Apparatus and method for predicting fishing vessel distribution using nighttime light satellite data

Abstract

The present invention relates to a device and method for predicting the distribution of fishing vessels using nighttime light satellite data. A fishing vessel distribution prediction device utilizing nighttime light satellite data according to one embodiment of the present invention comprises: an extraction unit that extracts light data identified as fishing vessels from nighttime light satellite data at a reference time transmitted from a satellite; a distribution map generation unit that generates a nighttime fishing vessel distribution map based on the light data identified as fishing vessels; and a prediction unit that predicts the nighttime fishing vessel distribution after the reference time based on the nighttime fishing vessel distribution map and the light data identified as fishing vessels prior to the reference time.

Description

Technology Field

[0001] The present invention relates to an apparatus and method for predicting the distribution of fishing vessels using nighttime light satellite data. More specifically, it relates to an apparatus and method for predicting the distribution of fishing vessels using nighttime light satellite data that can improve the visibility of fishing activities and maximize fishing efficiency by identifying the location of fishing vessels based on nighttime light satellite data and predicting future fishing vessel distribution by learning from past fishing vessel distribution data. Background Technology

[0002] Satellite data, including images and footage captured by satellites, is utilized in various fields such as weather forecasting, marine environment monitoring, urban development, and disaster monitoring.

[0003] In particular, the field of marine observation is actively utilizing satellite data for purposes such as monitoring fishing activities, managing marine resources, and monitoring illegal fishing.

[0004] In utilizing such satellite data, it is necessary to analyze both daytime and nighttime satellite data due to the nature of fishing activities, where fish species and methods vary depending on the time of day. In particular, while daytime satellite data facilitates the detection of fishing vessels by analyzing the visible light spectrum, the lack of natural light at night makes it difficult to effectively identify the location of vessels using visible light alone; therefore, methods based on lights in the open sea are generally used to detect fishing vessels.

[0005] Thus, if the distribution of fishing vessels, which is most closely related to the fishing environment, can be analyzed from satellite data, it becomes possible to maximize fishing efficiency by identifying trends in fishing grounds in real time, and furthermore, to manage marine resources more efficiently by predicting optimal fishing times through forecasting changes in fishing grounds.

[0006] However, due to recent climate change, the fishing environment in the waters off the Korean Peninsula, as well as globally, is changing rapidly. Consequently, the habitats of fish species are shifting, and predicting fishing seasons is becoming increasingly difficult.

[0007] Therefore, as variability in the fishing environment increases, the distribution of fishing vessels is naturally changing with significant variability as well; consequently, more rapid and accurate fishing vessel distribution prediction technology is required to maintain fishing efficiency and effectively manage marine resources.

[0008] However, it is practically impossible to directly observe and analyze the extensive marine environment in real time, and existing data-based analysis technologies also suffer from reduced effectiveness due to spatial and temporal constraints.

[0009] In addition, conventional technology generally identified fishing vessels based solely on the brightness of artificial lights in nighttime satellite data analysis. This presented limitations, such as the difficulty in distinguishing them from other light sources (e.g., offshore platforms, harbor lights, lights of other vessels, etc.) and the high likelihood of noise occurring due to changes in the marine environment or weather conditions.

[0010] Accordingly, there is a need for technology capable of more precise detection and prediction of fishing vessels from nighttime satellite data. Prior art literature

[0011] Republic of Korea Registered Patent Publication No. 10-2021-0018751 The problem to be solved

[0012] The present invention was devised to solve the problems described above, and the objective of the present invention is to provide a device and method for predicting the distribution of fishing vessels using nighttime light satellite data, which can distinguish between fishing vessels and other light sources by applying artificial intelligence to nighttime light satellite data.

[0013] The objectives of the present invention are not limited to the problems mentioned above, and other objectives and advantages of the present invention not mentioned may be understood from the following description and will be more clearly understood by the embodiments of the present invention. Furthermore, it will be understood that the objectives and advantages of the present invention can be realized by the means and combinations thereof set forth in the claims. means of solving the problem

[0014] For the above purpose, a fishing vessel distribution prediction device utilizing nighttime light satellite data according to one embodiment of the present invention comprises: an extraction unit that extracts light data identified as fishing vessels from nighttime light satellite data at a reference time transmitted from a satellite; a distribution map generation unit that generates a nighttime fishing vessel distribution map based on the light data identified as fishing vessels; and a prediction unit that predicts the nighttime fishing vessel distribution after the reference time based on the nighttime fishing vessel distribution map and the light data identified as fishing vessels before the reference time.

