Ice region navigation dynamic sea ice risk early warning method and system

Abstract

The application provides an ice area navigation dynamic sea ice risk early warning method and system, the method first acquires the navigation information of a target ship, the first navigation radar image information of a first time point and the second navigation radar image information of a second time point, and calculates the time interval of the two time points, then respectively extracts the interested region of the sea ice echo through a mask matrix, fills the pixels of the interested region of the sea ice echo and the pixels of other regions except the interested region of the sea ice echo with different colors to obtain the binary image of the sea ice echo of the two time points, then calculates the distance, azimuth angle of each piece of sea ice and the target ship, and the latitude and longitude coordinates of the sea ice centroid of the two time points by using a specific calculation method, calculates the sea ice collision probability and the sea ice collision consequence, further calculates the sea ice collision risk index, and performs corresponding different level sea ice risk early warning on the sea ice according to the comparison result of the sea ice collision risk index and the preset risk threshold of different levels.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of polar navigation and ice navigation, and particularly relates to an ice navigation dynamic sea ice risk early warning method and system. BACKGROUND

[0002] In recent years, with the development of the Polar Silk Road, polar navigation has become normalized. In order to ensure the safety of ship ice navigation, near-field sea ice perception and risk early warning technology is crucial.

[0003] Compared with macro sea ice information obtained by shaking image, ice chart or sea ice model, navigation radar can better grasp the near-field sea ice information in real time. In addition, most of the past sea ice risk early warning methods regard sea ice as static navigation obstacles, without considering the relative motion between the ship and the sea ice, and ignoring the difference in the ship's own ice resistance level, resulting in inaccurate risk prediction. At present, in the domestic and foreign navigation technology, it is impossible to perceive and warn the risk of near-field sea ice. SUMMARY

[0004] In order to solve the problems of low accuracy and inability to perceive near-field sea ice in the current near-field sea ice perception and risk early warning technology, the present application provides an ice navigation dynamic sea ice risk early warning method, which extracts sea ice echo information based on the navigation information of the ship and through the navigation radar images of two time points, constructs a dynamic ice navigation scene, and calculates the sea ice collision risk index in the dynamic ice navigation scene by using a specific calculation method, which can visualize the risk of each piece of sea ice relative to the ship, greatly improving the accuracy of near-field sea ice perception. The present application also relates to an ice navigation dynamic sea ice risk early warning system.

[0005] The technical scheme of the present application is as follows:

[0006] An ice navigation dynamic sea ice risk early warning method, characterized in that it comprises the following steps:

[0007] Navigation radar image information acquisition step: acquiring the navigation information of the target ship, the first navigation radar image information of the first time point and the second navigation radar image information of the second time point, and calculating the time interval between the two time points;

[0008] Sea ice echo extraction step: in the first navigation radar image information and the second navigation radar image information, the region of interest showing the sea ice echo is extracted by using a mask matrix, and the pixels of the region of interest of the sea ice echo and the pixels of other regions except the region of interest of the sea ice echo are filled with different colors respectively, to obtain the first binary image and the second binary image of the sea ice echo at the two time points;

[0009] The step of constructing a dynamic ice navigation scene includes the following steps: extracting the pixel coordinates of the center of each piece of sea ice in the first and second binary images, respectively, and extracting the pixel coordinates of the target ship according to the latitude and longitude coordinates of the target ship in the navigation information, calculating the distance and azimuth angle between each piece of sea ice and the target ship according to the pixel coordinates of the center of each piece of sea ice and the pixel coordinates of the target ship, respectively, and calculating the latitude and longitude coordinates of the center of each piece of sea ice at two time points according to the distance, the azimuth angle and the pixel coordinates of the target ship; then calculating the sea ice flow direction and the sea ice flow speed according to the latitude and longitude coordinates of the center of each piece of sea ice at two time points, and constructing a dynamic ice navigation scene.

