Biological growth measurement system for smart aquaculture

Abstract

The present invention relates to a smart aquaculture biological growth measurement system and comprises: an image acquisition device that irradiates a grid-patterned laser light source upon the occurrence of an event in a tank to acquire an image of a fish projected by the laser light source; a monitoring server that analyzes the image acquired by the image acquisition device to measure the growth status of the fish, predicts the time of shipment and the amount of feed required until shipment, and provides notifications; and a user terminal that interacts with the image acquisition device and the monitoring server, respectively, through a dedicated application.

Description

Technology Field

[0001] The present invention relates to a fish growth status monitoring technology for the automation of flow-through indoor fish farms, and more specifically, to a technology capable of estimating the growth status of fish by irradiating a fish farm tank with a grid pattern laser light source upon the occurrence of a specific event in the fish farm tank, photographing fish onto which the grid pattern is projected, and analyzing the image. Background Technology

[0003] Securing future food resources has become a major global issue. This is because concerns regarding food supply are growing due to the shutdown of production facilities and reduced trade activities resulting from unexpected situations such as climate change-induced disasters, epidemics, and wars. To prepare for such situations, agricultural technologies incorporating advanced scientific and technological advancements, such as smart farming, are receiving significant attention, particularly in developed nations. This is because such technologies enable the stable harvesting of crops without being constrained by external environmental factors.

[0004] In this context, smart aquaculture is emerging as a major global issue regarding the securing of food resources. This is because fish are the biological resource with the highest protein supply per unit area. While smart aquaculture is currently spreading, primarily in developed European nations, the current environment of domestic fish farms remains poor. The fishing population is aging rapidly, and labor shortages are intensifying in the fisheries sector. Furthermore, the working conditions in the fisheries industry are underdeveloped compared to other sectors, making it a profession that young people tend to avoid. Business conditions are worsening due to a decline in fisheries production caused by environmental changes such as rising water temperatures, as well as a high-cost, low-efficiency structure.

[0005] Therefore, Korea also faces an urgent need to develop smart aquaculture technologies capable of automating processes across the entire aquaculture industry. The development of such technologies may require techniques for managing the tank environment in which fish are active, managing and monitoring the inflow of water into the tanks, automated feeding technologies for supplying food to dozens of tanks in a farm, and technologies for monitoring fish growth status to estimate feed supply amounts and shipping dates. Prior art literature

[0007] Korean Registered Patent No. 10-2409745 (June 13, 2022) The problem to be solved

[0008] One embodiment of the present invention aims to reduce manager fatigue and provide foundational technology for the full-cycle automation of aquaculture farms by automatically measuring the length and volume of fish in the farm, predicting the time of shipment, and providing the manager with the amount of feed required until shipment.

[0009] One embodiment of the present invention aims to provide a technology capable of estimating the size and volume of a fish through pattern distortion analysis by projecting a laser light source with a grid pattern onto a fish and obtaining an image of the fish with the projected grid pattern. means of solving the problem

[0011] Among the embodiments, the smart aquaculture biological growth measurement system comprises: an image acquisition device that irradiates a grid-patterned laser light source upon the occurrence of an event in a tank to acquire an image of a fish projected by said laser light source; a monitoring server that analyzes the image acquired by said image acquisition device to measure the growth status of said fish, predicts the time of shipment and the amount of feed required until shipment, and provides notifications; and a user terminal that interacts with said image acquisition device and said monitoring server, respectively, through a dedicated application.

[0012] The image acquisition device may include: a camera module that is supported at a specific height above the water surface by a support member coupled to the upper side of a tank wall and positioned in a direction facing the water surface to photograph a specific area of ​​the water surface; a laser module that is coupled with the camera module and irradiates a laser light source of the grid pattern toward the specific area; and a control module that detects the movement of the fish on the water surface and photographs an image of the fish projected by the laser light source of the grid pattern.

[0013] The laser module is formed in multiple units according to the size of the specific area, and the grid pattern formed on the water surface overlaps with one another, thereby expanding the overall size of the pattern area.

