stores an instruction or data. For example, the memory 230 may temporarily store commands or data input to the central processing unit 220, commands or data output from the central processing unit 220, and the like. The memory 230 may include a RAM (Read Access Memory) or the like.

The storage 240 stores various data. For example, the storage 240 may store various kinds of data such as a program, data necessary for executing a program, data calculated in a program, image data, audio data, and the like. The coin amount automatic calculation program in the coin amount automatic calculation system according to the present invention can be stored in the storage 240. The storage 240 may include non-volatile memory, a hard disk, and the like.

3 (a) to 3 (d) are diagrams illustrating a process of automatically calculating a coin amount.

Hereinafter, the operation of the system for automatically calculating the amount of coins according to the present invention will be described in detail with reference to FIG. 1 to FIG.

The user requests execution of a coin amount automatic calculation program through a command input means (pointer, keyboard, touch input means, etc.) connected to the input / output interface 210. [ When the execution of the automatic amount calculating program is requested, the central processing unit 220 loads the automatic coin amount calculating program stored in the storage 240 into the memory 230 and drives the corresponding program.

When the coin amount automatic calculation program is executed, the central processing unit 220 loads the image selected by the user into the memory 230 among the images stored in the storage 240. In another embodiment, the central processing unit 220 may turn on the camera 120 at the request of the user to take an image including the coin, and load the captured image into the memory 230. [

The central processing unit 220 may analyze the image loaded in the memory 230 to identify the coins included in the image and calculate the amount of the identified coins.

The central processing unit 220 converts the image loaded in the memory 230 into a grayscale image, and filters the converted gray-scale image to remove noise. For example, the central processing unit 220 may remove noise from a gray scale image using a Gaussian filter. 3 (a) shows the original image (left) and the gray scale image (right) of the corresponding image. FIG. 3 (b) shows the gray scale image (left) and the Gaussian filter (Right).

After removing the noise, the central processing unit 220 detects the edge of the object included in the image by filtering the noise-removed image. For example, the central processing unit 220 can detect edges included in an image using a Canny filter. 3 (c) shows the image (left) obtained by removing the noise and the image (right) obtained by detecting the edge of the corresponding image using the cany filter.

After detecting the edge included in the image, the central processing unit 220 performs circular Hough Transform (CHT) on the detected edge image to detect a circular shape included in the image.

(또는,

)는 이들 파라미터에 대한 투표결과를 저장한다. 따라서 형태 검출은 임계값 이상을 갖는 축적셀에 해당하는 파라미터들을 찾는 것이 된다. The circular Hough transform is a method of detecting a specific form in an image using a voting procedure. Voting is done on the parameter space that models the shape, and the vote result is stored in the accumulator space divided by the parameter space. For example, assuming that the parameters for the coordinates (i, j) of the two-dimensional parameter space correspond to p _i and p _j ,

(or,

) Stores the voting results for these parameters. Therefore, the shape detection is to find the parameters corresponding to the accumulation cell having the threshold value or more.

The circular detection process using the circular Hough transform is performed for the range of the effective radius of the coin (r 1 r <r 2 (pixel)) and the angle 0 ≤ θ ≤ 2π in the edge extracted pixel coordinates (x, y) (I, j) of the circle satisfying 1 and the radius r. For example, for a Korean coin, the effective radius range of the coin may correspond to 24 ≤ r <45 (pixel).

는 해당 반지름 값과 산출된 해당 중심점을 기초로, 해당 중심점에서 해당 반지름으로 실제 원형을 이룰 확률을 나타낸다.When i, j is calculated by substituting the coordinate (x, y) value, the angle (?) Value and the radius (r) value in Equation 1, Is increased by one.

Represents a probability that an actual circle is formed from the corresponding center point to the corresponding radius based on the corresponding radius value and the calculated center point.

와 설정값을 비교하여

인 경우 원형이 검출된 것으로 판정한다. 즉,

이 설정값 이상인 경우, 중앙처리장치(220)는 해당 중심점에서 해당 반지름을 갖는 원형이 상기 영상 내에서 탐지된 것으로 판정한다.At this time, the central processing unit 220 calculates i and j values while changing θ and r with respect to one coordinate (x, y). For example, the central processing unit 220 calculates a center point of a circle including the pixel coordinates by changing the angle? And the radius r, and the angle? Is increased by a constant amount to 0??? , And the radius (r) is changed within the radius length range of the detection target coins to calculate the center point. When the calculation is completed for all pixel coordinates from which the edge has been extracted, the central processing unit 220

And the set value

, It is determined that the original has been detected. In other words,

If it is equal to or larger than the set value, the central processing unit 220 determines that a circle having the radius at the center point is detected in the image.