[0010] The step of calculating the sea ice collision probability includes the following steps: in the dynamic ice navigation scene, the closest encounter distance and the closest encounter time between a piece of sea ice and the target ship are calculated according to the distance and the azimuth angle between the piece of sea ice and the target ship, the change in the azimuth of the piece of sea ice relative to the target ship is calculated according to the azimuth angle between the piece of sea ice and the target ship at two time points, and the closest encounter distance, the closest encounter time, the change in the azimuth and the distance between the piece of sea ice and the target ship are normalized, respectively, to obtain the normalized closest encounter distance, the normalized closest encounter time, the normalized change in the azimuth and the normalized distance between the piece of sea ice and the target ship, and then the sea ice collision probability variable is obtained, the closest encounter distance weight, the closest encounter time weight, the change in the azimuth weight and the distance weight between the piece of sea ice and the target ship are set according to the closest encounter distance, the closest encounter time, the change in the azimuth and the distance between the piece of sea ice and the target ship, and then the sea ice collision probability weight is obtained, and the sea ice collision probability of the piece of sea ice is calculated according to the sea ice collision probability variable and the sea ice collision probability weight.

[0011] The step of calculating the sea ice collision risk index includes the following steps: the image matrix in which the piece of sea ice is located is obtained according to the latitude and longitude coordinates of the center of the piece of sea ice and the minimum circumscribed rectangle of the piece of sea ice, the sea ice area of the piece of sea ice is calculated according to the image matrix in which the piece of sea ice is located, the sea ice area and the ship ice resistance grade in the navigation information are normalized, respectively, to obtain the normalized sea ice area and the normalized ship ice resistance grade, and then the normalized sea ice collision consequence variable is obtained, the sea ice area weight and the ship ice resistance grade weight are set according to the sea ice area and the ship ice resistance grade, and then the sea ice collision consequence weight is obtained, the sea ice collision consequence of the piece of sea ice is calculated according to the sea ice collision consequence variable and the sea ice collision consequence weight, and the sea ice collision risk index of the piece of sea ice is calculated according to the sea ice collision probability and the sea ice collision consequence of the piece of sea ice.

[0012] The step of sea ice risk early warning includes the following steps: the sea ice collision risk index of each piece of sea ice is compared with the preset risk threshold of different levels, respectively, and the sea ice is given a corresponding different level of sea ice risk early warning according to the comparison result.

[0013] Preferably, in the step of acquiring the marine radar image information, the navigation information comprises radar range, distance scale circle interval, ship position, ship heading, ground heading and ground speed.

[0014] Preferably, in the step of constructing the dynamic ice region navigation scene, after the longitude and latitude coordinates of the sea ice mass centers at the two time points are calculated, the sea ice is preliminarily matched; the preliminary matching of the sea ice refers to taking each piece of sea ice in the second binary image as a template image, taking each piece of sea ice in the first binary image as a target image, and respectively calculating the similarity of each piece of sea ice in the template image and the target image by using a template matching algorithm, and matching the sea ice in the two images with a similarity less than a preset similarity threshold; the sea ice flow direction and the sea ice flow speed are calculated according to the longitude and latitude coordinates of the matched sea ice mass centers at the two time points.

[0015] Preferably, in the step of constructing the dynamic ice region navigation scene, after the preliminary matching of the sea ice, the matched sea ice is proofread; the proofreading of the matched sea ice refers to calculating the moving distance of a piece of sea ice within a time interval according to the longitude and latitude coordinates of the sea ice mass center of the piece of sea ice and the time interval, comparing the moving distance of the sea ice with a sea ice moving distance threshold set according to the time interval, retaining the matched sea ice with a moving distance less than the sea ice moving distance threshold, removing the redundant matching results, and then splicing the retained first binary image and the second binary image side by side, connecting the sea ice mass centers of the matched sea ice respectively, obtaining a plurality of matched sea ice line segments, and calculating the average slope of all the line segments, taking the average slope as a reference slope, and deleting the redundant matching according to the reference slope when a piece of sea ice is matched to multiple objects; the sea ice flow direction and the sea ice flow speed are calculated according to the longitude and latitude coordinates of the matched sea ice mass centers at the two time points.