[0014] The monitoring server can extract a fish image within the image of the fish, calculate the number of grids projected between the head and tail of the fish from the fish image, and estimate the length of the fish using the number of grids and the grid size.

[0015] The monitoring server can calculate the number of grids by reflecting the result of summing the ratios of the distances to the boundaries within each grid for grids that overlap with the boundaries of the head and tail of the fish, respectively.

[0016] The monitoring server can extract a fish image within the image of the fish, extract a line pattern laser light source passing through the fish from the grid pattern laser light source among the fish image, and estimate the volume of the fish using the straight distance and horizontal distance between different points of the line pattern laser light source on the image coordinates.

[0017] The monitoring server can build a deep learning model that estimates the growth state of the fish by generating training images of the fish onto which the laser light source of the grid pattern is projected based on the image of the fish and learning the pattern information of each training image.

[0018] The deep learning model above can be trained to receive feature information regarding changes in the grid pattern extracted from each of the training images as input and to generate length and volume regarding the growth state of the fish as output.

[0019] The monitoring server can record the feed dosage and the length and volume of the fish through a growth status monitoring program, and analyze the correlation between the feed dosage and the length and volume of the fish to predict the shipping time and the amount of feed required until shipping. Effects of the invention

[0021] The disclosed technology may have the following effects. However, this does not mean that a specific embodiment must include all of the following effects or only the following effects; therefore, the scope of the rights of the disclosed technology should not be understood as being limited by this.

[0022] A smart aquaculture biological growth measurement system according to one embodiment of the present invention can irradiate a grid-patterned laser light source into an aquaculture tank when a specific event occurs in the aquaculture tank, photograph a fish with a projected grid pattern, and estimate the growth status of the fish through image analysis.

[0023] Furthermore, by monitoring the growth status of fish and providing information on the time required until shipment and the amount of feed, this invention reduces work fatigue, enables the establishment of specific plans related to aquaculture farm operations, and allows for the management of multiple tanks within the farm with minimal manpower. Additionally, since this invention can be applied not only to fish farms but also to livestock breeding facilities or greenhouses cultivating specialty crops, it can be utilized across all sectors of future agriculture and livestock farming, such as smart farms. Brief explanation of the drawing

[0025] FIG. 1 is a diagram illustrating a biological growth measurement system according to the present invention. Figure 2 is a diagram illustrating the physical configuration of the image acquisition device of Figure 1. Figure 3 is a diagram illustrating the functional configuration of the monitoring server of Figure 1. FIG. 4 is a flowchart illustrating an example of a smart aquaculture biological growth measurement process according to the present invention. FIG. 5 is a diagram illustrating an embodiment of a biological growth measurement system according to the present invention. FIG. 6 is a drawing illustrating an embodiment of a method for installing a camera module and a laser module according to the present invention. FIG. 7 is a drawing illustrating an embodiment of a method for expanding the grating pattern of a laser light source according to the present invention. Figure 8 is a diagram illustrating the process of estimating fish length through image processing according to the present invention. FIG. 9 is a diagram illustrating the process of estimating fish volume through image processing according to the present invention. FIG. 10 is a diagram illustrating the process of estimating fish volume through AI image processing according to the present invention. Specific details for implementing the invention

[0026] The description of the present invention is merely an example for structural or functional explanation, and therefore the scope of the present invention should not be interpreted as being limited by the examples described in the text. That is, since the examples are subject to various modifications and may take various forms, the scope of the present invention should be understood to include equivalents capable of realizing the technical concept. Furthermore, the objectives or effects presented in the present invention do not imply that a specific example must include all of them or only such effects; therefore, the scope of the present invention should not be understood as being limited by them.

[0027] Meanwhile, the meaning of the terms described in this application should be understood as follows.

[0028] Terms such as "first," "second," etc., are intended to distinguish one component from another, and the scope of rights shall not be limited by these terms. For example, the first component may be named the second component, and similarly, the second component may be named the first component.