The central processing unit 220 post-processes the image subjected to the circular Huff transform and detects the detected center coins.

라 할 때,

는 원의 중심점 좌표이고, r은 반지름이다. 하나의 동전에 대해 음영과 동전 문양 때문에 복수 개의 원형 에지들이 생기고, 이들은 원형 허프 변환 적용과정을 통해 모두 다른 원형으로 검출된다. 특히, 하나의 원형 에지 일지라도 원 방정식과 일치하지 않아, 복수 개의 원형으로 검출되는 경우도 있다. 이러한 점들을 감안하면, 하나의 동전에 대해 원의 중심이나 반지름이 다른 n개의 원들이 생기므로, 이들을 처리하여 하나의 원 방정식에 해당하는 파라미터 값

을 구하는 후처리 과정이 필요하다.A parameter representing a circle corresponding to a coin

In other words,

Is the center point coordinate of the circle, and r is the radius. For a single coin, a number of circular edges are created due to shading and coin patterns, which are all detected in different circles through a circular Hough transform application process. In particular, even a single circular edge does not coincide with the original equation, and may be detected as a plurality of circles. Considering these points, since n circles with different radii or centers of circles are generated for one coin, they are processed to calculate the parameter values corresponding to one circle equation

A post-processing process is required.

When the central processing unit 220 performs a circular Hough transform and a set of parameters representing n circles detected for one coin is C, C can be expressed by the following equation (2).

At this time, the following equation (3) can be assumed for the center of n circles.

Where ρ is a constant. In other words, this assumption means that the circles created by one coin are formed within a certain range. Now, the process of performing the post-processing by the central processing unit 220 under this assumption will be described.

The central processing unit 220 performs a reference circle and a circular Hough transform to compare the n circles detected for one coin to remove a circle having a difference in positional difference between the reference circle and the center point and a difference between the radii of the circle and the reference circle.

즉, 해당 반지름 값과 산출된 해당 중심점을 기초로 해당 중심점에서 해당 반지름으로 실제 원형을 이룰 확률에 비례하여 커진다. 따라서, 중앙처리장치(220)는 가장 큰 축적셀 값(

)을 갖는 파라미터

를 기준 원형으로 삼는다. 이 때,

는 하기 수학식 4로 나타낼 수 있다.That is, the central processing unit 220 removes the non-circles corresponding to the coins in the first step of the post-processing process. To this end, the central processing unit 220 eliminates circles whose center point position or radius difference is greater than or equal to the threshold, compared to the circle with the highest probability of being coincident. The probability that a circle corresponds to a coin

In other words, it increases in proportion to the probability of obtaining the actual circle from the corresponding center point based on the calculated radius value and the calculated center point. Accordingly, the central processing unit 220 can determine the largest accumulation cell value (

)

As a reference circle. At this time,

Can be expressed by the following equation (4).

과 원형 허프 변환을 수행하여 하나의 동전에 대해 탐지된 n개의 원을 비교하여 중심점 위치나 반지름 크기가 각각 임계값 α와 β이상으로 차이가 나는 원들을 집합 C에서 제거하고 새로운 집합 C₁을 구성한다. 이를 수학식으로 표시하면 하기 수학식 5로 표현될 수 있다.The central processing unit 220 has a reference circular

And circular huff transformation to compare the detected n circles for one coin and remove the circles whose center point position or radius magnitude differs by more than the threshold value α and β respectively from the set C and construct a new set C ₁ do. This can be expressed by the following equation (5).

After the non-circles corresponding to the coins are removed in the first step of the post-processing process, the central processing unit 220 performs arithmetic averaging of the center point position and the radius values with respect to at least one of the remaining circles, , And a circle having an arithmetic mean center point and a radius is detected as a coin.