[0015] Meanwhile, a method for predicting the distribution of fishing vessels using nighttime light satellite data according to one embodiment of the present invention comprises: a step of extracting light data identified as fishing vessels from nighttime light satellite data at a reference time transmitted from a satellite; a step of generating a nighttime fishing vessel distribution map based on the light data identified as fishing vessels; and a step of predicting the nighttime fishing vessel distribution after the reference time based on the nighttime fishing vessel distribution map and the light data identified as fishing vessels before the reference time.

[0016] And, according to another embodiment of the present invention, a fishing vessel distribution prediction device utilizing nighttime light satellite data comprises: a memory; and at least one processor, wherein the at least one processor extracts light data identified as fishing vessels from nighttime light satellite data at a reference time transmitted from a satellite, generates a nighttime fishing vessel distribution map based on the light data identified as fishing vessels, and predicts the nighttime fishing vessel distribution after the reference time based on the nighttime fishing vessel distribution map and the light data identified as fishing vessels prior to the reference time. Effects of the invention

[0017] According to the present invention, by applying artificial intelligence to nighttime light satellite data, it is possible to effectively distinguish between fishing boats and other light sources.

[0018] In addition, according to the present invention, the accuracy of analysis of fishing vessels in nighttime light satellite data can be improved through the correction of missing data.

[0019] Furthermore, according to the present invention, by predicting the future distribution of fishing vessels, it has the effect of maximizing fishing efficiency and enabling effective management of marine resources. Brief explanation of the drawing

[0020] FIG. 1 is a diagram illustrating the entire process of predicting the distribution of fishing vessels according to one embodiment of the present invention. FIG. 2 is a configuration diagram of a fishing boat distribution prediction device according to one embodiment of the present invention. FIG. 3 is a flowchart of a method for predicting the distribution of fishing vessels according to one embodiment of the present invention. Specific details for implementing the invention

[0021] Hereinafter, embodiments disclosed in this specification will be described in detail with reference to the attached drawings. Identical or similar components regardless of drawing symbols are assigned the same reference number, and redundant descriptions thereof will be omitted. The suffixes "module" and "part" used for components in the following description are assigned or used interchangeably solely for the ease of drafting the specification and do not inherently possess distinct meanings or roles. Furthermore, in describing the embodiments disclosed in this specification, if it is determined that a detailed description of related prior art could obscure the essence of the embodiments disclosed in this specification, such detailed description will be omitted. Additionally, the attached drawings are intended only to facilitate understanding of the embodiments disclosed in this specification; the technical concept disclosed in this specification is not limited by the attached drawings, and it should be understood that they include all modifications, equivalents, and substitutions that fall within the spirit and technical scope of the invention.

[0022] Terms including ordinal numbers, such as first, second, etc., may be used to describe various components, but said components are not limited by said terms. These terms are used solely for the purpose of distinguishing one component from another.

[0023] When it is stated that one component is "connected" or "connected" to another component, it should be understood that while it may be directly connected or connected to that other component, there may also be other components in between. On the other hand, when it is stated that one component is "directly connected" or "directly connected" to another component, it should be understood that there are no other components in between.

[0024] One embodiment of the present invention deals with a fishing vessel distribution prediction technology utilizing nighttime light satellite data.

[0025] As variability in fishing activities increases due to recent climate change and rapid changes in the marine environment, there is a growing need for forecasting technologies that can more accurately analyze the movement routes of fishing vessels and respond quickly to changes in fishing grounds.

[0026] In particular, while daytime satellite data based on visible light makes it easier to detect fishing vessels, analysis utilizing the lights of fishing vessels at night is required; therefore, more precise fishing vessel detection and prediction technology utilizing nighttime light satellite data is needed.

[0027] For example, in conventional technology, the position of a fishing vessel was generally estimated by using nighttime satellite light data to detect lights of a certain brightness or higher.

[0028] However, this method had a problem where it was difficult to distinguish from offshore platforms, harbor lighting, and other vessels, and data accuracy was degraded due to factors such as clouds, atmospheric interference, and sea surface reflection.