[0029] When it is stated that one component is "connected" to another component, it should be understood that it may be directly connected to that other component, or that there may be other components in between. Conversely, when it is stated that one component is "directly connected" to another component, it should be understood that there are no other components in between. Meanwhile, other expressions describing the relationships between components, such as "between" and "exactly between," or "adjacent to" and "directly adjacent to," should be interpreted in the same way.

[0030] A singular expression should be understood to include a plural expression unless the context clearly indicates otherwise, and terms such as "include" or "have" are intended to specify the existence of the implemented features, numbers, steps, actions, components, parts, or combinations thereof, and should be understood not to preclude the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0031] In each step, identifiers (e.g., a, b, c, etc.) are used for convenience of explanation and do not describe the order of the steps; the steps may occur differently from the specified order unless a specific order is clearly indicated in the context. That is, the steps may occur in the same order as specified, may be performed substantially simultaneously, or may be performed in the reverse order.

[0032] The present invention may be implemented as computer-readable code on a computer-readable recording medium, and the computer-readable recording medium includes all types of recording devices in which data that can be read by a computer system is stored. Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device, etc. Additionally, the computer-readable recording medium may be distributed across networked computer systems, so that computer-readable code can be stored and executed in a distributed manner.

[0033] Unless otherwise defined, all terms used herein have the same meaning as generally understood by those skilled in the art to which this invention pertains. Terms defined in commonly used dictionaries should be interpreted as having meanings consistent with the context of the relevant technology and should not be interpreted as having an ideal or overly formal meaning unless explicitly defined in this application.

[0035] FIG. 1 is a diagram illustrating a biological growth measurement system according to the present invention.

[0036] Referring to FIG. 1, the biological growth measurement system (100) may include an image acquisition device (110), a monitoring server (130), and a user terminal (150).

[0037] The image acquisition device (110) is installed in a space for raising fish and may correspond to a device for measuring the growth status of fish being raised by photographing the said aquaculture space. The image acquisition device (110) may be implemented as one device constituting the smart aquaculture biological growth measurement system (100) according to the present invention, and the smart aquaculture biological growth measurement system (100) may be implemented in various modified forms depending on the purpose, such as analyzing the aquaculture environment or measuring the growth status using the image acquisition device (110). The image acquisition device (110) may be connected to a monitoring server (130) and a user terminal (170), respectively, through a network.

[0038] For example, an operator or manager of a fish farm can directly access the image acquisition device (110) through a user terminal (150) to monitor the current status of fish or the aquaculture environment within the aquaculture space, and can receive and check various information provided by the monitoring server (130) through a dedicated app or program.

[0039] In one embodiment, the image acquisition device (110) can acquire an image of a fish projected by a laser light source of a grid pattern by irradiating it when an event occurs in a tank located in a farming space. To this end, the image acquisition device (110) may be implemented by including various modules for acquiring an image of a fish in a tank. For example, the image acquisition device (110) may include a camera module, a laser module, and a control module, which will be explained in more detail with reference to FIG. 2.

[0040] The monitoring server (130) can be implemented as a server corresponding to a computer or program capable of collecting and analyzing images captured in a fish farming space by an image acquisition device (110), measuring the growth status of fish (e.g., fish, etc.) being farmed in the farming space, and predicting the time of shipment and the amount of feed required until shipment. For example, the monitoring server (130) can be implemented as a cloud server, but is not necessarily limited thereto. The monitoring server (130) can perform monitoring operations regarding the growth status of fish by measuring and recording the size and thickness of the fish through various image analyses. In addition, the monitoring server (130) can store and manage the captured images and measurement results in a database.

[0041] The monitoring server (130) can be connected to the image acquisition device (110) located in the aquaculture space via a wired network or a wireless network such as Bluetooth or WiFi, and can communicate with the image acquisition device (110) through the network. The monitoring server (130) can be connected to the user terminal (150) via a wireless network such as Bluetooth or WiFi, and can communicate with the user terminal (150) through the network. Meanwhile, the specific configuration and operation of the monitoring server (130) will be explained in more detail with reference to FIG. 3.