을 하기 수학식 6과 같이 표현할 수 있다.That is, the central processing unit 220 arithmetically averages the set C ₁ consisting of the circles representing the coins,

Can be expressed by the following equation (6).

을 수학식 6과 같이 산출한다. 도 3의 (d)는 에지가 탐지된 영상(좌)과 후처리를 통해 해당 영상에서 동전에 해당하는 원을 탐지한 영상(우)을 나타낸다.Considering that the pixels of the edges representing the circle are not arranged in a perfect match with the circular equations and that the coordinates of the pixels are integers, a difference from the actual equation values can not be avoided, so that the central processing unit 220 A set C ₁ consisting of circles representing coins is arithmetically averaged,

As shown in Equation (6). FIG. 3 (d) shows a video (left) in which an edge is detected (left) and a video (right) in which a circle corresponding to a coin is detected in a corresponding image through a post-process.

After the central processing unit 220 detects a circle corresponding to a coin through post-processing, the central processing unit 220 identifies the type of the detected coin and adds the amounts of the identified coins to calculate a sum amount . In one embodiment, the central processing unit 220 can identify the type of coin by comparing the radius of the detected coin with the radius of the previously stored coin type.

For example, in the case of 500 won, 100 won, 50 won and 10 won coins of Korea, the length of the radius as shown in Table 1 below may be stored in the storage 240 in advance.

4 is a flowchart illustrating a method for automatically calculating a coin amount according to an embodiment of the present invention.

When the coin amount automatic calculation program is executed at the request of the user, the central processing unit 220 analyzes the image photographed through the camera 120 or the image selected by the user among the images stored in the storage 240, Automatically calculate the amount of coins included.

The central processing unit 220 converts the photographed image or the selected image into a grayscale image (step S410), and filters the converted grayscale image to remove noise (step S420). For example, the central processing unit 220 may remove noise using a Gaussian filter.

After removing the noise, the central processing unit 220 detects the edge by filtering the noise-removed image (step S430). For example, the central processing unit 220 may detect edges using a canny filter.

After detecting the edge, the central processing unit 220 performs circular Hough transform (CHT) on the basis of the detected edge (step S440), postprocesses the circular Hough transformed image, and detects coins contained in the image (Step S450). In one embodiment, the step of performing post-processing comprises the steps of: comparing the reference circle with the detected circle so as to remove a circle whose difference in position and radius between the reference circle and the center point is equal to or greater than the set value; Arithmetically averaging the center point position and the radius value with respect to the circle, and detecting a circle having the arithmetic mean center point and the radius as the coin.

The process of performing the circular Hough transform and the process of performing the post-process are as described in Equation (1) and Equation (2) to Equation (6), respectively.

When the coin contained in the image is detected, the central processing unit 220 identifies the type of the detected coin (step S460), and automatically calculates the sum of the coin amounts by summing the amounts of the respective coins identified S470). In one embodiment, identifying the type of coin identifies the type of coin by comparing the radius of the detected coin with the radius of the stored coin type.

FIG. 5 is a view showing coins identified using the automatic coin amount calculation method according to an embodiment of the present invention. The left image shows the original image, and the right image shows the mark according to the type of the coin identified on the original image. For example, a 500 won coin is shown as a circle, 100 won as a square, 50 won as a straight line, and 10 won as a cross. 5, it can be seen that the method of automatically calculating the amount of coins according to the present invention accurately identifies the type of coins.

6 is a diagram showing a result of coin identification according to the arrangement type of coins. FIG. 6 is a result of an experiment for discriminating the coin amount automatic calculating method according to the present invention while varying the arrangement such as the number of coins, the position, the interval between coins, and confirming that the recognition rate is reduced only when there is no gap between coins . This is because proper circle detection is not achieved due to overlap between adjacent circles

7 is a view for comparing the coin identification result of the conventional system with the coin identification result of the present invention. The left image of FIG. 7 represents the coin identification result according to the existing system, and the right image represents the coin identification result according to the present invention. Referring to FIG. 7, it can be seen that the method of automatically calculating the amount of coins according to the present invention improves the coin discrimination performance compared to the conventional method by using circular Hough transform. In the conventional method, an object other than a coin is recognized as a coin, so that the discrimination power can be confirmed.

While the present invention has been described in connection with certain exemplary embodiments, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, but, on the contrary, it is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the invention.

