Method and apparatus for generating spatiotemporal maps of estimated vessel traffic within an area - Patents.com

Abstract

The present invention relates to a computing device or method for generating a spatio-temporal map (50) of estimated vessel traffic within an area (51), the method comprising the steps of: acquiring (110) a plurality of satellite radar images of the area (51), where the plurality of satellite radar images are generated over a period of twelve months or more with a median frequency of less than ten per week; analyzing (120) generating vessel data representative of historical vessel traffic by analysing the plurality of satellite radar images; modifying (130) the generated vessel data based on the frequency of the plurality of satellite radar images; and generating (180) one or more spatio-temporal vessel distribution maps representative of estimated vessel traffic within the area from the modified data, the spatio-temporal distribution maps including vessel density values ​​for each time period.

Description

[Technical Field]

[0001] The present invention relates to the field of marine traffic, and in particular to the field of vessel traffic analysis. The present invention provides a method and a computing device for generating a spatiotemporal map of estimated vessel traffic (vessel traffic volume estimation) within an area (region). [Background technology]

[0002] Maritime travel, fishing, and the transport of goods are increasing worldwide, which has led to a growing interest in monitoring the marine environment and human activities at sea, including, for example, tracking and monitoring illegal vessel activity, shipping routes, and monitoring vessel movements and oil spills.

[0003] Monitoring vessel activity has been of particular interest, and various systems have been developed to identify vessel presence and distinguish vessel activity. One example is the automatic identification system (AIS), which intelligently utilizes identification data and information transmitted by vessels equipped with the system. Terrestrial AIS data is received by ground stations, while space-based AIS uses satellites to collect signals from vessels. MarineTraffic® is a leading provider of AIS-based vessel tracking. This vessel tracking is primarily based on data collected from a network of coastal AIS receiving stations, supplemented by satellite receivers. MarineTraffic® uses AIS technology to receive, analyze, and store the positions of a vast number of vessels daily. However, the resulting data lacks information on vessels not equipped with AIS.

[0004] The use of space-based imagery for maritime monitoring has also been proposed. Satellite-based radar imagery, typically collected by synthetic aperture radar (SAR), is becoming increasingly widespread for maritime monitoring because of its ability to acquire data in all weather conditions, day or night. SAR's microwave-active sensors illuminate targets with a focused, directed beam of energy, which produces unique scattering effects depending on the orientation of the detected object. The backscattering response of surface materials to microwave energy (also called the "backscatter signal") differs significantly from the spectral reflectance of the same material in visible sunlight. Therefore, SAR systems can provide unique information over a wide area, regardless of weather or other conditions.

[0005] In the field of Earth observation, SAR has become a valuable tool for security and environmental services, where vessel detection and identification based on SAR data is an important element for services and systems dealing with, for example, maritime traffic (maritime traffic), illegal fishing, or maritime border activity, or marine coastal management such as oil spill detection and monitoring, in real time or near real time.

[0006] One of the main challenges in ship detection is the presence of sea clutter inherent in coherent images. Ship detection algorithms, such as constant false alarm rate (CFAR), α-stable algorithms, and wavelet transforms, have been implemented. However, these algorithms have limitations due to the presence of sea clutter and speckle noise in SAR images. Therefore, in recent years, methods have been proposed to integrate synthetic aperture radar (SAR) images with data obtained from an Automatic Identification System (AIS) to effectively monitor maritime activities. In particular, one method estimates the false alarm and missed detection rates by confirming the number of ships detected by SAR using AIS as a ground truth data source (Graziano et al. Integration of Automatic Identification System (AIS) Data and Single-Channel Synthetic Aperture Radar (SAR) Images by SAR-Based Ship Velocity Estimation for Maritime Situational Awareness. Remote Sens. 2019, 11(19), 2196). The proposed method limits the distance between AIS reporting and SAR detection to less than 150 meters.

[0007] However, vessel detection by AIS still presents challenges, with many vessels of various sizes and types either not having AIS or having their AIS turned off.

[0008] Furthermore, while these methods may seem suitable for real-time or immediate analysis of vessel traffic, they are not suited to studies assessing environmental and social impacts. When planning activities in marine areas, especially coastal areas, it is tempting to obtain general information about vessel traffic so that activities can be optimally planned. In fact, maritime traffic, including fishing and tourism, has a significant impact on offshore activities such as seismic exploration, drilling, and the installation of offshore infrastructure. Fishing vessels in particular are often problematic due to their lack of location information (they do not have AIS), the shifting fishing grounds throughout the year, and the seasonality of their activities.

[0009] Therefore, in order to fully understand the changes in vessel traffic in an area over time, a solution is needed that can generate a spatiotemporal map of estimated vessel traffic in that area. Summary of the Invention

[0010] The following is a simplified summary of selected aspects, embodiments, and examples of the present invention to facilitate a basic understanding of the present invention. However, this summary does not constitute an exhaustive list of all aspects, embodiments, and examples of the present invention. It should be noted that the summary merely provides selected aspects, embodiments, and examples of the present invention as a concise introduction to the detailed description of the aspects, embodiments, and examples of the present invention.

[0011] The present invention aims to overcome the drawbacks of the prior art. In particular, the present invention provides a method implemented by one or more processors for generating a spatiotemporal map of estimated vessel traffic in an area, preferably an offshore area, the method comprising: acquiring a plurality of satellite radar images of the area, wherein the plurality of satellite radar images are generated over a period of twelve months or more at a median frequency of less than ten per week and have a resolution in the range of five meters to twenty meters; an analysis step of generating vessel data representing vessel traffic history including vessels not equipped with an automatic identification system by analyzing the plurality of satellite radar images; verifying or correcting the generated vessel data based on the frequency of the plurality of satellite radar images; generating one or more spatial and temporal distribution maps of vessels representing estimated vessel traffic within the area from the verified or corrected vessel data, the spatial and temporal distribution maps including vessel density values ​​for each time period; Includes:

[0012] The advantage of this methodology is that it provides a complete statistical picture of maritime traffic changes in a given area over time. All vessels (including those without AIS) of a size that fit within the satellite's spatial resolution can be detected. This means that a spatiotemporal map can be generated that shows the estimated vessel traffic in an area. Such a map can be used to generate forecasts of vessel traffic depending on the period of interest and other parameters. Estimating traffic in this way makes it possible to quickly identify how traffic in an area changes depending on various parameters such as vessel size and period of time.