[0042] The user terminal (150) may be a computing device capable of receiving various monitoring information and notifications regarding the prediction of the shipping time and the amount of feed required until shipping through the biological growth measurement system (100) for smart aquaculture. The user terminal (150) may be implemented as a smartphone, laptop, or computer that can be operated in connection with the monitoring server (130), but is not necessarily limited thereto and may also be implemented as various devices such as a tablet PC.

[0043] Additionally, the user terminal (150) may install and run a dedicated program or application to link with the monitoring server (130). Through this, the user terminal (150) can use various services provided by the monitoring server (130). Meanwhile, the user terminal (150) may be directly connected to the image acquisition device (110), and may also install and run a dedicated program or application for this purpose.

[0045] Figure 2 is a diagram illustrating the physical configuration of the image acquisition device of Figure 1.

[0046] Referring to FIG. 2, the image acquisition device (110) may be configured to include modules for acquiring an image of a fish projected by a laser light source by irradiating a laser light source of a grid pattern when an event occurs in the water tank. The image acquisition device (110) may include a camera module (210), a laser module (230), and a control module (250).

[0047] The camera module (210) is supported by a support attached to the upper side of one side of the water tank wall at a specific height above the water surface and is positioned in a direction facing the water surface to capture a specific area of ​​the water surface. Here, the support may correspond to a rod-shaped structure fixedly installed on the water tank wall (or column surface) to support the camera module (210). A clamp for attachment to the water tank wall may be fixedly attached to one end of the support, and a joint connected to another support may be attached to the other end. The camera module (210) may be supported through the support so as to be perpendicular to the water surface of the water tank. That is, the camera module (210) may be positioned at a specific height above the water surface, and the height above the water surface may be adjusted to a height at which the laser light source is clearly visible. The camera module (210) can generate an image by capturing a certain area in a direction facing the water surface, and the field of view of the image may be determined according to the height above the water surface.

[0048] The laser module (230) is linked with the camera module (210) and can irradiate a laser light source with a grid pattern toward a specific area. The laser module (230) can be implemented to operate in conjunction with the operation of the camera module (210) and can be supported by a support at a specific height vertically away from the water surface, just like the camera module (210). The laser module (230) can irradiate a laser light source forming a specific pattern toward a specific area of ​​the water surface being photographed by the camera module (210). To this end, the laser module (230) can be combined with a pattern module for forming a specific pattern.

[0049] In one embodiment, the laser module (230) is formed in multiple units according to the size of a specific area, and the grid pattern formed on the water surface overlaps with one another, thereby expanding the overall size of the pattern area. That is, the laser module (230) can be installed in multiple units as needed, and the grid pattern irradiated by each laser module (230) can form independent pattern areas on the water surface depending on the installation position and direction of the module. For example, when each laser module (230) forms a square-shaped grid pattern of a specific size on the water surface, the boundaries of each grid pattern can be arranged to overlap each other to form a single large square grid pattern, thereby expanding the size of the pattern area formed on the water surface. This will be explained in more detail with reference to FIG. 7.

[0050] The control module (250) can control the overall operation of the image acquisition device (110) and manage the control flow or data flow between the camera module (210) and the laser module (230). More specifically, the control module (250) can detect the movement of fish on the water surface and capture an image of the fish with a grid pattern laser light source projected onto it. The control module (250) can detect the movement of the water surface through the camera module (210), and when the movement of the fish is detected, it can control the laser module (230) to irradiate a grid pattern laser light source, and can capture an image of the fish coming up to the surface while the laser light source is being irradiated.

[0052] Figure 3 is a diagram illustrating the functional configuration of the monitoring server of Figure 1.