[0016] Preferably, in the step of sea ice risk early warning, the comparison of the sea ice collision risk index of each piece of sea ice with the preset risk threshold of different levels comprises:

[0017] When the sea ice collision risk index value of a piece of sea ice is less than or equal to a first risk threshold, a low-risk green early warning is given to the piece of sea ice; when the sea ice collision risk index value of the piece of sea ice is greater than the first risk threshold and less than or equal to a second risk threshold, a medium-risk yellow early warning is given to the piece of sea ice; and when the sea ice collision risk index value of the piece of sea ice is greater than the second risk threshold, a high-risk red early warning is given to the piece of sea ice.

[0018] The application discloses a dynamic sea ice risk warning system for ice area navigation, and belongs to the field of navigation safety.

[0019] The navigation radar image information acquisition module acquires the navigation information of the target ship, the first navigation radar image information at the first time point and the second navigation radar image information at the second time point, and calculates the time interval between the two time points.

[0020] The sea ice echo extraction module extracts the sea ice echo region of interest in the first navigation radar image information and the second navigation radar image information respectively through a mask matrix, fills the pixels of the sea ice echo region of interest and the pixels of other regions except the sea ice echo region of interest with different colors respectively, and obtains the first binary image and the second binary image of the sea ice echo at the two time points.

[0021] The latitude and longitude coordinate calculation module extracts the pixel coordinates of the center of each piece of sea ice in the first binary image and the second binary image respectively, extracts the pixel coordinates of the target ship according to the latitude and longitude coordinates of the target ship in the navigation information, calculates the distance and the azimuth angle between each piece of sea ice and the target ship according to the pixel coordinates of the center of each piece of sea ice and the pixel coordinates of the target ship respectively, and calculates the latitude and longitude coordinates of the center of each piece of sea ice at the two time points according to the distance, the azimuth angle and the pixel coordinates of the target ship.

[0022] The motion parameter calculation module calculates the sea ice flow direction and the sea ice flow speed according to the latitude and longitude coordinates of the center of each piece of sea ice at the two time points, and further constructs the dynamic ice area navigation scene.

[0023] The sea ice collision probability calculation module calculates the closest encounter distance and the closest encounter time between the target ship and the sea ice according to the distance and the azimuth angle between the target ship and the sea ice in the dynamic ice navigation scene, calculates the azimuth change of the sea ice relative to the target ship according to the azimuth angle between the target ship and the sea ice at two time points, and performs normalization processing on the closest encounter distance, the closest encounter time, the azimuth change, and the distance between the target ship and the sea ice, respectively, to obtain the normalized closest encounter distance, the normalized closest encounter time, the normalized azimuth change, and the normalized distance between the target ship and the sea ice, and then obtain the sea ice collision probability variable. The closest encounter distance weight, the closest encounter time weight, the azimuth change weight, and the distance weight between the target ship and the sea ice are set according to the closest encounter distance, the closest encounter time, the azimuth change, and the distance between the target ship and the sea ice, and then the sea ice collision probability weight is obtained. The sea ice collision probability of the sea ice is calculated according to the sea ice collision probability variable and the sea ice collision probability weight.

[0024] The sea ice collision risk index calculation module obtains the image matrix in which the sea ice is located according to the latitude and longitude coordinates of the sea ice centroid and the minimum circumscribed rectangle of the sea ice, calculates the sea ice area of the sea ice according to the image matrix in which the sea ice is located, performs normalization processing on the sea ice area and the ship ice resistance level in the navigation information respectively to obtain the normalized sea ice area and the normalized ship ice resistance level, and then obtains the normalized sea ice collision consequence variable. The sea ice area weight and the ship ice resistance level weight are set according to the sea ice area and the ship ice resistance level, and then the sea ice collision consequence weight is obtained. The sea ice collision consequence of the sea ice is calculated according to the sea ice collision consequence variable and the sea ice collision consequence weight. The sea ice collision risk index of the sea ice is calculated according to the sea ice collision probability and the sea ice collision consequence of the sea ice.