[0013] According to other optional features of the method according to the invention, the method may include one or more of the following characteristics, either alone or in combination: - The generated vessel data includes the vessel position, the estimated vessel length, and time information. The generated vessel data representing vessel traffic includes vessels with an overall length of less than 100 meters, preferably less than 50 meters. While AIS detection can only target vessels that have enabled AIS and have a sufficient overall length (e.g., 100 meters or more), the present invention can also estimate traffic volume for small vessels (e.g., vessels with an overall length of less than 100 meters). - further comprising a step of detecting outliers, preferably the step of detecting outliers may comprise comparing a plurality of satellite radar images of a selected sub-area at a certain point in time and detecting an outlier when a vessel of the same length is detected in the same position at a certain point in time. - Acquire at least 100 satellite radar images of the same area at a frequency of at least two, preferably at least five per month. - further comprising a data refinement step using one or more data selected from Automatic Identification System data, nighttime satellite imagery, and high resolution satellite imagery from said area. - integrating data into a geographic information system related to said area. - defining a plurality of layers associated with one or more portions of the area, wherein the plurality of layers correspond to a plurality of layer versions, and each of the plurality of layer versions includes data associated with a different value of vessel length and / or time. the plurality of layers include one or more attributes associated with the type of information contained in each of the plurality of layers, and optionally, the one or more attributes include a confidence attribute, a value of the confidence attribute corresponding to a level of confidence in the accuracy of the data contained in that layer, and further optionally, the confidence attribute may be based at least in part on the number of satellite radar images used to generate that layer of the plurality of layers, the source of the satellite radar images, or the method used to analyze the satellite radar images. - the plurality of layers includes one or more layers of automatic vessel detection results, AIS layer, nighttime optical satellite imagery, high-resolution satellite imagery, offshore infrastructure, and other associated data according to geographic area in a geographic information system project. - the plurality of layers includes a local coordinate system layer, the local coordinate system layer containing information associated with a coordinate system used by a vessel to determine its local position and to convert the local position of the vessel into a distance between the vessel and a point of interest within the area. - the spatiotemporal distribution map comprises multiple layers, one of which is dedicated to vessel density over time; - further comprising a step of generating a marine traffic hazard score, wherein the step of generating the marine traffic hazard score comprises: determining a geographic portion of a survey area from global positioning satellite (GPS) information; calculating a marine traffic hazard score based on seasonal variations in vessel frequency and vessel length for said geographical portion; Includes:

[0014] In another aspect, the invention may relate to a method for determining a suitable route, the method including using a spatiotemporal map generated in accordance with the invention.

[0015] The present invention may also relate to a method of defining suitable offshore development areas, the method including the step of using a spatiotemporal map generated in accordance with the present invention. Such a method of defining suitable offshore development areas may include the step of defining drilling areas.

[0016] In another aspect, the present invention may relate to one or more tangible, non-transitory computer-readable media that store computer-readable instructions that, when executed by one or more processors, cause the one or more processors to perform the methods of the present invention.

[0017] In another aspect, the present invention also provides a computing device including one or more processors and one or more tangible, non-transitory computer-readable media storing instructions that, when executed by the one or more processors, configure the one or more processors to perform operations including: acquiring a plurality of satellite radar images of an area, wherein the plurality of satellite radar images are generated over a period of 12 months or more at a median frequency of less than 10 images per week and have a resolution in the range of 5 meters to 20 meters; An analysis step of generating vessel data representing vessel traffic history, including vessels not equipped with an Automatic Identification System, by analyzing the plurality of satellite radar images. verifying or correcting the generated vessel data based on the frequency of the plurality of satellite radar images. A generating step of generating one or more spatiotemporal distribution maps of vessels representing estimated vessel traffic within the area from the verified or corrected vessel data, the spatiotemporal distribution maps including vessel density values ​​for each time period. [Brief explanation of the drawings]

[0018] These and other objects, features, and advantages of the present invention will become more apparent from the following detailed description taken in conjunction with the accompanying drawings. [Figure 1] FIG. 1 is a schematic diagram illustrating a method for generating a spatiotemporal map of estimated vessel traffic in an area according to the present invention. [Figure 2] A spatiotemporal map of the area is shown. [Figure 3] This shows a satellite radar image. The white dots in A correspond to ships with large backscattering responses. B is the same radar image as A, with the results of automatic ship extraction (black dots) superimposed. [Figure 4] A spatiotemporal map of the area is shown with vessel locations detected from satellite radar imagery overlaid. [Figure 5] A, B, C, and D all show density maps for an example of an output cell of 1000 m and a search radius of 25 km. [Figure 6] 1 is a diagram of a computing device according to the present invention; [Figure 7] This shows the spatial distribution of detected vessels by month over a five-year period from April 2015 to April 2020. DETAILED DESCRIPTION OF THE INVENTION

[0019] Aspects of the present invention are disclosed with reference to flowcharts and / or block diagrams that illustrate methods, apparatus, and computer program products according to embodiments of the invention.

[0020] The drawings illustrate, via flow diagrams and / or block diagrams, the architecture, functionality, and possible apparatus or systems or methods and computer program products according to embodiments of the present invention.

[0021] To this end, boxes shown in flow diagrams or block diagrams may represent systems, devices, modules, or code containing executable instructions for implementing particular logical function(s).

[0022] In some aspects, the functions associated with the boxes may appear in a different order than shown in the drawings.

[0023] For example, if two boxes are shown side by side, they may execute substantially synchronously, or the boxes may execute in reverse order depending on the functionality involved.