[0053] Referring to FIG. 3, the monitoring server (130) can perform the operation of analyzing an image acquired by an image acquisition device to measure the growth status of fish, predict the time of shipment and the amount of feed required until shipment, and provide a notification. To this end, the monitoring server (130) may include an image receiving unit (310), a growth measuring unit (330), a shipment prediction unit (350), and a control unit (370).

[0054] At this time, embodiments of the present invention are not required to include all of the above components simultaneously; depending on each embodiment, some of the components may be omitted, or some or all of the components may be selectively included. The operation of each component will be described in detail below.

[0055] The image receiving unit (310) can receive images of fish from the image acquisition device (110). Here, the fish may correspond to fish raised in a tank of a fish farm, but is not necessarily limited thereto and may include various animals and plants whose growth status can be measured through image analysis. The image receiving unit (310) can receive images of fish from multiple image acquisition devices (110), and in this case, it may receive a single image in which the image signals captured by each image acquisition device (110) are mixed by a video mixer. Additionally, the image receiving unit (310) can store the images of fish received from the image acquisition device (110) in a database, and may store them after performing image preprocessing or image classification operations as needed.

[0056] The growth measurement unit (330) can measure the growth status of fish by analyzing images received from the image acquisition device (110). At this time, the growth status of fish may include the length (size) and thickness (volume) of the fish. That is, the growth measurement unit (330) can collect data regarding the growth status of fish through image analysis and can record and store the collected data in a database by date. In addition, the growth measurement unit (330) can estimate the types, number, and movement characteristics of fish present in the tank through image analysis.

[0057] In one embodiment, the growth measurement unit (330) can extract a fish image within a fish image, calculate the number of grids projected between the head and tail of the fish from the fish image, and estimate the length of the fish using the number of grids and the grid size. Here, the fish image may correspond to an image of an area containing the fish within the image. The growth measurement unit (330) can extract a bounding box containing the fish through object detection on the fish image and generate an image of the bounding box as a fish image.

[0058] Additionally, the growth measuring unit (330) can identify the head and tail of the fish in the fish image and count the number of grids based on a single straight line extending from the tip of the head to the tip of the tail. The growth measuring unit (330) can calculate the length of the fish by multiplying the number of grids and the grid size by utilizing the fact that the size of the unit grid forming the grid pattern is constant.

[0059] In one embodiment, the growth measuring unit (330) can calculate the number of grids by reflecting the result of summing the ratios of the distances to the boundaries within each grid for grids that overlap with the boundaries of the head and tail of the fish, respectively. This will be explained in more detail with reference to FIG. 8.

[0060] In one embodiment, the growth measuring unit (330) can extract a fish image within the image of the fish, extract a line pattern laser light source passing through the fish from the fish image among the grid pattern laser light sources, and estimate the volume of the fish using the straight distance and horizontal distance between different points of the line pattern laser light source on the image coordinates. This will be explained in more detail with reference to FIG. 9.

[0061] In one embodiment, the growth measurement unit (330) may build a deep learning model that estimates the growth state of a fish by generating training images of a fish onto which a laser light source of a grid pattern is projected based on an image of a fish and learning the pattern information of each training image. In this case, the deep learning model may be trained to receive feature information regarding changes in the grid pattern extracted from each training image as input and to generate length and volume regarding the growth state of the fish as output. This will be explained in more detail with reference to FIG. 10.

[0062] The shipment prediction unit (350) can predict the shipment time and the amount of feed required until shipment based on the measured growth status and provide a notification to the user terminal (150). At this time, the user can establish a feed supply plan and a shipment plan based on the predicted shipment time and the amount of feed required until shipment, and efficiently operate the fish farm according to the plan. In one embodiment, the shipment prediction unit (350) can generate a two-dimensional function graph that can check the growth status of fish based on growth data recorded by date using a regression algorithm.

[0063] Here, regression can be considered a type of supervised learning in machine learning; unlike classification, where predicted values ​​are continuous, the predicted values ​​can be discrete. In machine learning, regression can be defined as the process of modeling the correlation between multiple independent variables and a single dependent variable, and it can proceed as a process of predicting the dependent variable by finding the optimal values ​​of regression coefficients that influence the independent variables.

