Device and method providing feature information of target fish

Abstract

According to one embodiment, a method and device for providing fish object characteristic information for a target fish object are disclosed, comprising: a receiving unit acquiring a target fish object image for a target fish object; a processor acquiring depth information for the target fish object based on the target fish object image; the processor determining a plurality of key points determined according to the shape of the target fish object and used to represent the characteristics of the target fish object using the target fish object image and the depth information; the processor acquiring fish object characteristic information including information on at least one of a plurality of edge points, weight, total length, body length, body height, head length, caudal peduncle height, caudal peduncle length, and body width for the target fish object based on the plurality of key points; and an output unit providing the fish object characteristic information.

Description

Technology Field

[0001] The technical field of the present disclosure relates to characteristic information regarding fish objects, and more specifically, to a technology that can easily provide characteristic information related to fish by obtaining various trait information corresponding to a target fish from an image of a fish object obtained from a photographic device. Background Technology

[0003] Traditionally, fish identification tasks required a large workforce and significant time investment, and there were many inconveniences in obtaining precise fish trait information in real time. Furthermore, 3D modeling of fish required expensive equipment, and due to a lack of systematized processes, it was difficult to complete sufficient 3D modeling simply by using fish object images. Accordingly, there is a demand for customized service technology that can acquire fish object images and provide managers or technicians with trait information corresponding to various fish, as well as detailed information for 3D modeling. Prior art literature

[0005] Korean Published Patent 10-2023-0126986 A (2023.08.31) Sorting system for bycatch fish by size and type The problem to be solved

[0006] The present disclosure discloses a device and a method for acquiring an image of a target fish object, determining a plurality of key points thereon, and providing fish object characteristic information for the fish object based on the acquired plurality of key points.

[0007] The problems to be solved in this disclosure are not limited to those mentioned above, and other unmentioned problems will be clearly understood by a person skilled in the art from the description below. means of solving the problem

[0009] As a technical means for achieving the technical problem described above, a method according to the first aspect of the present disclosure may include: a step of acquiring a target fish object image for a target fish object; a step in which a processor acquires depth information for the target fish object based on the target fish object image; a step in which the processor determines a plurality of key points determined according to the shape of the target fish object and used to represent the characteristics of the target fish object using the target fish object image and the depth information; a step in which the processor acquires fish object characteristic information including information on at least one of a plurality of edge points, weight, total length, body length, body height, head length, caudal peduncle height, caudal peduncle length, and body width for the target fish object based on the plurality of key points; and a step in which an output unit provides the fish object characteristic information.

[0010] Additionally, the method may further include the step of obtaining a mask image that distinguishes and displays the internal region of the target fish object and the external region of the target fish object using the target fish object image.

[0011] Additionally, the step of acquiring the fish object characteristic information may include: a step of acquiring a first depth value for the outer region using the mask image; a step of acquiring a second depth value for the most derived point in the inner region using the mask image; and a step of determining the body width based on the difference between the first depth value and the second depth value.

[0012] Additionally, the method may further include the step of obtaining a plurality of segmented triangles based on the connection between the plurality of edge points; the step of obtaining a plurality of internal segmented triangles in which a plurality of external segmented triangles corresponding to an external region among the plurality of segmented triangles are excluded using the mask image; and the step of obtaining 3D modeling information for the target fish object using the plurality of internal segmented triangles.

[0013] Additionally, the method may further include the steps of: determining a plurality of edge projection points corresponding to each of the plurality of edge points and representing points on which the plurality of edge points are projected; determining a plurality of edge triangles obtained based on the plurality of edge projection points and the plurality of edge points; performing lateral modeling of the target fish object based on the plurality of edge triangles; and providing lateral modeling information obtained as the lateral modeling is performed by the output unit.

[0014] Additionally, the method may further include the step of the processor acquiring a similar fish object image corresponding to the type of the target fish object; and the step of the processor acquiring a segment image in which a plurality of internal regions of the target fish object are divided into segment lines according to the comparison result between the target fish object image and the similar fish object image, wherein the plurality of internal regions may include parts of the body, upper fin, lower fin, and caudal fin of the target fish object, and the segment line may include at least one of the plurality of key points.