[0024] Each box in a flow diagram or block diagram, and combinations of boxes in a flow diagram or block diagram, may be implemented by a particular system that performs a particular function or operation or that implements a particular combination of equipment and computer instructions.

[0025] A description of exemplary embodiments of the present invention follows.

[0026] As used herein, the term "vessels" may be considered to refer to any object that floats and can be steered / moved. This term may refer not only to ships, but also to boats, platforms, and barges.

[0027] Within the meaning of this invention, the terms "process," "compute," "determine," "display," "extract," "compare," or more broadly, "executable operation" refer to actions that can be performed by a computing device or processor, unless the context requires otherwise. These actions relate to operations and / or methods of a data processing system (such as a computing system or electronic computing device) that can store, transmit, or display information by manipulating and transforming data represented by physical (electronic) quantities in memory within the computing system or other device. In particular, a computational operation can be performed by the device's processor, resulting data can be entered into corresponding fields in a data memory, and the field or fields can be returned to a user, for example, via an HMI (human-machine interface) that formats the data. These actions can be application or software-based.

[0028] The words or phrases "application," "software," "program code," and "executable code" mean any expression, code, or notation of a set of instructions that is designed to cause data processing to perform a particular function, either directly or indirectly (e.g., after translation into other code). Examples of program code include, but are not limited to, subprograms, functions, executable applications, source code, object code, libraries, and / or other sequences of instructions designed to be executed on a computing system.

[0029] By "processor" in the sense of the present invention is meant one or more hardware circuits configured to perform operations according to instructions stored in a code. The hardware circuits may be integrated circuits. Examples of processors include, but are not limited to, central processing units (CPUs), graphic processors, application specific integrated circuits ("ASICs" in Anglo-Saxon terminology), programmable logic circuits, etc. A single processor or multiple other units may be used to implement the present invention.

[0030] In the sense of the present invention, "coupled" means directly or indirectly connected with one or more intermediate elements. When two elements are coupled, they may be mechanically or electrically coupled, or may be connected by a communication channel.

[0031] Within the meaning of the present invention, the expression "human-machine interface (HMI)" corresponds to any element that allows a human to communicate with a computer. This includes, but is not limited to, a keyboard and a means for displaying commands entered via the keyboard and, if desired, allowing the selection of items displayed on the screen via a mouse or touchpad. Another embodiment may be a touchscreen for directly selecting elements on the screen by touching it with a finger or an object, and may optionally display a virtual keyboard.

[0032] By "computing device" is understood any device that includes a processing unit or processor, which may for example be in the form of a microcontroller cooperating with a data memory (which may be a program memory). Such memory may be separable. The processing unit may cooperate with said memory by means of an internal communication bus.

[0033] As used herein, the term "about" allows for variation in value or range, for example, within 10%, within 5%, or within 1% of the boundary value of the stated value or range.

[0034] As used herein, "substantially" means the majority or most of the time, and may be, for example, about 50%, 60%, 70%, 80%, 90%, 95%, 96%, 97%, 98%, 99%, 99.5%, 99.9%, 99.99%, or at least about 99.999% or more.

[0035] In the sense of the present invention, a "sub-area" refers to a predetermined part of an area. In particular, an area may be divisible into one or more sub-areas.

[0036] The expression "route within the area" refers to a route (flight) from one point to another. This route may be defined as a line passing through a group of different points, each of which may be characterized by its geographic coordinates. Conversely,

[0037] The term "offshore area" refers to an area of ​​the sea that does not include coastal areas, while the term "coastal area" refers to an area of ​​the sea that does include coastal areas.

[0038] As mentioned above, most of the methods developed so far for maritime traffic analysis are specialized in real-time or immediate analysis of vessel traffic. However, these methods are not suitable for environmental and social impact assessment studies, nor are they suitable for understanding the overall picture of maritime traffic, including vessels without AIS. Therefore, there is a need for a spatiotemporal map of estimated vessel traffic in an area that can be used to forecast vessel traffic depending on the period of interest and other parameters.

[0039] A method based on satellite radar images has been developed. The analysis is carried out from infrequently acquired satellite radar images, which allows for long-term analyses (more than one year) to be carried out with acceptable computation times. The solution of the present invention allows for a rapid survey with vessel size estimation, which has the practical benefit of being able to distinguish between fishing vessels and large tanker traffic.

[0040] The present invention relates to the generation of estimated traffic volumes, which may take the form of maps. Estimating traffic volumes in this manner allows for rapid identification of changes in traffic volumes within an area as a function of various parameters, such as vessel size and time period. Previous methods have not been able to accurately estimate traffic volumes. The solution preferably includes generating vessel density data for each area and time period.

[0041] A first aspect of the present invention relates to a method 100 for generating a spatiotemporal map 50 of estimated vessel traffic. In particular, the map 50 of estimated vessel traffic can be tailored to a particular region of interest 51. As noted above, coastal areas present challenges. Preferably, the region of interest 51 is an offshore or coastal area. The region of interest 51 may include a marine area 53 and a land area 54.

[0042] The method 100 is preferably implemented by one or more processors 10. A processor 10 implementing the method of the present invention may be incorporated into a computing device (such as a computer or computer server) configured to receive a plurality of satellite radar images. Computing devices 1 suitable for implementing the present invention or configured to implement the method of the present invention are described in further detail below.

[0043] The method according to the invention allows to carry out long analyses with acceptable computation times.

[0044] As particularly shown in FIG. 1 , a method 100 for generating a spatiotemporal map 50 of estimated vessel traffic in accordance with the present invention includes the steps of: acquiring 110 a plurality of satellite radar images of an area 51; analyzing 120 the plurality of satellite radar images to generate vessel data representative of vessel traffic history (past vessel traffic); validating or correcting 130 the generated data; and generating 160 one or more spatiotemporal distribution maps of vessels representative of estimated vessel traffic within the area.