[0064] The control unit (370) controls the overall operation of the monitoring server (130) and can manage the control flow or data flow between the image receiving unit (310), the growth measuring unit (330), and the shipment prediction unit (350).

[0066] FIG. 4 is a flowchart illustrating an example of a smart aquaculture biological growth measurement process according to the present invention.

[0067] Referring to FIG. 4, the biological growth measurement system (100) can acquire an image of a fish through an image acquisition device (110) (step S410). That is, the image acquisition device (110) can acquire an image of a fish projected by irradiating a laser light source with a grid pattern when an event occurs in the tank.

[0068] Additionally, the biological growth measurement system (100) can measure the growth status of fish by analyzing an image acquired by an image acquisition device through a monitoring server (130) (step S430). That is, the monitoring server (130) can extract a fish image from the image of the fish and analyze the fish image to estimate the length and volume of the fish, respectively.

[0069] Additionally, the biological growth measurement system (100) can predict the time of shipment and the amount of feed required until shipment through the monitoring server (130) (step S450). That is, the monitoring server (130) can record the feed dosage and the length and volume of the fish through a growth status monitoring program, and analyze the correlation between the feed dosage and the length and volume of the fish to predict the time of shipment and the amount of feed required until shipment. The growth status monitoring program can calculate the number of remaining days to reach the target body size and length based on the feed dosage and length and volume information, and can predict the time of shipment based on this.

[0070] Accordingly, the aquaculture farm manager can check the amount of feed currently in storage and adjust the feed supply rate required for farm operations. Meanwhile, the growth status monitoring program can be implemented to run on a web server, allowing the aquaculture farm manager to monitor the growth status of the fish from anywhere in an internet environment.

[0072] FIG. 5 is a diagram illustrating an embodiment of a biological growth measurement system according to the present invention.

[0073] Referring to FIG. 5, the biological growth measurement system (100) can be applied to a fish farm tank (510) for raising flounder with a diameter of about 5 cm. A boom is installed on one side of the tank, and a grid laser device (520) can be installed on the boom. The grid laser device (520) can be controlled by a growth status monitoring program, and when a fish is detected in the image of a variable focus camera (530) installed vertically above the water surface of the fish farm tank (510), a grid pattern laser beam (540) with a width and height of 30 cm each, composed of unit grids where each grid is 5 mm in width and height, can be projected onto the fish. The variable focus camera (530) can generate photos of the fish taken periodically (e.g., 2 photos per second) over a specific period of time and can be stored in a fish farm management computer (550).

[0074] The fish farm management computer (550) can be connected to a cloud server (560), and photos stored in the fish farm management computer (550) can be uploaded to the cloud server (560) in real time. The growth status monitoring program can also operate on the cloud server (560), and can analyze the uploaded photos to estimate the size and volume of the fish, and estimate the shipment date and the amount of feed required from the size and volume data. Meanwhile, the grid laser control box (570) can turn on and off the power connected to the grid laser device (520) by receiving an on / off signal from the growth status monitoring program.

[0076] FIG. 6 is a drawing illustrating an embodiment of a method for installing a camera module and a laser module according to the present invention, and FIG. 7 is a drawing illustrating an embodiment of a method for expanding a grid pattern of a laser light source according to the present invention.

[0077] Referring to FIG. 6, a boom (i.e., a rod-shaped pole) can be installed using one column face of the water tank to install the laser module (230), and the laser module (230) can be installed using a clamp so as to be perpendicular to the water surface of the water tank. At this time, the height relative to the water surface can be adjusted to a height where the laser light source is clearly visible.

[0078] Referring to FIG. 7, the total size of the pattern area formed by the grid pattern can be expanded by installing a plurality of laser modules (230) according to the size of the tank and arranging the grid patterns irradiated by each laser module (230) so that they overlap each other at the boundary portion.