[0015] Additionally, if the ratio of the body width to the total length of the target fish object is greater than or equal to a threshold ratio, the processor may further include the step of determining a characteristic information error probability representing an error probability regarding the fish object characteristic information; and the output unit may further include the step of providing an information abnormality notification regarding the characteristic information error probability, wherein the information abnormality notification may be determined differently depending on the difference value between the ratio and the threshold ratio.

[0016] Additionally, the step of acquiring the fish object characteristic information may further include the step of determining an obesity rate representing the health status of the target fish corresponding to the target fish object according to the ratio of the length value for the body height and body length of the target fish object and the weight value for the weight.

[0017] A device for providing fish object characteristic information for a target fish object according to a second aspect of the present disclosure may include: a receiving unit for acquiring a target fish object image for a target fish object; a processor for acquiring depth information for a target fish object based on the target fish object image and using the target fish object image and the depth information to determine a plurality of key points determined according to the shape of the target fish object and used to represent the characteristics of the target fish object, and acquiring fish object characteristic information including information on at least one of a plurality of edge points, weight, total length, body length, body height, head length, caudal peduncle height, caudal peduncle length, and body width for the target fish object based on the plurality of key points; and an output unit for providing the fish object characteristic information.

[0018] Additionally, a third aspect of the present disclosure may provide a computer-readable recording medium having a program for executing the method of the first aspect on a computer. Effects of the invention

[0020] According to one embodiment of the present disclosure, by acquiring characteristic information corresponding to a fish based on a plurality of key points determined from an acquired fish object image, the inconvenience of requiring a large number of personnel, expensive equipment, and a long time can be eliminated.

[0021] The effects of the present disclosure are not limited to those mentioned above, and other unmentioned effects will be clearly understood by a person skilled in the art from the description below. Brief explanation of the drawing

[0023] FIG. 1 is a block diagram schematically illustrating the configuration of a device that provides fish object characteristic information for a target fish object according to one embodiment. FIG. 2 is a flowchart illustrating a method for providing fish object characteristic information for a target fish object according to one embodiment. FIG. 3 is a diagram illustrating the interaction between a shooting device, a database, and a manager device according to one embodiment. FIG. 4 is a schematic diagram illustrating an example of obtaining a mask image from a target fish object image according to one embodiment. FIG. 5 is a drawing for explaining an example of determining a segment line and a plurality of key points appearing in a segment image according to one embodiment. FIG. 6 is a schematic diagram illustrating an example of obtaining multiple edge triangles based on multiple edge points and multiple edge projection points according to one embodiment. FIG. 7 is a diagram illustrating an example of obtaining fish object characteristic information based on a plurality of key points according to one embodiment. FIG. 8 is a diagram illustrating an example of obtaining 3D modeling information for a target fish object using the Delaunay Triangulation technique and the Ear-Clipping algorithm according to one embodiment. Specific details for implementing the invention

[0024] The advantages and features of the present disclosure and the methods for achieving them will become clear by referring to the embodiments described below in detail together with the accompanying drawings. However, the present disclosure is not limited to the embodiments disclosed below but can be implemented in various different forms, and the embodiments provided are merely to make the disclosure complete and to fully inform those skilled in the art of the scope of the present disclosure.

[0025] The terms used in this specification are for describing embodiments and are not intended to limit the disclosure. In this specification, the singular form includes the plural form unless specifically stated otherwise in the text. The terms “comprises” and / or “comprising” as used in this specification do not exclude the presence or addition of one or more other components in addition to the components mentioned. Throughout the specification, the same reference numerals refer to the same components, and “and / or” includes each of the mentioned components and all combinations of one or more. Although terms such as “first,” “second,” etc., are used to describe various components, these components are not limited by these terms. These terms are used merely to distinguish one component from another. Accordingly, the first component mentioned below may be the second component within the technical scope of this disclosure.

[0026] Unless otherwise defined, all terms used herein (including technical and scientific terms) may be used in a sense that is commonly understood by a person skilled in the art. Additionally, terms defined in commonly used dictionaries are not to be interpreted ideally or excessively unless explicitly and specifically defined otherwise. Spatially relative terms such as "below," "beneath," "lower," "above," and "upper" may be used to easily describe the relationship between one component and other components, as illustrated in the drawings.