[0045] The method may further include the steps of outlier detection 135, data refinement 140, generating a marine traffic obstruction score 170, and / or integrating the data into a geographic information system 150.

[0046] As shown in FIG. 1, the method 100 of the present invention includes the step 110 of acquiring a plurality of satellite radar images, in particular a plurality of satellite radar images of an area 51 .

[0047] As shown in Figure 2, an area 51 may correspond to several satellite radar images 52. Such image tiling allows for the analysis of large areas that would be inaccessible with a single satellite radar image.

[0048] The method of the present invention may include the use of several satellite radar images covering different portions of the area, and thus several satellite radar images covering different points in time within the period of interest. Preferably, 100 or more satellite radar images may be acquired at different times over the same area 51. More preferably, 200 or more satellite radar images 52 may be acquired at different times over the same area 51. Even more preferably, 400 or more satellite radar images 52 may be acquired at different times over the same area 51. Preferably, the satellite radar images are historical data associated with the area of ​​interest 51. For example, the satellite radar images may be generated over a period of 12 months or more, preferably over a period of 24 months or more, more preferably over a period of 36 months or more, and even more preferably over a period of 48 months or more. This scale of historical data allows the present invention to perform seasonal analysis of vessel traffic and, therefore, to evaluate trends in vessel traffic in the area 51.

[0049] This step can be specifically designed to load into memory all the images necessary to implement the present invention.

[0050] Prior art methods have been designed to maximize the accuracy of vessel traffic analysis at the expense of speed and cost. Thus, the present invention preferably acquires and uses a median frequency of fewer than 10 satellite radar images per week. More preferably, the present invention acquires and uses a median frequency of fewer than 8 satellite radar images per week, and even more preferably, fewer than 6 satellite radar images per week.

[0051] However, to accurately estimate vessel traffic within an area, the median frequency of the multiple satellite radar images is preferably greater than or equal to two per month, more preferably greater than or equal to four per month, and even more preferably greater than or equal to six per month.

[0052] The satellite radar images used in the present invention preferably have a resolution of 20 meters or less, and more preferably have a resolution in the range of 5 meters to 20 meters.

[0053] While prior art has been developed to analyze high resolution satellite radar images with high accuracy and in real time, the advantage of the present invention is that it computes multiple satellite radar images to provide a robust estimation of vessels within area 51 over time.

[0054] That is, it is preferable to acquire 100 or more satellite radar images 52 for the same area 51. It is even more preferable to acquire 200 or more satellite radar images 52 for the same area 51. It is even more preferable to acquire 400 or more satellite radar images 52 for the same area 51.

[0055] As shown in Figure 1, the method 100 according to the present invention includes a step 120 of analyzing a number of satellite radar images, which is specifically designed to generate data relating to ships. An advantage of the present invention is that the proposed solution also includes the analysis of ships that are not equipped with an Automatic Identification System.

[0056] The data generated about ships in this way can be thought of as representing the history of ship traffic. However, because satellite radar imagery is not acquired continuously, the generated data only represents a portion of the history of ship traffic. In fact, the lack of raw data used means that much information about real-world ship traffic is not captured in this analysis.

[0057] The data generated about the vessel may include data about the vessel's position and / or data about the vessel's estimated length, which may be associated with time information, which may in particular be the time at which the satellite image was taken.

[0058] Preferably, the data generated regarding the vessel includes the position of the vessel, which will be reported based on a global navigation satellite system, such as the Global Positioning System (GPS), the Global Navigation Satellite System (GLONASS), the BeiDou Navigation Satellite System (BDS), or the Galileo system.

[0059] Preferably, the data generated about the vessel includes an estimated length of the vessel, which can be calculated through image processing.

[0060] As shown in Figures 3A and 3B, in satellite radar images, ships are associated with points of light and diffraction.

[0061] Various methods can be used to generate vessel data from satellite radar imagery. These methods are preferably configured to identify vessels and generate an estimate of their overall length. For example, as shown in Figure 3B, a computer vision algorithm can be used to detect vessels and calculate an estimate of their overall length using a box bounding a bright spot assigned to the vessel.

[0062] Moreover, the data generated for vessels representing vessel traffic may include vessels of an overall length of less than 100 meters, preferably less than 50 meters, and more preferably less than 30 meters. In particular, the data generated for vessels representing vessel traffic may include all vessels of an overall length of 20 meters or greater.

[0063] In some embodiments, computer vision algorithms may be pre-trained or pre-constructed. Currently, machine learning is widely adopted in the field of computer vision. Trained models can be broadly divided into models generated by unsupervised learning methods and models generated by supervised learning methods. Unsupervised learning methods allow for the determination of observation sets without prior knowledge. This allows for the formation of such sets without requiring label values ​​for the input data. In contrast, supervised learning methods link inputs to outputs based on examples of input-output pairs.

[0064] The present invention uses machine learning techniques to create supervised predictive models configured to calculate vessel position and / or estimate vessel length. Among the supervised learning methods, neural networks, classification or regression trees, nearest neighbor searches, and random forests are some of the most robust and efficient machine learning techniques in accordance with the present invention.

[0065] There are several methods available for detecting ships from satellite radar imagery. Ship detection methods using SAR data are usually based on either CFAR detectors or image transforms (typically wavelet transforms). An evaluation of operational detection systems was carried out by the Detection and Classification of Maritime Traffic from Space project; see Harm Greidanus, Benchmarking operational SAR ship detection, IEEE Int'l. Geosci. Remote Sens. Symp. 2004 6, pp. 4215-4218, 2004.

[0066] The CFAR algorithm is widely used for computing adaptive thresholds, and its use generally involves determining a probability density function (PDF) that can adequately describe the statistical characteristics of the background.

[0067] Multi-look SAR images can be used in the method according to the invention. A commonly used PDF for multi-look SAR images is the Gaussian distribution.