[0025] The sea ice risk early warning module compares the sea ice collision risk index of each sea ice with the preset risk threshold of different levels respectively, and performs the sea ice risk early warning of different levels according to the comparison result.

[0026] Preferably, the navigation information includes radar range, distance scale circle interval, ship position, ship heading, ground heading, and ground speed.

[0027] Preferably, the module for constructing a dynamic ice region navigation scene further comprises a preliminary matching sea ice sub-module, the longitude and latitude coordinate calculation module is connected to the motion parameter calculation module through the preliminary matching sea ice sub-module, after the longitude and latitude coordinates of the sea ice centroids at two time points are calculated by the longitude and latitude coordinate calculation module, the preliminary matching sea ice sub-module takes each piece of sea ice in the second binary image as a template image and each piece of sea ice in the first binary image as a target image, and the similarity of each piece of sea ice in the template image and the target image is calculated respectively by using a template matching algorithm, and the sea ice with a similarity less than a preset similarity threshold in the two images is matched; and the motion parameter calculation module calculates the sea ice flow direction and the sea ice flow speed according to the longitude and latitude coordinates of the matched sea ice centroids at two time points.

[0028] Preferably, the module for constructing a dynamic ice region navigation scene further comprises a preliminary matching sea ice sub-module, the longitude and latitude coordinate calculation module is connected to the motion parameter calculation module through the preliminary matching sea ice sub-module, after the longitude and latitude coordinates of the sea ice centroids at two time points are calculated by the longitude and latitude coordinate calculation module, the preliminary matching sea ice sub-module takes each piece of sea ice in the second binary image as a template image and each piece of sea ice in the first binary image as a target image, and the similarity of each piece of sea ice in the template image and the target image is calculated respectively by using a template matching algorithm, and the sea ice with a similarity less than a preset similarity threshold in the two images is matched; and the motion parameter calculation module calculates the sea ice flow direction and the sea ice flow speed according to the longitude and latitude coordinates of the matched sea ice centroids at two time points.

[0029] Preferably, in the sea ice risk early warning module, the comparison of the sea ice collision risk index value of each piece of sea ice with the preset risk threshold of different levels comprises:

[0030] When the sea ice collision risk index value of a piece of sea ice is less than or equal to the first risk threshold, a low-risk green early warning is given to the piece of sea ice, when the sea ice collision risk index value of the piece of sea ice is greater than the first risk threshold and less than or equal to the second risk threshold, a medium-risk yellow early warning is given to the piece of sea ice, and when the sea ice collision risk index value of the piece of sea ice is greater than the second risk threshold, a high-risk red early warning is given to the piece of sea ice.

[0031] The present application has the following advantages:

[0032] The application provides an ice area navigation dynamic sea ice risk early warning method, which comprises the following steps: first, obtaining the navigation information of a target ship, the first navigation radar image information at a first time point and the second navigation radar image information at a second time point, and calculating the time interval between the two time points; then, in the first navigation radar image information and the second navigation radar image information, the interested area of sea ice echo is extracted by using a mask matrix, and the pixels of the interested area of sea ice echo and the pixels of other areas except the interested area of sea ice echo are filled with different colors respectively to obtain the first binary image and the second binary image of sea ice echo at the two time points, so as to distinguish sea ice from other areas; then, the pixel coordinates of the centroid of each piece of sea ice in the first binary image and the second binary image are extracted respectively, and the distance and the azimuth angle of each piece of sea ice from the target ship are calculated according to the pixel coordinates and the latitude and longitude coordinates of the target ship in the navigation information; the latitude and longitude coordinates of the centroid of each piece of sea ice at the two time points are calculated according to the distance, the azimuth angle and the latitude and longitude coordinates of the target ship, so as to obtain the relative spatial information of each piece of sea ice from the target ship; the sea ice flow direction and the sea ice flow speed are calculated according to the latitude and longitude coordinates of the centroid of each piece of sea ice at the two time points, and a dynamic ice area navigation scene is constructed; then, in the dynamic ice area navigation scene, the closest encounter distance and the closest encounter time of each piece of sea ice from the target ship are calculated according to the distance and the azimuth angle of each piece of sea ice from the target ship, and then the sea ice collision probability variable and the sea ice collision probability weight are obtained; the sea ice collision probability of each piece of sea ice is calculated according to the sea ice collision probability variable and the sea ice collision probability weight, and the possibility of collision between each piece of sea ice and the target ship is estimated; then, the image matrix of each piece of sea ice is obtained according to the latitude and longitude coordinates of the centroid of each piece of sea ice, the sea ice area of each piece of sea ice is calculated according to the image matrix of each piece of sea ice, and then the sea ice collision consequence variable and the sea ice collision consequence weight are obtained; the sea ice collision consequence of each piece of sea ice is calculated according to the sea ice collision consequence variable and the sea ice collision consequence weight, the sea ice collision risk index of each piece of sea ice is calculated according to the sea ice collision probability and the sea ice collision consequence, and the severity of collision between each piece of sea ice and the target ship is estimated; finally, the sea ice collision risk index value of each piece of sea ice is compared with the first risk threshold and the second risk threshold to perform different levels of sea ice early warning on the sea ice, the relative motion of the sea ice and the target ship is considered, and the anti-ice level of the target ship is considered, the risk of each piece of sea ice relative to the ship is visualized, the accuracy of near-field sea ice perception can be effectively improved, and the problem that the navigation risk is caused by the negligence or subjective misjudgment of the driver on sea ice is solved. The application constructs a dynamic ice area navigation scene based on the navigation radar images at different time points as the sea ice perception method, proposes dynamic sea ice collision risk assessment and early warning according to the motion state of the sea ice and the target ship and the characteristics of the sea ice and the target ship, and provides technical support for ice area ship navigation auxiliary decision-making.

[0033] The application also relates to an ice region navigation dynamic sea ice risk early warning system, which corresponds to the ice region navigation dynamic sea ice risk early warning method, can be understood as a system for realizing the ice region navigation dynamic sea ice risk early warning method, and comprises a navigation radar image information acquisition module, a sea ice echo extraction module, a dynamic ice region navigation scene construction module, a sea ice collision probability calculation module, a sea ice collision risk index calculation module and a sea ice risk early warning module which are sequentially connected, the dynamic ice region navigation scene construction module comprises a longitude and latitude coordinate calculation module and a motion parameter calculation module, each module works cooperatively, sea ice echo information is extracted based on the navigation information of a ship and through two-time-point navigation radar images, a dynamic ice region navigation scene is constructed, and the sea ice collision probability and the sea ice collision consequence are calculated in the dynamic ice region navigation scene by using a specific calculation method, then the sea ice collision risk index is calculated, the risk of each piece of sea ice relative to the ship can be visually early warned, and the accuracy of near-field sea ice perception is greatly improved. BRIEF DESCRIPTION OF DRAWINGS

[0034] Figure 1 is a flow chart of the ice region navigation dynamic sea ice risk early warning method of the application.

[0035] Figure 2 is a sea ice preliminary matching result schematic diagram.

[0036] Figure 3 is a sea ice proofreading matching result schematic diagram.

[0037] Figure 4 is a dynamic sea ice scene construction schematic diagram. DETAILED DESCRIPTION

[0038] The application will be described below in combination with the drawings.