[0029] Furthermore, the high likelihood of missing values ​​in satellite imagery data has raised the issue that it is difficult to reliably analyze the actual movement paths of fishing vessels based solely on light detection.

[0030] Accordingly, there is a need for technology that analyzes the lights of fishing vessels more precisely, utilizes artificial intelligence to effectively distinguish differences from other light sources, and corrects missing data to improve prediction accuracy.

[0031] Accordingly, in one embodiment of the present invention, a new method is proposed to effectively detect fishing vessels by analyzing nighttime light satellite data based on artificial intelligence, and to predict future fishing vessel distribution by analyzing the movement patterns of fishing vessels through learning past fishing vessel distribution data.

[0032] In this regard, FIG. 1 is a diagram illustrating the entire process of predicting the distribution of fishing vessels according to one embodiment of the present invention, FIG. 2 is a configuration diagram of a fishing vessel distribution prediction device according to one embodiment of the present invention, and FIG. 3 is a flowchart of a fishing vessel distribution prediction method according to one embodiment of the present invention.

[0033] Referring to FIGS. 1 to 3, a fishing boat distribution prediction device (100) according to one embodiment of the present invention may be configured to include an extraction unit (110), a distribution map generation unit (120), a prediction unit (130), etc.

[0034] A fishing vessel distribution prediction device (100) according to one embodiment of the present invention identifies the lights of fishing vessels in nighttime light satellite data using an extraction unit (110), generates a nighttime fishing vessel distribution map based on the location information of fishing vessels through a distribution map generation unit (120), and predicts the distribution of fishing vessels at a later point in time by analyzing past fishing vessel distribution data through a prediction unit (130), thereby enabling more reliable detection and prediction of fishing vessels.

[0035] Below, we will provide a more detailed explanation of the configuration of each function, such as the extraction unit (110), distribution map generation unit (120), and prediction unit (130), for realizing this.

[0036] The extraction unit (110) can extract light data identified as fishing boats from night light satellite data at a reference time transmitted from a satellite (see step S210).

[0037] Here, the reference point in time may refer to the time when nighttime light satellite data collected from satellites at a specific time is analyzed.

[0038] However, the reference point can be set in various ways in addition to the aforementioned analysis point, and can be classified into the latest standard, a specific point in time standard, or a periodic data analysis standard depending on the purpose of the analysis.

[0039] For example, the latest standard may refer to the time when the latest night light satellite data is received, and can be used when real-time or near-real-time monitoring of fishing vessels is required.

[0040] A specific point in time standard may refer to nighttime light satellite data collected during specific fishing periods in the past, and may be used when analyzing climate change or long-term fishing ground movement patterns, such as nighttime light satellite data collected on March 1 of each year from 2020 to 2024 or nighttime light satellite data measured during red tide periods from 2020 to 2024.

[0041] Periodic data analysis criteria may refer to night light satellite data collected over a certain period, and can be used for daily and weekly analysis and prediction of fishing vessel distribution, such as based on 23:00 UTC every day or 00:00 UTC every Monday.

[0042] Next, the extraction unit (110) inputs nighttime light satellite data at a reference time into the first artificial intelligence model to extract light data identified as a fishing vessel.

[0043] Here, the first artificial intelligence model may be a deep learning-based model trained to identify only the light data of fishing boats from the learning nighttime light satellite data.

[0044] For example, the first artificial intelligence model according to one embodiment of the present invention may be an artificial intelligence model based on an artificial neural network (ANN) or a convolutional neural network (CNN), and may be configured to extract light data of fishing vessels from nighttime light satellite data.

[0045] If the first artificial intelligence model is a model based on an ANN, characteristic data such as brightness, color, and intensity of night lights included in night light satellite data is used as input values, and the trained neural network is used to distinguish between fishing boat lights and other light sources and to extract fishing boat light data.

[0046] For example, a first artificial intelligence model based on an ANN learns that the lights of fishing vessels are distinguished from other lights (e.g., harbor lights, vessels other than fishing vessels, etc.) according to specific factors (e.g., clouds over time, changes in light brightness due to changes in atmospheric conditions, changes in the intensity of fishing vessel lights depending on whether fishing is taking place, etc.), thereby enabling the extraction of fishing vessel light data.

[0047] As such, it can be seen that utilizing ANN allows for the partial correction of brightness fluctuations or environmental noise that may occur in nighttime satellite light data, thereby enabling the extraction of accurate fishing vessel light data.