[0068] Ship detection models using various beam modes of C-band SAR have also been developed. See Touzi et al., "Optimization of the Degree of Polarization for Enhanced Ship Detection Using Polarimetric RADARSAT-2," in IEEE Transactions on Geoscience and Remote Sensing, vol. 53, no. 10, pp. 5403–5424, October 2015. This model was used to estimate the minimum overall length of a ship based on a comparison of the ship's radar cross-section (RCS) with that of the ocean. However, successful SAR detection of a ship depends on the ship's size and type, prevailing wind conditions, the SAR resolution used, and the field of view. Lin and Khoo present a general concept for estimating ship speed using the concept of azimuth shift due to a known course from the range direction. Jiang et al. also proposed distributions such as the gamma distribution and the K distribution, but these distributions also suffer from the same limitations as the Gaussian distribution. Generally, sea clutter in SAR images always exhibits a steep peak or "heavy-tailed" characteristic, so in many practical applications, the above distribution often fails to describe the heavy-tailed sea clutter.

[0069] Another method relies on an improved version of the Search for Unidentified Maritime Objects (SUMO) detector introduced by the European Community's Joint Research Centre. In this approach, the detection method is applied to small sections (1500 x 400 pixels) of the SAR image. For each tile, the mean intensity and standard deviation are calculated and compared to three different thresholds to determine whether it should be targeted.

[0070] As shown in FIG. 1, the generating method 100 of the present invention includes a step 130 of verifying or correcting the generated data.

[0071] Applicant has demonstrated that data derived from analysis of satellite imagery over a period of more than twelve months should be corrected because the frequency of satellite radar imagery can vary both temporally and spatially.

[0072] The step 130 of verifying or correcting the generated data may include, as a first sub-step, identifying frequency variations in the satellite radar imagery. If no variations exist, the generated data can be verified. If frequency variations are identified, the generated data can be corrected.

[0073] In particular, this step may involve normalizing the generated data according to the frequency of the satellite images, for example the temporal sampling rate of the satellite radar images may be used to correct the estimated number of vessels in the area over time.

[0074] As shown in FIG. 1, the generating method 100 of the present invention may further include a step 135 of detecting outliers.

[0075] While the above analysis steps allow for optimal performance, outlier detection and outlier suppression may be required in some cases.

[0076] The outlier detection step 135 preferably involves comparing multiple satellite radar images of a selected sub-area at several points in time.

[0077] For example, outlier detection can occur when a vessel of the same length is detected in the same position at several times.

[0078] As shown in Figure 1, the generation method 100 of the present invention may further include a data refinement step 140. The refinement may be performed on the generated data, on the verified data, or on the corrected data.

[0079] The present invention is primarily based on satellite radar imagery, however the solution may benefit from other data, such as Automatic Identification System data, satellite infrared night imagery, and / or satellite high resolution imagery.

[0080] For example, data from an Automatic Identification System (AIS) may be used to complement and normalize the generated data.

[0081] Satellite infrared nighttime imagery may also be used to capture low-level radiation sources, both natural and anthropogenic, allowing for the detection of nighttime maritime activity.

[0082] It can also use high-resolution satellite imagery to capture small vessels (8-15 meters) and detect maritime activity at night.

[0083] As shown in FIG. 1, a generating method 100 according to the present invention may include a step 150 of integrating data into a geographic information system.

[0084] An advantage of the present invention is that it can provide information about estimated vessel traffic over an area depending on time period and location (e.g., sub-area).

[0085] As shown in Figure 4, analysis of multiple satellite radar images120 can generate data for each vessel ticketed during the process, including the vessel's location, time, and estimated length. It shows constant vessel traffic throughout the year, concentrated near the coast and continental shelf. Further offshore, maritime traffic appears to be more moderate. This information can be linked to maps and included in a geographic information system.

[0086] The geographic information system data may also include static feature values ​​associated with vessels in the area, such as vessel width.

[0087] The method may further include defining a plurality of layers associated with one or more portions of the zone, the plurality of layers corresponding to layer versions, for example, each layer version including data associated with a different vessel length value and / or a different time value.

[0088] Figures 5A, 5B, 5C, and 5D each show a layer based on a defined density of vessels 56 as a function of time (e.g., Figure 5A shows summer, Figure 5B shows fall, Figure 5C shows winter, and Figure 5D shows spring), and each diagram can be associated with a particular layer assigned to a particular season.

[0089] These multiple layers may relate to one or more of the following layers (preferably layers in a geographic information system): Automatic vessel detection results, AIS layer, nighttime optical satellite imagery, high resolution satellite imagery, offshore infrastructure, and other ancillary data depending on the geographic area.

[0090] The results of the automatic vessel detection may correspond to data generated about vessels to represent historical vessel traffic.

[0091] Advantageously, the plurality of layers includes a local coordinate system layer containing information related to a coordinate system used by a vessel to determine its local position and to convert the vessel's local position into distances between the vessel and points of interest within the area.

[0092] More preferably, the spatiotemporal map includes multiple layers, one of which may be dedicated to the density of vessels 56 over time.

[0093] Layers according to the present invention may include one or more attributes associated with the type of information contained in each layer.

[0094] The one or more attributes may include, for example, a confidence attribute, whose value may correspond to a confidence level in the accuracy of the data contained in the layer, and whose calculation may be based at least in part on the number of satellite radar images used to generate the layer of multiple layers, the source of the satellite radar images, or the method used to analyze the satellite radar images.

[0095] As shown in FIG. 1, the method 100 of the present invention includes generating 160 one or more spatiotemporal distribution maps of vessels 55 representing estimated vessel traffic within an area.

[0096] This step can be specifically designed to allow information generated over a long period of time to be aggregated in a comprehensive manner.

[0097] As shown in Figures 5A, 5B, 5C, and 5D, the generated spatiotemporal distribution map includes vessel density values ​​by period generation 160.

[0098] Preferably, the spatial and temporal distribution map also includes information regarding vessel length by time period and / or by area, and may be configured to allow for the export of vessel data according to a selected time period or a selected area.

[0099] As shown in FIG. 1, a generating method 100 according to the present invention may include generating 170 a marine hazard score.