[0079] Meanwhile, the environment of the fish farm may correspond to a relatively dark environment without lighting, and since most of the fish in the farm have dark colors, a laser light source irradiated through the laser module (230) may be a green laser with good visibility. For example, the laser light source may be generated at a wavelength of 510 nm by a laser module (230) having an input voltage of 3 V and an output power of 10 mW.

[0080] When power is supplied to the laser module (230), a laser light source with a grid pattern similar to a checkerboard can be directed toward the water surface. The laser module (230) can operate in conjunction with the camera module (210), and if a specific event occurs, such as feeding, while the camera module (210) is capturing an image of the water surface, the movement of the fish is detected and the operation of the laser module (230) can be initiated. When a fish moves into the grid pattern projected by the laser module (230), the camera module (210) can acquire an image of the fish with the grid pattern projected onto it. In particular, the camera module (210) may use a model capable of variable focus to increase image clarity and may be pre-configured to capture the best image according to height.

[0081] Meanwhile, the control module (250) connected to the camera module (210) and the laser module (230) may be implemented by being included inside a computing device, and in this case, the camera module (210) and the laser module (230) may each be connected to the computing device.

[0083] Figure 8 is a diagram illustrating the process of estimating fish length through image processing according to the present invention.

[0084] Referring to FIG. 8, the image acquisition device (110) can acquire an image of a fish onto which a laser light source of a grid pattern is projected, and can transmit the image of the fish to the monitoring server (130). The monitoring server (130) can execute a biological growth measurement program to analyze the image of the fish and perform length estimation. Here, the biological growth measurement program can perform the operation of measuring and recording the growth status of the fish based on image analysis. The monitoring server (130) can extract a fish image from the image of the fish and can identify the head and tail of the fish within the fish image. The monitoring server (130) can count the number of grids from the head to the tail of the fish.

[0085] At this time, the monitoring server (130) can calculate the percentage of the length of the area overlapping with the grid that overlaps with the boundary of the head and tail portions of the fish, and then reflect this in the number of grids. The monitoring server (130) can estimate the actual length information of the fish as a result of multiplying the number of grids by the length of the actual unit grid.

[0086] In Fig. 8, the grid count of the head remainder within the grid overlapping with the boundary of the head portion is 0.8, and the grid count of the tail remainder is 0.2, so the grid counts of the head and tail portions can be added together to add a count of 1. Additionally, if the total grid number between the head and tail of the fish is 21 and the real size of the unit grid is 5 mm, the fish length can be estimated as 5 × 21 = 105 mm.

[0088] FIG. 9 is a diagram illustrating the process of estimating fish volume through image processing according to the present invention.

[0089] Referring to FIG. 9, the image acquisition device (110) can acquire an image of a fish onto which a grid pattern laser light source is projected and can transmit the image of the fish to a monitoring server (130). The monitoring server (130) can execute a biological growth measurement program to analyze the image of the fish and perform volume estimation. The monitoring server (130) can extract a fish image from the image of the fish and can extract a line pattern laser light source passing through the fish from the image of the fish among the grid pattern laser light sources.

[0090] Figure 9 corresponds to a case where a single line laser projects onto a fish. In this case, if the straight-line distance between the pointers of the line laser is denoted as a, the distance between the pointers on a curve with volume can be measured as the horizontal distance b in a two-dimensional image. Therefore, if p0 is a reference representing the height in a flat state, the height of d1 corresponding to p1 can be simply calculated using the law of triangles. Additionally, the height corresponding to p2 can be calculated through d1+d2.

[0092] FIG. 10 is a diagram illustrating the process of estimating fish volume through AI image processing according to the present invention.

[0093] Referring to FIG. 10, a monitoring server (130) can build a deep learning model that estimates the growth state of a fish by generating training images of a fish onto which a laser light source of a grid pattern is projected based on an image of a fish and learning the pattern information of each training image. In this case, the deep learning model can be trained to receive feature information regarding changes in the grid pattern extracted from each training image as input and to generate length and volume regarding the growth state of the fish as output.