[0027] Spatially relative terms should be understood as terms that include different orientations of components during use or operation, in addition to the orientations depicted in the drawings. For example, if a component depicted in a drawing is inverted, a component described as "below" or "beneath" another component may be placed "above" of that other component. Therefore, the exemplary term "below" may include both the downward and upward directions. Components may also be oriented in other directions, and accordingly, spatially relative terms may be interpreted according to the orientation.

[0028] Various embodiments are described in detail below with reference to the drawings.

[0030] FIG. 1 is a block diagram schematically illustrating the configuration of a device that provides fish object characteristic information for a target fish object according to one embodiment.

[0031] Referring to the drawing, the device (100) may include a receiving unit (110), a processor (120), and a transmitting unit (130).

[0032] A person skilled in the art will understand that, in addition to the components shown in FIG. 1, other general-purpose components may be further included in the device (100). For example, the device (100) may further include a memory (not shown). Or, according to other embodiments, a person skilled in the art will understand that some of the components shown in FIG. 1 may be omitted.

[0034] A receiving unit (110) according to one embodiment can acquire a target fish object image for a target fish object. The target fish object image may include the target fish and the floor surface where the target fish is located. The floor surface is not limited to a specific surface and can be interpreted as an external and internal area where the target fish is located for the target fish object image to be captured, such as a workbench and a conveyor belt. Additionally, the receiving unit (100) can acquire multiple target fish object images for the same target fish object depending on the position of the shooting device. Furthermore, the receiving unit (100) is not limited to a specific type of image and can acquire the target fish object in various ways, such as video or clips, depending on the type of shooting device.

[0035] A processor (120) according to one embodiment can obtain fish object characteristic information based on a plurality of key points obtained using a target fish object image and depth information. The number of key points is not limited to a specific number and can be determined as a location where the internal area of ​​the target fish is distinguished or where differences in height and color are distinguished, and the number of key points may vary depending on the type and shape of the target fish. Based on the determined plurality of key points, the processor (120) can obtain fish object characteristic information including information on at least one of a plurality of edge points, weight, total length, body length, body height, head length, caudal peduncle height, caudal peduncle length, and body width for the target fish object. The fish object characteristic information is not limited to specific information and may additionally include fish-related trait information collected by observing the target fish, including the type, size, weight, and condition of the fish.

[0036] An output unit (130) according to one embodiment may provide fish object characteristic information. The concept of outputting fish object characteristic information may be interpreted as transmitting it to an administrator and technician device. Not limited thereto, the output unit (130) may provide fish object characteristic information in various ways that allow administrator and worker accounts to visually verify it.

[0038] FIG. 2 is a flowchart illustrating a method for providing fish object characteristic information for a target fish object according to one embodiment.

[0039] Referring to step S210, the device (100) acquires a target fish object image for the target fish object.

[0040] For example, the device (100) may acquire an image of a target fish object from a database or acquire the target fish by capturing it in real time from a shooting device. Additionally, when the image of the target fish object is acquired, the device (100) may perform a preliminary inspection of the image of the target fish object. The preliminary inspection may be interpreted as a review process to check for distortion in the image itself or for the shape of the target fish object not being recognized.

[0042] Referring to step S220, the device (100) obtains depth information for a target fish object based on an image of a target fish object.

[0043] For example, the depth information may include information about the distance between the shooting device and the bottom surface where the target fish object is located, and information about the distance between the shooting device and a certain point of the target fish object. Not limited thereto, the device (100) may acquire distance information for a specific object recognized by the shooting device and acquire it as depth information in a database. For example, the depth information may be acquired using various methods such as a depth measuring camera, infrared, RGB shooting, and stereo vision, and is not limited to a specific method.

[0045] A device (100) according to one embodiment obtains a mask image that distinguishes between an internal region of a target fish object and an external region of the target fish object using a target fish object image.

[0046] For example, the outer region of the target fish object can be interpreted as a region that does not include the region corresponding to the target fish object, and in the example described above, it may mean the bottom surface where the target fish object is located. The inner region of the target fish object may correspond to the head, body, and fins, which are parts of the shape of the target fish object. Accordingly, the device (100) can obtain a mask image that distinguishes between the outer region and the inner region of the target fish object, and an example thereof is described in detail in FIG. 4.