[0100] An advantage of the present invention is that historical data on vessel traffic can be analyzed to generate useful processed data in the context of marine projects that have low impact on the environment and human activity at sea.

[0101] In particular, the invention may include generating a marine traffic hazard score that reflects a predicted level of marine activity as a function of time. The marine traffic hazard score may be calculated for an area, and more preferably for each sub-area or route within the area.

[0102] In particular, the method may include calculating a marine traffic obstruction score for a geographical portion or sub-area based on seasonality of vessel occurrence and vessel length, which may relate to an estimate of the number of vessels over a period of time.

[0103] The method according to the invention preferably includes the step of calculating the spatial distribution of vessels by month, the number of vessels being preferably normalised by the frequency of images during the observation period, as mentioned above.

[0104] Preferably, the method according to the invention comprises the step of calculating the annual variation, more preferably the method according to the invention may comprise the step of calculating the trend regarding the evolution of the traffic vessels over a period of time.

[0105] The solution according to the present invention may be useful in many coastal or offshore developments. In particular, the solution may be useful in environmental and social impact studies and in any offshore operations. In fact, the generated spatiotemporal maps provide a much more accurate estimation of vessel traffic than those obtained via AIS.

[0106] The use of such a solution can also be applied to the definition of suitable shipping routes and suitable marine development areas.

[0107] Thus, in another aspect, the present invention relates to a method for determining a suitable route, the method including the step of using a spatiotemporal map generated in accordance with the present invention.

[0108] The method may include integrating definitions of upstream criteria, such as the duration of a road generation or the length of time that the "road is quiet," into the spatiotemporal map.

[0109] In another aspect, the invention relates to a method for determining suitable marine development areas, the method including using a spatiotemporal map generated in accordance with the invention, which method can be used in determining the location of an offshore wind farm, a liquefied gas extraction plant, or an offshore platform.

[0110] That is, the present invention may relate to a method of defining a suitable offshore development area 300, including the step of defining a drilling area.

[0111] As will be appreciated by one skilled in the art, aspects of the present invention may be embodied as an apparatus, system, method, or computer program product. Accordingly, aspects of the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment (including firmware, resident software, microcode, etc.), or mode of operation. Additionally, aspects of the present invention may take the form of a computer program product embodied in one or more computer-readable medium(s) having computer-readable program code embedded therein.

[0112] Accordingly, in another aspect, the present invention relates to one or more computer-readable media storing computer-readable instructions that, when executed by one or more processors 10, cause the one or more processors to perform the method of the present invention. The computer-readable media is preferably a tangible, non-transitory computer-readable medium.

[0113] For purposes of this disclosure, a computer-readable medium may include any means or collection of means capable of retaining data and / or instructions for a period of time. Computer-readable media may include, for example, but are not limited to, storage media as well as communication media. Such storage media include direct access storage devices (such as hard disk drives and floppy disk drives), sequential access storage devices (such as tape disk drives), compact discs, CD-ROMs, DVDs, RAM, ROM, electrically erasable programmable read-only memory (EEPROM), and / or flash memory. Such communication media also include wire, fiber optics, microwaves, radio waves, and other electromagnetic and / or optical transport means. Computer-readable media may also include any combination of the foregoing.

[0114] In particular, any combination of one or more computer-readable media can be used. In the context of this specification, a computer-readable medium can be any tangible medium that can contain or store a program used by or associated with a system, apparatus, or device that executes instructions. A computer-readable medium can be, for example, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the above. More specific examples (a non-exhaustive list) of computer-readable storage media would include hard disks, random access memory (RAM), etc.

[0115] Computer program code for performing operations according to aspects of the present invention may be written in any combination of one or more programming languages, including object-oriented programming languages ​​(e.g., Java, C++), the programming language "C" or similar programming languages, scripting languages ​​(e.g., Perl), and / or functional languages ​​(e.g., Meta Language). The program code may be executed entirely on the user's computer, partially on the user's computer and partially on a remote computer, or entirely on a computer or a remote server. In the latter case, the remote computer may be connected to the user's computer via any type of network, such as a local area network (LAN) or a wide area network (WAN).

[0116] These computer program instructions can be stored on a computer readable medium that can control a computing device (i.e., a computer, a server, etc.), such that the instructions stored on the computer readable medium provide a computing device configured to carry out the present invention.

[0117] Thus, in another aspect, the present invention relates to a computing device 1 configured to generate a spatiotemporal map 50 of estimated vessel traffic in an area 51, preferably an offshore area.

[0118] In particular, the computing device 1 is configured to be able to carry out the method according to the invention.

[0119] For purposes of this disclosure, a computing device 1 according to the present invention may include any means or collection of means operable to compute, classify, process, transmit, receive, acquire, emit, switch, store, display, reveal, detect, record, reproduce, handle or utilize any type of information, sensitive information or data.

[0120] A computing device may be, for example, a personal computer, a network storage device, or other suitable device and may vary in size, shape, performance, functionality, and price. A computing device may include one or more processing resources, such as random access memory (RAM), a central processing unit (CPU), or hardware or software control logic, ROM, and / or other types of non-volatile memory. Components that may be added to a computing device include one or more disk drives, one or more network ports for communicating with external devices, and various input / output (I / O) devices (e.g., keyboard, mouse, video display, etc.). Additionally, a computing device may have one or more buses that can carry communications between various hardware components.

[0121] FIG. 6 is a block diagram illustrating various hardware components that may be utilized in a computing device 1 according to the present invention.

[0122] 6, the computing device 1 may include one or more memory components 20 configured to store a plurality of satellite radar images of an area 51 and instructions for the processor(s), one or more communication interfaces 30 configured to acquire the plurality of satellite radar images of the area 51, and one or more processors 10 configured to process the plurality of satellite radar images to generate one or more spatiotemporal distribution maps. The computing device 1 may also include one or more user interfaces 40.

[0123] The computing device 1 for generating a spatiotemporal map 50 of estimated vessel traffic within an area 51 may include a memory component 20 .