[0048] In addition, since the first artificial intelligence model based on ANN learns patterns based on structured characteristics, it can provide higher classification performance than simple threshold-based analysis.

[0049] However, for the characteristic data that is the input value of the first artificial intelligence model based on the ANN, the extraction unit (110) may be configured to perform a separate data preprocessing process on the night light data.

[0050] Meanwhile, if the first artificial intelligence model is a CNN-based model, it can be configured to receive nighttime light satellite data as image data, analyze spatial features, and extract light data of fishing vessels.

[0051] Here, a first artificial intelligence model based on CNN can be trained to extract light data of fishing vessels from nighttime light satellite data by utilizing the fact that the shape of the light appears differently depending on the size, type of fishing, and movement characteristics of the fishing vessel.

[0052] For example, small fishing vessels can utilize small-sized forms, large fishing vessels can utilize forms with long, extended lights or multiple lights appearing continuously, and cluster fishing can utilize forms where lights form clusters at regular intervals.

[0053] In addition, the first AI model based on CNN may be trained to extract the lights of fishing vessels from nighttime light satellite data by utilizing the fact that fishing vessel patterns appear differently depending on the fishing method.

[0054] For example, since jigging vessels use strong lights to attract plankton, the intensity of the light appears stronger compared to other vessels in the same area; and since they stay in a certain area for a certain period of time, patterns in which the light is continuously observed at a specific location can be utilized, as can patterns in which lights are arranged to form clusters in circles or at regular intervals for school fishing.

[0055] In addition, since fishing vessels have the characteristic of constantly changing positions unlike harbor lights or offshore platforms, the first AI model based on CNN can learn the characteristics of fishing vessel lights moving in satellite images from multiple viewpoints, thereby distinguishing between fixed light sources such as harbor lights or offshore platforms and the lights of moving fishing vessels.

[0056] As such, the first artificial intelligence model based on CNN can be said to be useful for extracting fishing vessel light data based on the shape of the fishing vessel lights, fishing vessel operation patterns, and movement characteristics of the fishing vessel.

[0057] In addition, for generating a night fishing boat distribution map of the distribution map generation unit (120) to be described later, the extracted fishing boat light data can be configured to include not only simple intensity information but also location information (latitude, longitude, etc.) where the light was detected.

[0058] Accordingly, the extraction unit (110) includes the coordinate values ​​of the actual sea area where the fishing vessel is located in the light data of each extracted fishing vessel.

[0059] Next, a distribution map generating unit (120) according to one embodiment of the present invention generates a night fishing boat distribution map based on light data identified as fishing boats (see step S220).

[0060] For example, a night fishing vessel distribution map may be a map generated to visually represent the distribution of fishing vessels in a specific sea area based on the location data of fishing vessels from the light data of fishing vessels extracted from night light satellite data.

[0061] First, the distribution map generation unit (120) applies a gap filling technique to the extracted light data of the fishing boat to correct missing values.

[0062] Here, missing values ​​may refer to data where light data of some fishing vessels is missing from satellite imagery due to various causes such as clouds, atmospheric interference, and reduced sensor sensitivity.

[0063] In addition, gap filling is a process of supplementing missing parts in satellite data, and can refer to a technique that corrects missing values ​​by utilizing the movement patterns of fishing vessels and surrounding data.

[0064] For example, gap filling includes correction techniques based on surrounding data, correction techniques based on time series data, and correction techniques based on clustering, and the distribution map generation unit (120) can correct missing values ​​based on at least one of the gap filling techniques described above.

[0065] First, the correction technique based on surrounding data is a method that corrects missing values ​​by predicting the location of missing fishing vessels using fishing vessel light data detected in adjacent waters. It can predict the location of missing fishing vessels and correct missing values ​​by using spatial interpolation techniques such as Kriging Interpolation and IDW (Inverse Distance Weighting).

[0066] Alternatively, a correction technique based on time-series data is a method that corrects missing fishing vessel locations by comparing past and present data; for instance, if a fishing vessel repeatedly visits a specific fishing area, missing values ​​can be corrected by predicting the fishing vessel operating at that location based on past data.