[0048] A device (100) according to one embodiment obtains a first depth value for an outer region using a mask image, obtains a second depth value for the most derived point in an inner region using a mask image, and determines a body width based on the difference between the first depth value and the second depth value.

[0049] For example, the first depth value can be interpreted as the distance from the capturing device to the bottom surface where the target fish object is projected. Additionally, the second depth value can be interpreted as the part of the body of the target fish object closest to the capturing device that is most protruded from the bottom surface. The first and second depth values ​​are not limited to specific values ​​or units and can be obtained as depth values ​​corresponding to physical units or bit values ​​corresponding to pixels.

[0050] For example, the first depth value, which is the length from the shooting device to the bottom surface, means having a higher value than the second depth value, which is the highest value among the bodies of the target fish object. Accordingly, the difference value obtained by subtracting the second depth value from the first depth value, that is, the distance between the bottom surface where the target fish object is located and the highest point among the target fish object, can be represented, and the device (100) can determine the difference value as the body width of the target fish object. In the case of a bit value corresponding to a pixel, the device (100) can determine the body width of the target fish object by converting it into a depth value corresponding to a physical unit according to the number between bits.

[0051] For example, if the second depth value is greater than the first depth value, the device (100) may provide a depth error notification to the user account indicating that the depth information for the target fish object is an error, and may re-acquire a mask image for the target fish object according to the user account's selection input.

[0053] Referring to step S230, the device (100) uses the target fish object image and depth information to determine a plurality of key points that are determined according to the shape of the target fish object and used to represent the characteristics of the target fish object.

[0054] For example, a key point can be interpreted as a key point included within the shape of a target fish object appearing within an image of a target fish object. Additionally, the device (100) can obtain fish object characteristic information for the target fish object along a connecting line extending from one of the multiple key points to another.

[0055] A device (100) according to one embodiment acquires a similar fish object image corresponding to a type of target fish object and acquires a segment image in which a plurality of internal regions of the target fish object are divided into segment lines according to the comparison result between the target fish object image and the similar fish object image, wherein the plurality of internal regions include parts of the body, upper fin, lower fin, and tail fin of the target fish object and the plurality of segment lines include at least one of a plurality of key points.

[0056] For example, if the similarity level indicating similarity to the type of similar fish object similar to the type of target fish object is greater than or equal to a preset similarity level, the device (100) can obtain a similar fish object image corresponding to the similar fish object from a database. That is, the target fish object image and the similar fish object image can be interpreted as having similar positions of head, body, and fins to some extent, and by performing a comparison, the device (100) can obtain a segment image of the target fish object. The segment image can be interpreted as an image in which multiple internal regions of the target fish object are divided by segment lines, and a detailed explanation thereof is provided later in FIG. 5.

[0058] Referring to step S240, the device (100) obtains fish object characteristic information including information on at least one of a plurality of edge points, weight, total length, body length, body height, head length, caudal peduncle height, caudal peduncle length, and body width for a target fish object based on a plurality of key points.

[0059] For example, fish object characteristic information including multiple edge points, weight, body length, body height, head length, caudal peduncle height, caudal peduncle length, and body width for a target fish object obtained based on multiple key points is described later in FIG. 7. In addition, the fish object characteristic information is not limited to the specific information described above, and depending on the type of fish object, the device (100) may additionally obtain fish object characteristic information corresponding thereto.

[0060] A device (100) according to one embodiment determines an obesity rate representing the health status of a target fish corresponding to a target fish object based on the ratio of the length value to the height and length of the target fish object and the weight value to the weight.

[0061] For example, among the acquired fish object characteristic information, the obesity rate of the target fish may be determined based on the body length, body height, and weight of the target fish object. Additionally, based on the determined obesity rate, the device (100) may provide a notification regarding the obesity rate to an administrator or worker account to inform them of the health status of the target fish.

[0062] For example, the device (100) can determine the obesity rate for a target fish by comparing the ratio of the summed length and height values ​​of the target fish object to the weight value of the target fish object with a preset standard obesity rate corresponding to the type of target fish received from the database. The method of determining the obesity rate is not limited to a specific method and can be changed and updated differently depending on the method entered from the administrator and worker accounts.

[0064] Referring to step S250, the device (100) provides fish object characteristic information.