[0124] The storage component 20 may include any computer-readable medium known in the art, such as volatile memory (e.g., static random access memory (SRAM) or dynamic random access memory (DRAM)) and / or non-volatile memory (e.g., read-only memory, flash memory, hard disk, optical disk, magnetic tape, etc.). The storage component 20 may include instructions, modules, or applications for performing various functions. That is, the storage component 20 may implement routines, programs, or matrix-type data structures. Preferably, the storage component 20 includes a medium readable by a computing system in the form of volatile memory, such as random access memory (RAM) or cache memory. The storage component 20, like the other modules, may be connected to other components of the computing device 1 via, for example, a communication bus and one or more data transfer interfaces.

[0125] The storage component 20 is preferably configured to store multiple satellite radar images, and may be configured to store the models used and / or the data generated.

[0126] Furthermore, the memory component 20 is preferably configured to store instructions capable of carrying out the method of the present invention.

[0127] The computing device 1 may further include a communications interface 30. The communications interface 30 is preferably configured to enable data transmission over one or more communications networks, which may be wired or wireless. The communications interface 30 enables the computing device 1 to communicate with other devices or computing systems, particularly clients. For example, the communications interface 30 may enable the computing device 1 to receive satellite radar images from another computer. Preferably, communication occurs via wireless protocols such as Wi-Fi, 3G, 4G, 5G, and / or Bluetooth®. Such data exchange may take the form of sending and receiving files. For example, the communications interface 30 may be configured to transmit printable files. In particular, the communications interface may be configured to enable communication with remote devices, including clients. A client is generally any hardware and / or software capable of communicating with the computing device 1.

[0128] The communication interface 30 according to the present invention is particularly adapted to allow data exchange with third party devices or systems.

[0129] The communication interface 30 is configured to allow acquisition of multiple satellite radar images.

[0130] A computing device 1 for generating a spatiotemporal map 50 of estimated vessel traffic may include one or more processors 10. The processors 10 are operatively coupled to a memory component 20 and are operable to execute instructions encoded in a program to implement the techniques of the present disclosure, and more particularly, to implement the methods of the present disclosure.

[0131] The coded instructions may be stored in any suitable article of manufacture, such as storage component 20, including one or more tangible, non-transitory computer-readable media that may at least collectively store such instructions or routines. That is, storage component 20 may include a set of instructions that, when executed by processor 10, perform a method of the present invention.

[0132] The storage component 20 may include any number of databases or similar storage media that may be queried by the processor 10 as needed to perform the methods of the present invention. In particular, the processor 10 may be configured to generate vessel data representing historical vessel traffic, including vessels not equipped with an Automatic Identification System, by analyzing a plurality of satellite radar images.

[0133] The processor 10 may also be configured to apply corrections to the generated data based on the frequency of multiple satellite radar images.

[0134] The processor 10 may also be configured to generate, from the corrected data, one or more spatiotemporal vessel distribution maps representing estimated vessel traffic within the area, the spatiotemporal distribution maps including vessel density values ​​for each time period.

[0135] While Figure 6 depicts these various modules or components separately, the present invention contemplates various arrangements, including a single module that combines all of the above-described functions. Similarly, these modules or components may be split across several electronic boards or integrated onto a single electronic board.

[0136] The computing device 1 of the present invention may be incorporated into a computing system to communicate with one or more external devices, such as a keyboard, a pointer device, a display, or any device that allows a user to interact with the device 1.

[0137] The computing device 1 may also be configured to communicate with or through a human machine interface, i.e., in some embodiments of the invention, the computing device 1 may be coupled to a human machine interface (HMI) that allows for the transmission of parameters to the device and, conversely, makes available to the user the values ​​of data measured or calculated by the device.

[0138] Generally, an HMI is communicatively coupled to a processor and may include a user output interface and a user input interface, which may include an interface for audio and visual output and various indicators (e.g., visual, audio, tactile, etc.).

[0139] The user input interface may include a keyboard, mouse, or other navigation module, including a touch screen, touch pad, stylus input interface, and a microphone for inputting audio signals, such as user utterances, data, and commands, that are recognizable by the processor. [Example]

[0140] [Acquisition process] In an exemplary embodiment of the invention, over 900 Sentinel-1 radar images were used to cover a very large area of ​​the ocean surface (170 x 250 km) over a 48-month period. Image coverage is good, with 50-400 images per station (average 228). Spatial coverage is very uneven, with higher coverage of radar data in the coastal and northern parts of the display.

[0141] In the method according to the present invention, approximately 35,000 km 2It allows detection of over 50,000 vessels within a given area. After image registration, land may be masked out. A pre-processing step may be required to obtain more accurately calibrated SAR data. The next step is a detection process, which uses a Constant False Alarm Rate (CFAR). A discrimination process is then used to reject false alarms if target measurements or characterization of the ocean or meteorological phenomenon are available. It detects all vessels over 20 meters in size and generates an estimate of the vessel's overall length.

[0142] [Acquired data] Table 1 below shows the number of images analyzed per year.

[0143] [Table 1]

[0144] Table 1 shows that the distribution of monthly image counts increased over the study period, from 7 or fewer images per month in 2015, which is the beginning of Table 1, to 12 or fewer images per month from March 2017 to June 2018, and finally to 27 or fewer images per month from June 2018 onwards.

[0145] Our method allows us to normalize the monthly distribution of detected vessels by the number of images, which shows that marine traffic volume is seasonal, with vessel traffic decreasing in summer (see Figure 7).

[0146] The size of the detected vessels ranges from 30 meters to 260 meters in estimated overall length. Figure 7 shows that "small" vessels in the 30-50 meter length range and "medium" vessels in the 50-100 meter length range make up the majority of the vessel traffic. AIS-equipped tankers (e.g., 200 meters long) appear to be a very limited part of the maritime traffic in this area. Vessels between 100 meters and 150 meters in size are present along the coast and on the continental shelf.