[0067] Alternatively, a clustering-based correction technique is a method for correcting missing fishing vessels by analyzing the pattern of fishing vessels forming clusters during operation. If there is a pattern of fishing vessels operating at regular intervals, missing values ​​can be corrected to maintain that interval when missing data occurs, and this can be performed based on algorithms such as DBSCAN (Density-Based Spatial Clustering of Applications with Noise).

[0068] The distribution map generation unit (120) generates a night fishing boat distribution map based on location data from the fishing boat light data with missing values ​​corrected.

[0069] For example, the distribution map generation unit (120) can generate a night fishing boat distribution map by mapping the location of individual fishing boats on the map based on location data and displaying them in the form of points.

[0070] Additionally, the distribution map generation unit (120) can analyze the density of fishing vessels operating in a specific sea area and represent areas where fishing is concentrated and areas where fishing is relatively less frequent on a night fishing vessel distribution map using color differences.

[0071] For example, sea areas with a high density of fishing vessels can be represented in relatively darker colors, allowing users to visually distinguish areas where specific fishing grounds are formed.

[0072] In addition, the distribution map generation unit (120) may distinguish and display fishing boats on the night fishing boat distribution map according to the fishing method.

[0073] As such, the night fishing vessel distribution map generated by the distribution map generation unit (120) can be generated by comprehensively analyzing the location of individual fishing vessels, fishing vessel density, fishing method, etc., and can more accurately monitor fishing activities in a specific sea area.

[0074] The prediction unit (130) predicts the distribution of night fishing boats after the reference point based on the night fishing boat distribution map and light data identified as fishing boats before the reference point (see step S230).

[0075] That is, the prediction unit (130) analyzes a night fishing boat distribution map generated based on fishing boat light data collected before the reference point, learns the pattern of fishing boats moving or fishing in a specific sea area, and can predict the distribution of fishing boats after the reference point.

[0076] In addition, the prediction unit (130) can predict where fishing vessels are likely to fish in the future by taking into account the tendency of fishing patterns in a specific sea area to change according to seasonal and environmental factors in past data (night fishing vessel distribution map prior to the reference point).

[0077] To this end, the prediction unit (130) inputs the night fishing boat distribution map and light data identified as fishing boats prior to the reference point into the second artificial intelligence model to predict the night fishing boat distribution after the reference point.

[0078] At this time, the second artificial intelligence model may be a deep learning-based model trained to predict the distribution of night fishing boats from a learning map of night fishing boat distribution and past night light satellite data.

[0079] For example, the second artificial intelligence model may be a model based on a Long-Term Memory Network (LSTM) or a Transformer, and can learn the movement paths and fishing patterns of fishing vessels to predict the distribution of fishing vessels after a reference point in time.

[0080] If the second artificial intelligence model is a model based on LSTM, it uses the fishing boat distribution map and fishing boat light data prior to the reference point as input values ​​to learn time series fluctuation patterns and predict the movement path and distribution of fishing boats after the reference point.

[0081] For example, a second artificial intelligence model based on LSTM can learn the fishing patterns of fishing vessels that repeatedly appear in a specific sea area over a certain period of time in the past, and predict the distribution of fishing vessels expected in the same sea area in the future.

[0082] In addition, the second artificial intelligence model based on LSTM enables the calculation of the expected movement path of fishing vessels in a specific time period after the reference point, based on the movement pattern, when it is confirmed that fishing vessels with a specific fishing pattern tend to move along a certain path in the past time prior to the reference point.

[0083] As such, it can be seen that when the second artificial intelligence model is a model based on LSTM, it can perform more precise predictions by analyzing the temporal variability of the night fishing vessel distribution and reflecting the periodically repeating fishing patterns.

[0084] Meanwhile, if the second artificial intelligence model is a model based on a Transformer, it can be configured to receive fishing vessel distribution maps and fishing vessel light data prior to a reference point, learn the movement patterns of fishing vessels at specific time intervals, and analyze long-term dependencies within the time series data by utilizing an attention mechanism.

[0085] For example, a second AI model based on Transformers learns the fishing methods of fishing vessels moving in a certain pattern in a specific sea area, and by reflecting past fishing patterns and environmental factors, it enables the prediction of areas with high concentrations of fishing vessels at a specific point in time after a reference point.

[0086] In addition, it may be configured to utilize the Multi-Head Attention technique to simultaneously analyze the movement patterns of multiple fishing vessels operating in different sea areas and to make predictions by considering the interactions between the vessels.