110: computing device
120: camera

Claims (14)

A camera for capturing an image;
A central processing unit;
Memory; And
Includes storage that stores a coin amount automatic calculation program,
The central processing unit executes the coin amount automatic calculation program
The image selected by the user among the images photographed by the camera is loaded into the memory,
Converts the image loaded into the memory into a gray scale image,
Filtering the converted gray-scale image to remove noise,
An edge of the object included in the image is detected by filtering the noise-
A circular Huff transform is performed on the edge-detected image to detect a circular shape included in the image,
Transforming the detected circular-shaped coin by post-processing the image subjected to the circular Huff transform,
Identifying a type of the detected coin,
The sum of the identified coins is added to calculate the sum
Automatic calculation system of coin amount. The method according to claim 1,
The central processing unit
Wherein the noise is removed from the converted gray-scale image using a Gaussian filter. The method according to claim 1,
The central processing unit
And the edge of the object is detected in the image from which the noise is removed by using the canny filter. The method according to claim 1,
The central processing unit
Wherein a circular center included in the image is detected by calculating a center point of a circle including pixel coordinates of the extracted edge through Equation (1): &quot; (1) &quot;
[Equation 1]

Where (x, y) is the coordinate value of the pixel, (i, j) is the center point coordinate value of the circle, r is the radius, 5. The method of claim 4,
The central processing unit
Calculating a center point of the circle including the pixel coordinates while changing the angle and radius r while calculating the center point while changing the radius r within the radius range of the detection target coins Automatic calculation system of coin amount. 6. The method of claim 5,
The central processing unit
Wherein the probability calculating unit calculates a probability that an actual circle is formed from the center point to the corresponding radius based on the radius value and the calculated at least one center point. The method according to claim 6,
The central processing unit
If the probability is equal to or greater than a predetermined value, detecting a circular shape having a corresponding radius at the corresponding center point as a circular shape included in the image.
&Quot; (2) &quot;

Where A (i, j, r) is the probability of achieving a circle with a radius of r at the center point i, j, k is the probability The method according to claim 1,
The central processing unit
And comparing the reference circle and the detected circle to remove a circle whose difference in position and radius between the reference circle and the center point is equal to or greater than the set value,
Arithmetically averages the center point position and the radial value with respect to at least one of the remaining circles removed, and detects a circle having the arithmetic mean center point and the radius as the coin. 9. The method of claim 8,
The central processing unit
Wherein a prototype having the highest probability of forming an actual prototype among the detected protoplasts is selected as a reference prototype. The method according to claim 1,
The central processing unit
Wherein the type of coin is identified by comparing the radius of each coin type and the detected radius of the coin. Executing a coin amount automatic calculation program by a central processing unit;
Loading an image selected by a user among images photographed by a camera into a memory;
Converting the image loaded in the memory into a grayscale image by the central processing unit;
The central processing unit filtering the converted gray-scale image to remove noise,
The central processing unit filtering the noise-removed image to detect an edge of an object included in the image;
The central processing unit performing circular Hough transform on the detected image of the edge to detect a circular shape included in the image;
Processing the image on which the circular Huff transform has been performed to detect the detected middle round coin; And
And the central processing unit identifies the type of the detected coin and adds the amounts of the identified coins to calculate a sum amount. 12. The method of claim 11,
The step of performing the circular Hough transform to detect a circular shape included in the image
And calculating a center point of a circle including the pixel coordinates of the extracted edge through the following expression (3), thereby detecting the original included in the image.
&Quot; (3) &quot;

Where (x, y) is the coordinate value of the pixel, (i, j) is the center point coordinate value of the circle, r is the radius, 12. The method of claim 11,
The step of post-processing the image subjected to the circular Huff transform
Comparing the reference circle with the detected circle, and removing a circle having a difference between a position difference between the reference circle and the center point and a difference between the radii of the reference circle and the center point is equal to or larger than a set value; And
Calculating arithmetic mean of the center point position and the radius value for at least one remaining circle that has been removed, and detecting the circle having the arithmetic mean center point and radius as the coin. 12. The method of claim 11,
The step of identifying a type of the detected coin
Wherein the type of the coin is identified by comparing the radius of the coin with the radius of the detected coin.

Images