[0065] A device (100) according to one embodiment determines a characteristic information error probability indicating an error probability for fish object characteristic information when the ratio of body width to total length of a target fish object is greater than or equal to a threshold ratio, provides an information abnormality notification regarding the characteristic information error probability, and the information abnormality notification is determined differently depending on the difference value between the ratio and the threshold ratio.

[0066] For example, if the ratio of body width to total length of a target fish object exceeds a threshold ratio, it can be interpreted that there is a high probability that some of the acquired fish object characteristic information will appear as errors. Accordingly, notifications regarding the probability of characteristic information errors are provided to administrator and worker accounts differently depending on the difference between the ratio of body width to total length and the threshold ratio, thereby enabling a rapid response to errors in the fish object characteristic information. The probability of characteristic information errors can be calculated as a ratio obtained by dividing the body width of the target fish object by the total length, and the device (100) can perform a comparison between the calculated ratio and the threshold ratio. The method of calculating the ratio is not limited to a specific method, and the method of calculating the ratio can be changed and updated according to the selection input of the administrator and worker accounts.

[0067] As another example, if the probability of a characteristic information error is greater than or equal to the first probability, the device (100) may determine the abnormal information notification to reacquire the image, and if the probability of a characteristic information error is greater than or equal to the second probability, the abnormal information notification may determine the reacquiring of key points, and the first probability may be a higher probability than the second probability. That is, in the case of the first probability, it can be interpreted to mean that the ratio of the body width to the total length of the target fish object is significantly different from the threshold ratio, and since it is determined that the degree of error is large and there is a high probability that there is a problem with the target fish object image itself, the device (100) may reacquire the target fish object image according to the selection input by providing an abnormal information notification to the manager and the operator accordingly. In addition, in the case of the second probability, it may mean a lower probability than the first probability, and the device (100) may reacquire multiple key points using an existing target fish object image with a lower intensity than the performance of reacquiring the target fish object image. Abnormal information notifications provided based on the first and second probabilities may be changed and updated according to the notification methods pre-configured in the administrator and worker accounts.

[0069] FIG. 3 is a diagram illustrating the interaction between a device (100) and a shooting device (330), a database (320), and an administrator device (310) according to one embodiment.

[0070] Referring to FIG. 3, the device (100) can obtain a target fish object image for a target fish object from the database (320). Additionally, the device (100) can obtain a target fish object image captured in real time from the shooting device (330). The database (320) can store one or more target fish object images and can transmit a similar fish object image for a similar fish object corresponding to the target fish object upon a request from the device (100). Additionally, the database (320) can store fish object characteristic information corresponding to the target fish object. Fish object characteristic information according to the type of target fish can be classified and stored in the database (320) using indexing, and the existing fish object characteristic information can be updated whenever additional fish object characteristic information is obtained. Additionally, as multiple images of target fish objects are acquired, the device (100) can extract features of the target fish objects by learning the images using a Convolutional Neural Network (CNN) technique used in deep learning. The technique used in deep learning is not limited to a specific technique, and the technique can be updated according to the selection input of the administrator and worker accounts.

[0072] FIG. 4 is a schematic diagram illustrating an example in which a device (100) according to one embodiment obtains a mask image (420) from a target fish object image (410).

[0073] Referring to FIG. 4 and the above-described step S220, the device (100) can obtain a mask image (420) in which an inner region and an outer region of a target fish object are distinguished from a target fish object image (410), and thereby determine a first depth value for the outer region and a second depth value, which is the most derived point in the inner region.

[0074] For example, the device (100) can obtain a 3D Point Cloud, which is a set of points corresponding to a 3D space for distinguishing an inner region and an outer region from a target fish object image (410) obtained from a shooting device (330), and can obtain a minimum perimeter bounding box corresponding to the perimeter of the target fish object. Accordingly, the device (100) can convert the 3D Point Cloud into a 2D coordinate system using the following Equation 1 based on an Intrinsic parameter representing the degree of movement or tilt of the target fish object from the shooting device (330).

[0076]

[0078] Additionally, the device (100) can acquire an external 3D point corresponding to an external region. That is, the external region can be interpreted as the bottom surface where the target fish object is located, and the device (100) can remove abnormal values ​​from the external 3D point and perform a flattening transformation on the value using the Ransac Regression algorithm to represent the flattened bottom. Accordingly, the device (100) can acquire a mask image (420) in which the internal region of the target fish object and the external region of the target fish object are distinguished.