[0087] As such, the second AI model based on Transformers can perform predictions by comprehensively considering the movement patterns of fishing vessels and environmental variables, and can have an advantage over existing LSTM-based models in performing long-term predictions.

[0089] As described above, according to an embodiment of the present invention, a fishing vessel distribution prediction device analyzes nighttime light data captured by a satellite to extract light data identified as fishing vessels, generates a nighttime fishing vessel distribution map based on this, and enables the prediction of future fishing vessel distribution by utilizing time series data.

[0090] In addition, a gap filling technique is applied to correct missing values ​​in fishing vessel light data to obtain more reliable data, and based on the corrected data, the density of fishing vessels in a specific sea area is analyzed and an accurate night fishing vessel distribution map is generated by reflecting fishing patterns.

[0091] Furthermore, the artificial intelligence-based prediction unit (130), which has learned past fishing vessel distribution data and fishing vessel light data, analyzes the fishing vessel movement path and fishing pattern to predict the expected distribution of fishing vessels in a specific sea area in the future.

[0092] As such, it can be seen that the fishing vessel distribution prediction device according to one embodiment of the present invention can be utilized in fields such as fishing optimization, marine resource management, and illegal fishing monitoring by monitoring changes in the distribution of fishing vessels in real time and predicting future fishing vessel concentration areas, and can contribute to improving fishing productivity and strengthening marine safety.

[0094] Meanwhile, the functional operations and embodiments of the subject described herein may be implemented as digital electronic circuits, or as computer software, firmware, or hardware comprising the structures and structural equivalents disclosed herein, or as a combination of one or more of these. The embodiments of the subject described herein may be implemented as one or more computer program products, that is, as one or more modules relating to computer program instructions encoded on a tangible program storage medium for processing the operation of a processing system or for execution by means of such processing.

[0095] A computer-readable medium may be a machine-readable storage device, a machine-readable storage substrate, a memory device, or a combination of one or more of these.

[0096] In this specification, the terms “system” or “device” encompass all apparatuses, devices, and machines for processing data, including, for example, programmable processors, computers, or multiple processors or computers. A processing system may, in addition to hardware, include code that forms an execution environment for a computer program upon request, such as, for example, code constituting processor firmware, a protocol stack, a database management system, an operating system, or a combination of one or more of these.

[0097] A computer program (also known as a program, software, software application, script, or code) may be written in any form of a programming language, including compiled or interpreted languages, or a priori or procedural languages, and may be deployed in any form, including a standalone program, a module, a component, a subroutine, or other unit suitable for use in a computer environment. A computer program does not necessarily correspond to a file in a file system. A program may be stored in a single file provided to the requested program, in multiple interacting files (e.g., a file storing one or more modules, subprograms, or parts of code), or in a part of a file containing other programs or data (e.g., one or more scripts stored within a markup language document). A computer program may be deployed to be executed on multiple computers or a single computer located at a single site or distributed across multiple sites and interconnected by a communication network.

[0098] Meanwhile, computer-readable media suitable for storing computer program instructions and data may include all forms of non-volatile memory, media, and memory devices, such as semiconductor memory devices like EPROM, EEPROM, and flash memory devices, magnetic disks like internal hard disks or external disks, magneto-optical disks, and CD-ROM and DVD-ROM disks. Processors and memory may be supplemented by or integrated with special-purpose logic circuits.

[0099] In the specification of this disclosure (particularly in the claims), the use of the term "above" and similar descriptive terms may be in both singular and plural. Furthermore, where a range is described in this disclosure, it is to include an invention to which individual values ​​belonging to said range are applied (unless otherwise stated), as is equivalent to describing each individual value constituting said range in the detailed description of the invention.

[0100] Unless explicitly stated otherwise, the steps constituting the method according to the present disclosure may be performed in a suitable order. The present disclosure is not necessarily limited by the order in which the steps are described. The use of all examples or exemplary terms (e.g., etc.) in the present disclosure is merely for the purpose of describing the present disclosure in detail, and the scope of the present disclosure is not limited by such examples or exemplary terms unless limited by the claims. Furthermore, a person skilled in the art will understand that various modifications, combinations, and changes may be made according to design conditions and factors within the scope of the claims or equivalents to which they are added.

[0101] Accordingly, the scope of the present disclosure should not be limited to the embodiments described above, and all scopes equivalent to or equivalently modified from the claims set forth below, as well as the claims set forth below, shall be considered to fall within the scope of the scope of the present disclosure.