[0147] Very high vessel densities are detected from November to March, mainly along the west coast and further south, with May to August being the period with the least vessel traffic.

[0148] These maps can be layered by vessel size for further analysis, for example using ArcMap's Kernel density toolbox to calculate density maps with an output cell of 1000 meters and a search radius of 25 km.

[0149] The data generated by the present invention can be used to detect large-scale vessel traffic along coasts and continental shelves. It is useful to aggregate the generated data over a period of months or years.

[0150] In this way, it is possible to identify maritime traffic according to the time horizon and predict future traffic. While current methods aim to generate highly accurate real-time data for vessel traffic assessment, the use of satellite radar imagery provides additional information compared to AIS (Automatic Identification System) data used by many maritime traffic providers, such as MarineTraffic. The spatial and temporal distribution of vessels allows for optimizing operations and reducing HSE risks. Furthermore, the low frequency of imagery used allows for rapid surveys to be carried out to predict vessel density according to estimated length and time horizon.

Claims

1. A method (100) implemented by one or more processors (10) for generating a spatiotemporal map (50) of estimated vessel traffic within an area (51), comprising: acquiring (110) a plurality of satellite radar images (52) of the area (51), wherein the plurality of satellite radar images (52) are generated at a median frequency of less than 10 per week over a period of 12 months or more and have a resolution in the range of 5 meters to 20 meters; an analyzing step (120) for analyzing the plurality of satellite radar images (52) to generate vessel data representing vessel traffic history, including vessels not equipped with an Automatic Identification System; verifying or correcting (130) the generated vessel data based on the frequency of the plurality of satellite radar images; generating (160) one or more spatiotemporal distribution maps (55) of vessels representing estimated vessel traffic within the area from the verified or corrected vessel data, the spatiotemporal distribution maps including vessel density values ​​for each time period; A method comprising:

2. The method of claim 1 , wherein the generated vessel data includes the vessel's position, the vessel's estimated overall length, and time information.

3. 3. The method of claim 1 or 2, wherein the generated vessel data representing vessel traffic includes vessels having an overall length of less than 100 meters.

4. moreover an outlier detection step (135) for detecting outliers from the verified or corrected data obtained from the verification or correction step (130); 3. The method of claim 1 or 2, comprising:

5. The outlier detection step (135) comparing multiple satellite radar images of a selected sub-area at a time; Outlier detection when ships of the same length are detected at the same location at a certain time 5. The method of claim 4, comprising:

6. 3. The method of claim 1, wherein one hundred or more satellite radar images (52) are acquired of the same area at a frequency of two or more per month.

7. moreover a data refinement step (140) using one or more types of data selected from Automatic Identification System data, nighttime satellite imagery, and high-resolution satellite imagery from said area (51); 3. The method of claim 1 or 2, comprising:

8. moreover Integrating (150) data into a geographic information system related to said area (51).

3. The method of claim 1 or 2, comprising:

9. moreover defining a plurality of layers associated with one or more portions of the area, wherein the plurality of layers correspond to a plurality of layer versions, and each of the plurality of layer versions includes data associated with a different value of vessel length and / or time; 2. The method of claim 1, comprising:

10. The plurality of layers includes one or more attributes associated with the type of information included in each of the plurality of layers.

10. The method according to claim 9.

11. The method of claim 10 , wherein the one or more attributes include a confidence attribute, a value of the confidence attribute corresponding to a level of confidence in the accuracy of data contained in the stratum.

12. 12. The method of claim 11, wherein the confidence attribute is based at least in part on the number of satellite radar images (52) used to generate the layer of the plurality of layers, the source of the satellite radar images, or the method used to analyze the satellite radar images.

13. 11. The method of claim 9 or 10, wherein the plurality of layers includes one or more layers of automatic vessel detection results, AIS layer, nighttime optical satellite imagery, high-resolution satellite imagery, offshore infrastructure, and other associated data according to geographic area in a geographic information system project.

14. the plurality of layers includes a local coordinate system layer; The local coordinate system layer includes information associated with a coordinate system used by a vessel to determine its local position and to convert the vessel's local position into a distance between the vessel and points of interest within the area.

11. The method according to claim 9 or 10.

15. 3. The method of claim 1 or 2, wherein the spatiotemporal distribution map comprises multiple layers, one of which is dedicated to vessel density over time (56).

16. moreover generating a marine traffic obstruction score; and generating a marine traffic hazard score includes: determining a geographic portion of a survey area from global positioning satellite information; calculating a marine traffic hazard score based on seasonal variations in vessel frequency and vessel length for said geographical portion; Contains 3. The method according to claim 1 or 2, characterized in that

17. using the spatiotemporal map generated according to claim 1 or 2 How to determine the appropriate route, including:

18. using the spatiotemporal map generated according to claim 1 or 2 Methods for determining appropriate marine development areas, including:

19. Steps for defining the excavation area 20. The method of claim 18, comprising:

20. One or more tangible, non-transitory computer-readable storage media storing computer-readable instructions that, when executed by one or more processors (10), cause the one or more processors to perform the method of claim 1 or 2.

21. A computing device (1), one or more processors (10); one or more tangible, non-transitory, computer-readable storage media (20) storing instructions; Including, The instructions, when executed by the one or more processors, cause the one or more processors to: acquiring a plurality of satellite radar images (52) of an area (51), wherein the plurality of satellite radar images (52) are generated at a median frequency of less than 10 per week over a period of 12 months or more and have a resolution in the range of 5 meters to 20 meters; an analysis step of generating vessel data representing vessel traffic history including vessels not equipped with an Automatic Identification System by analyzing the plurality of satellite radar images (52); verifying or correcting the generated vessel data based on the frequency of the plurality of satellite radar images; generating one or more spatiotemporal distribution maps of vessels representing estimated vessel traffic within the area from the verified or corrected vessel data, the spatiotemporal distribution maps including vessel density values ​​for each time period; configured to perform operations including A computing device characterized by:

Images