[0080] FIG. 5 is a drawing for explaining an example in which a device (100) according to one embodiment determines a segment line (510) and a plurality of key points (520) appearing in a segment image.

[0081] Referring to FIG. 5 and the above-described step S230, the device (100) can determine a plurality of key points (520) that are determined according to the shape of the target fish object and used to indicate the characteristics of the target object.

[0082] For example, from a mask image (420) in which the inner and outer regions of the target fish object described in FIG. 4 are distinguished, the device (100) can obtain a plurality of key points (520) for the target fish object using an Artificial Neural Network (ANN) algorithm. Based on the plurality of key points (520), the device (100) can perform conversion into key point 3D coordinate values ​​corresponding to the plurality of key points (520) and including X, Y, and Z axes. Accordingly, based on the obtained key point 3D coordinate values ​​and the outer region of the target fish object, the device (100) can perform relocation by updating the origin of the key point 3D coordinate values ​​for the mask image (420) through the following Equation 2. That is, the device (100) can determine the direction represented by the X, Y, and Z axes of the coordinates through Equation 2 corresponding to the Transformation Matrix, and can determine the updated origin by performing an operation with the coordinates on the inverse matrix of the Transformation Matrix using Equation 3 below.

[0084]

[0086]

[0088] Accordingly, the device (100) can obtain fish object characteristic information including information on at least one of a plurality of edge points, weight, total length, body length, body height, head length, caudal peduncle height, caudal peduncle length, and body width for a target fish object based on 3D coordinate values ​​corresponding to a plurality of relocated key points (520). The algorithms and techniques for the plurality of key points (520) are not limited to any specific algorithms and techniques and can be changed and updated according to the selection input of the administrator and worker accounts.

[0089] As another example, the device (100) can obtain a similar fish image for a similar fish object similar to the type of target fish object received from the database (320), and can obtain a segment image in which a plurality of internal regions of the target fish object are divided into segment lines (510) according to the result of comparison with the target fish object image (410).

[0090] Referring to FIG. 5, the segment line (510) can be interpreted as a boundary line in which the body and tail fin are divided among a plurality of internal regions of a target fish object, but is not limited thereto, and may further include a segment line corresponding to a different part of a plurality of internal regions of a target fish object. Accordingly, the device (100) can obtain fish object characteristic information by connecting one or more key points (520) using coordinate values ​​of the X, Y, and Z axes based on some of the plurality of key points (520) included in the segment line (510).

[0092] FIG. 6 is a schematic diagram illustrating an example in which a device (100) according to one embodiment obtains a plurality of edge triangles based on a plurality of edge points (610) and a plurality of edge projection points (620).

[0093] A device (100) according to one embodiment determines a plurality of edge projection points (620) corresponding to each of a plurality of edge points (610) and representing points projected by the plurality of edge points (610), determines a plurality of edge triangles obtained based on the plurality of edge projection points (620) and the plurality of edge points (610), performs side-face modeling for a target fish object based on the plurality of edge triangles, and provides side-face modeling information obtained as the side-face modeling is performed.

[0094] Referring to FIG. 6, a plurality of edge points (610) can be interpreted as internal regions of a target fish object obtained from a shooting device, and a plurality of edge projection points (620) can be interpreted as the bottom surface where the target fish object is located. Accordingly, the device (100) can obtain a plurality of edge triangles by connecting some of the plurality of edge points (610) with the plurality of edge projection points (620). That is, the shape represented by the plurality of edge triangles can be interpreted as the side surface of the target fish object. Additionally, some of the plurality of edge triangles may be Upward and some may be Downward. For example, an edge triangle may be created with two edge points (610) and one edge projection point (620), or with one edge point (610) and two edge projection points (620). Accordingly, the device (100) can perform side-face modeling by connecting the area and perimeter length of the faces of a plurality of acquired edge triangles and can provide side-face modeling information for a target fish object.

[0096] FIG. 7 is a diagram illustrating an example in which a device (100) according to one embodiment obtains fish object characteristic information based on a plurality of key points (520).