[0094] For example, in Fig. 10, volume estimation can be performed by using a fish thickness model that has learned the changes in patterns according to the volume of the fish, matching the pattern most similar to the pattern information of the image data collected in real time, and estimating the volume value of the matched pattern as the volume of the fish. That is, the curvature of the grid pattern projected onto the fish can vary as the thickness of the fish changes, and by constructing a fish thickness model by generating training data from images collected by projecting a grid pattern laser light source onto each fish of different volumes and sizes, the length and volume of the fish can be estimated with increased precision even in a non-linear relationship between shape and body size.

[0095] In addition, since the reliability of result data can be increased as the amount of training data increases for deep learning models, training data needs to be generated using as much video data as possible for each type of fish.

[0097] Although the present invention has been described above with reference to preferred embodiments, those skilled in the art will understand that various modifications and changes can be made to the invention without departing from the spirit and scope of the invention as described in the following claims. Explanation of the symbols

[0099] 100: Biological Growth Measurement System 110: Image acquisition device 130: Monitoring server 150: User terminal 210: Camera module 230: Laser module 250: Control Module 310: Image receiver 330: Growth measuring unit 350: Shipment Forecasting Unit 370: Control Unit

Claims

Claim 1 An image acquisition device that acquires an image of a fish projected by a grid pattern laser light source by irradiating the grid pattern laser light source when an event occurs in a tank; a monitoring server that analyzes the image acquired by the image acquisition device to measure the growth state of the fish, extracts a fish image within the image of the fish, extracts a line pattern laser light source passing through the fish among the grid pattern laser light sources from the fish image, estimates the volume of the fish using the straight and horizontal distances between different points of the line pattern laser light source on the image coordinates, predicts the time of shipment and the amount of feed required until shipment, and provides a notification; and a user terminal that links with the image acquisition device and the monitoring server, respectively, through a dedicated application; wherein the image acquisition device comprises: a camera module that is supported at a specific height from the water surface by a support attached to the upper side of one end of the tank wall and is positioned in a direction facing the water surface to photograph a specific area of ​​the water surface; and a laser module that links with the camera module and irradiates the grid pattern laser light source toward the specific area. A smart aquaculture biological growth measurement system characterized by including a control module that detects the movement of the fish on the water surface and captures an image of the fish onto which a laser light source of the grid pattern is projected. Claim 2 delete Claim 3 A smart aquaculture biological growth measurement system according to claim 1, wherein the laser module is formed in a plurality according to the size of the specific area, and the grid pattern formed on the water surface overlaps with one another to expand the overall size of the pattern area. Claim 4 A smart aquaculture biological growth measurement system according to claim 1, wherein the monitoring server extracts a fish image within the image of the fish, calculates the number of grids projected between the head and tail of the fish from the fish image, and estimates the length of the fish using the number of grids and the size of the grids. Claim 5 A smart aquaculture biological growth measurement system according to claim 4, wherein the monitoring server calculates the number of grids by reflecting the result of summing the ratios of the distances to the boundaries within each grid in the case of grids that overlap with the boundaries of the head and tail of the fish, respectively. Claim 6 delete Claim 7 A smart aquaculture biological growth measurement system according to claim 1, characterized in that the monitoring server generates learning images of fish onto which a laser light source of the grid pattern is projected based on an image of the fish, and constructs a deep learning model that learns pattern information of each learning image to estimate the growth state of the fish. Claim 8 A smart aquaculture biological growth measurement system according to claim 7, characterized in that the deep learning model receives as input feature information regarding changes in the grid pattern extracted from each of the training images and is trained to generate as output length and volume regarding the growth state of the fish. Claim 9 A smart aquaculture biological growth measurement system according to claim 1, characterized in that the monitoring server records the feed dosage and the length and volume of the fish through a growth status monitoring program, and analyzes the correlation between the feed dosage and the length and volume of the fish to predict the shipping time and the amount of feed required until shipping.

Images