Claims

Claim 1 A fishing vessel distribution prediction device utilizing nighttime light satellite data, comprising: an extraction unit that extracts light data identified as fishing vessels from nighttime light satellite data at a reference point transmitted from a satellite; a distribution map generation unit that generates a nighttime fishing vessel distribution map based on the light data identified as fishing vessels; and a prediction unit that predicts the nighttime fishing vessel distribution after the reference point based on the nighttime fishing vessel distribution map and the light data identified as fishing vessels before the reference point. Claim 2 In claim 1, the extraction unit inputs nighttime light satellite data at the reference time into a first artificial intelligence model and extracts light data identified as fishing vessels, a fishing vessel distribution prediction device utilizing nighttime light satellite data. Claim 3 In paragraph 2, the first artificial intelligence model is a deep learning-based model trained to identify only the light data of fishing vessels from the learning night light satellite data, a fishing vessel distribution prediction device utilizing night light satellite data. Claim 4 A fishing vessel distribution prediction device utilizing night light satellite data, wherein the prediction unit inputs the night fishing vessel distribution map and light data identified as fishing vessels prior to the reference point into a second artificial intelligence model to predict the night fishing vessel distribution after the reference point. Claim 5 In paragraph 4, the second artificial intelligence model is a deep learning-based model trained to predict the distribution of night fishing vessels from a learning night fishing vessel distribution map and past night light satellite data, a fishing vessel distribution prediction device utilizing night light satellite data. Claim 6 In claim 1, the distribution map generating unit corrects missing values ​​by applying a gap filling technique to the extracted light data of the fishing vessel, a fishing vessel distribution prediction device utilizing night light satellite data. Claim 7 In claim 6, the distribution map generating unit generates the night fishing boat distribution map based on location data from the fishing boat light data in which the missing values ​​are corrected, a fishing boat distribution prediction device utilizing night light satellite data. Claim 8 A method for predicting the distribution of fishing vessels using nighttime light satellite data, comprising: a step of extracting light data identified as fishing vessels from nighttime light satellite data at a reference point transmitted from a satellite; a step of generating a nighttime fishing vessel distribution map based on the light data identified as fishing vessels; and a step of predicting the nighttime fishing vessel distribution after the reference point based on the nighttime fishing vessel distribution map and the light data identified as fishing vessels before the reference point. Claim 9 In claim 8, the extraction step comprises the process of inputting nighttime light satellite data at the reference time into a first artificial intelligence model to extract light data identified as fishing vessels, a method for predicting the distribution of fishing vessels using nighttime light satellite data. Claim 10 In claim 9, the above-mentioned first artificial intelligence model is a deep learning-based model trained to identify only the light data of fishing vessels in the learning night light satellite data, a method for predicting the distribution of fishing vessels using night light satellite data. Claim 11 A method for predicting fishing vessel distribution using night light satellite data, wherein the predicting step comprises inputting the night fishing vessel distribution map and light data identified as fishing vessels prior to the reference point into a second artificial intelligence model to predict the night fishing vessel distribution after the reference point. Claim 12 In claim 11, the above-mentioned second artificial intelligence model is a deep learning-based model trained to predict the distribution of night fishing vessels from a learning night fishing vessel distribution map and past night light satellite data, a method for predicting the distribution of fishing vessels using night light satellite data. Claim 13 A method for predicting the distribution of fishing vessels using nighttime light satellite data, wherein the distribution map generation step comprises a process of correcting missing values ​​by applying a gap filling technique to the extracted light data of fishing vessels. Claim 14 In claim 13, the distribution map generation step further comprises the process of generating the night fishing vessel distribution map based on location data from the fishing vessel light data in which the missing values ​​have been corrected, a method for predicting fishing vessel distribution using night light satellite data. Claim 15 A fishing vessel distribution prediction device utilizing nighttime light satellite data comprises: a memory; and at least one processor, wherein the at least one processor extracts light data identified as fishing vessels from nighttime light satellite data at a reference point transmitted from a satellite, generates a nighttime fishing vessel distribution map based on the light data identified as fishing vessels, and predicts the nighttime fishing vessel distribution after the reference point based on the nighttime fishing vessel distribution map and the light data identified as fishing vessels before the reference point.

Images