[0097] Referring to FIG. 7, the fish object characteristic information obtained based on a plurality of key points (520) may include information on at least one of a plurality of edge points (610), weight, total body length, body height, head length, caudal peduncle height, caudal peduncle length, and body width for the target fish object. Not limited thereto, the device (100) may obtain a plurality of different key points (520) depending on the type or state of the target fish object, and may obtain additional fish object characteristic information corresponding thereto as some of the plurality of key points are connected.

[0099] FIG. 8 is a drawing illustrating an example in which a device (100) according to one embodiment obtains 3D modeling information for a target fish object using a Delaunay triangulation (810) technique and an Ear-Clipping algorithm (820).

[0100] A device (100) according to one embodiment acquires a plurality of divided triangles based on connections between a plurality of edge points, acquires a plurality of internal divided triangles (822) in which a plurality of external divided triangles (821) corresponding to an external region among the plurality of divided triangles are excluded using a mask image, and acquires 3D modeling information for a target fish object using the plurality of internal divided triangles.

[0101] Referring to FIG. 8, the device (100) can obtain 3D modeling information for the bottom surface by using a plurality of edge projection points described in FIG. 6. Additionally, by using the Delaunay triangulation technique (810), the device (100) can obtain a plurality of internal division triangles (821) and a plurality of external division triangles (822) corresponding to an internal area and a part of an external area of ​​the target fish object. Thus, by obtaining a plurality of internal division triangles (822) excluding the plurality of external division triangles (821) corresponding to a part of the external area, the device (100) can obtain 3D modeling information for the target fish object using the plurality of internal division triangles (822).

[0102] In another example, the device (100) can obtain 3D modeling information for a target fish object using an Ear-Clipping algorithm (820). Based on two vectors generated from a plurality of segmented triangles, the device (100) can determine a plurality of external segmented triangles (821) corresponding to the direction indicated by the vectors as the same as the direction of the external region of the target fish object to be excluded.

[0103] Accordingly, the device (100) can obtain 3D modeling information by performing a Delaunay triangulation technique (810) or an Ear-Clipping algorithm (820) to connect multiple internal dividing triangles (822), excluding multiple external dividing triangles (821) corresponding to the external region among multiple dividing triangles. Additionally, by summing the side modeling information described in FIG. 6, the device (100) can provide the final 3D modeling information, which is the merging of the 3D modeling information and the side modeling information, to the administrator and worker accounts.

[0105] Various embodiments of the present disclosure may be implemented as software comprising one or more instructions stored in a storage medium (e.g., memory) readable by a machine (e.g., a display device or a computer). For example, a processor of the machine (e.g., processor (120)) may call at least one of the one or more instructions stored from the storage medium and execute it. This enables the machine to be operated to perform at least one function according to the at least one called instruction. The one or more instructions may include code generated by a compiler or code that can be executed by an interpreter. The storage medium readable by the machine may be provided in the form of a non-transitory storage medium. Here, 'non-temporary' merely means that the storage medium is a tangible device and does not contain a signal (e.g., electromagnetic waves), and this term does not distinguish between cases where data is stored semi-permanently and cases where it is stored temporarily.

[0106] According to one embodiment, the method according to the various embodiments disclosed herein may be provided by being included in a computer program product. The computer program product may be traded between a seller and a buyer as a product. The computer program product may be distributed in the form of a device-readable storage medium (e.g., compact disc read-only memory (CD-ROM)), or distributed online (e.g., download or upload) through an application store (e.g., Play Store™) or directly between two user devices (e.g., smartphones). In the case of online distribution, at least a portion of the computer program product may be temporarily stored or temporarily created in a device-readable storage medium, such as the memory of a manufacturer's server, an application store's server, or a relay server.

[0107] Those skilled in the art related to the embodiments described above will understand that they may be implemented in modified forms without departing from the essential characteristics of the description. Therefore, the disclosed methods should be considered in an illustrative rather than a restrictive sense. The scope of the invention is defined by the claims, not by the foregoing description, and all variations within the scope of equivalence should be interpreted as being included in the invention. Explanation of the symbols

[0109] 100: Device 110: Receiver 120: Processor 130: Output section 310: Administrator Device 320: Database 330: Shooting device 410: Target fish object image 420: Mask image 510: Segment line 520: Key point 610: Multiple edge points 620: Multiple edge projection points 810: Delaunay Triangulation 820: Ear-Clipping Algorithm 821: External dividing triangle 822: Internal dividing triangle

Claims

Claim 1 A method for a device to provide fish object characteristic information for a target fish object comprises: a receiving unit acquiring a target fish object image for the target fish object; a processor acquiring depth information for the target fish object based on the target fish object image; the processor determining a plurality of key points determined according to the shape of the target fish object and used to represent the characteristics of the target fish object using the target fish object image and the depth information; and the processor acquiring fish object characteristic information including information on at least one of a plurality of edge points, weight, total length, body length, body height, head length, caudal peduncle height, caudal peduncle length, and body width for the target fish object based on the plurality of key points. A method comprising: a step in which an output unit provides the fish object characteristic information; and further comprising a step in which the processor obtains a mask image that distinguishes and represents an internal region of the target fish object and an external region of the target fish object using the target fish object image; wherein the step of obtaining the fish object characteristic information comprises: a step in which the processor obtains a first depth value for the external region using the mask image; a step in which the processor obtains a second depth value for the most derived point in the internal region using the mask image; and a step in which the processor determines the body width based on the difference between the first depth value and the second depth value. Claim 2 delete Claim 3 delete Claim 4 A method according to claim 1, further comprising: a step of obtaining a plurality of segmented triangles based on connections between the plurality of edge points; a step of obtaining a plurality of internal segmented triangles in which a plurality of external segmented triangles corresponding to an external region among the plurality of segmented triangles are excluded using the mask image; and a step of obtaining 3D modeling information for the target fish object using the plurality of internal segmented triangles. Claim 5 A method according to claim 1, further comprising: a step in which the processor determines a plurality of edge projection points corresponding to each of the plurality of edge points and representing points on which the plurality of edge points are projected; a step in which the processor determines a plurality of edge triangles obtained based on the plurality of edge projection points and the plurality of edge points; a step in which the processor performs lateral modeling for the target fish object based on the plurality of edge triangles; and a step in which the output unit provides lateral modeling information obtained as the lateral modeling is performed. Claim 6 The method according to claim 1 further comprises: a step in which the processor acquires a similar fish object image corresponding to the type of the target fish object; and a step in which the processor acquires a segment image in which a plurality of internal regions of the target fish object are divided into segment lines according to the result of comparison between the target fish object image and the similar fish object image, wherein the plurality of internal regions include a part of the body, upper fin, lower fin, and tail fin of the target fish object, and the segment line includes at least one of the plurality of key points. Claim 7 A method according to claim 1, further comprising: a step in which, when the ratio of the body width to the total length of the target fish object is greater than or equal to a threshold ratio, the processor determines a characteristic information error probability representing an error probability regarding the fish object characteristic information; and a step in which the output unit provides an information abnormality notification regarding the characteristic information error probability; wherein the information abnormality notification is determined differently depending on the difference value between the ratio and the threshold ratio. Claim 8 A method according to claim 1, further comprising the step of determining an obesity rate representing the health status of a target fish corresponding to the target fish object according to the ratio of the length value for the body height and body length of the target fish object and the weight value for the weight. Claim 9 A device for providing fish object characteristic information for a target fish object, comprising: a receiver for acquiring a target fish object image for a target fish object; and a processor for acquiring depth information for a target fish object based on the target fish object image, determining a plurality of key points determined according to the shape of the target fish object and used to represent the characteristics of the target fish object using the target fish object image and the depth information, and acquiring fish object characteristic information including information on at least one of a plurality of edge points, weight, total length, body length, body height, head length, caudal peduncle height, caudal peduncle length, and body width for the target fish object based on the plurality of key points. A device comprising: an output unit providing the above-mentioned fish object characteristic information; wherein the processor obtains a mask image that distinguishes and represents an internal region of the target fish object and an external region of the target fish object using the above-mentioned target fish object image, obtains a first depth value for the external region using the mask image, obtains a second depth value for the most derived point in the internal region using the mask image, and determines the body width based on the difference between the first depth value and the second depth value. Claim 10 A computer-readable recording medium having a program stored on it for executing the method of any one of claims 1 and 4 through 8 on a computer.

Images