[0039] The application relates to an ice region navigation dynamic sea ice risk early warning method, a flow chart of the method is shown in Figure 1 and comprises the following steps in sequence:

[0040] 1. The marine radar image information acquisition step: acquiring the navigation information of the target ship, the first marine radar image information at the first time point and the second marine radar image information at the second time point, and calculating the time interval between the two time points. Specifically, first, connect the PostgreSQL database using psycopg2 (a PostgreSQL database interface for Python language) in Python language, query the navigation information of the target ship from the PostgreSQL database, then use the marine radar to acquire the first marine radar image information at the first time point t0 and the second marine radar image information at the second time point t1, and calculate the time interval t1-t0 between the two time points. Preferably, the navigation information includes radar range, distance scale circle interval, ship position, ship heading, ground heading and ground speed, etc.

[0041] 2. The sea ice echo extraction step includes extracting the region of interest and extracting the sea ice echo: respectively in the first marine radar image information and the second marine radar image information, the region of interest showing the sea ice echo is extracted by the mask matrix, and the pixels of the region of interest of the sea ice echo and the pixels of other regions except the region of interest of the sea ice echo are filled with different colors respectively, obtaining the first binary image and the second binary image of the sea ice echo at the two time points.

[0042] Specifically, first, the region of interest ROI showing the sea ice echo is extracted in the first marine radar image information and the second marine radar image information respectively by the mask matrix, wherein the mask matrix expression is as follows:

[0043] dst(I)=src1(I)∧src2(I) if mask(I)≠0 (1)

[0044] In the above formula, dst is the output image matrix, I is the multi-dimensional index of the image matrix element, src1 and src2 are two image matrices with the same size, and mask is the mask matrix.

[0045] Then the pixels of the region of interest of the sea ice echo and the pixels of other regions except the region of interest of the sea ice echo are filled with different colors respectively, that is, the pixels of the sea ice echo region are set to white and the other redundant regions are set to black, obtaining the first binary image Img0 and the second binary image Img1 of the sea ice echo at the two time points. Wherein, the range expression of the region of interest of the sea ice echo is as follows:

[0046] dst(I)=lowerb(I)≤src(I)≤upperb(I) (2)

[0047] In the above formula, lowerb is the lower boundary threshold of the region of interest in the sea ice echo, src is the input image matrix, upperb is the upper boundary threshold of the region of interest in the sea ice echo, and I is the multidimensional index of the image matrix element.

[0048] 3. Steps for constructing a dynamic ice zone navigation scenario: Extract the pixel coordinates of the centroid of each sea ice block from the first and second binary images respectively. Extract the pixel coordinates of the target ship based on the latitude and longitude coordinates of the target ship in the navigation information. Calculate the distance and azimuth between each sea ice block and the target ship based on the pixel coordinates of the sea ice centroid and the target ship respectively. Calculate the latitude and longitude coordinates of the sea ice centroid at two time points based on the distance, azimuth, and the target ship's pixel coordinates respectively. Then, calculate the sea ice flow direction and sea ice flow velocity based on the latitude and longitude coordinates of the sea ice centroid at the two time points, thereby constructing a dynamic ice zone navigation scenario.

[0049] 3.1 Calculate the latitude and longitude coordinates of the sea ice mass center.

[0050] Specifically, the pixel coordinates (u) of the centroid of each sea ice block are first extracted from the first binary image Img0 and the second binary image Img1. t ,v t ), and extract the pixel coordinates (u) of the target ship based on the latitude and longitude coordinates of the target ship in the navigation information. s ,v s Based on the pixel coordinates (u) of the sea ice centroid t ,v t ) and the pixel coordinates of the target ship (u s ,v s ) Calculate the distance D and azimuth Br between each piece of sea ice and the target ship, and then use the distance D, azimuth Br, and the pixel coordinates (u) of the target ship. s ,v s ) Calculate the latitude and longitude coordinates (T) of the mass center of each sea ice block. lon ,T lat Among them, the distance D, azimuth Br, and latitude and longitude coordinates (T) of the mass center of each sea ice block. lon ,T lat Calculate according to the following formulas respectively:

[0051]

[0052]

[0053]

[0054]

[0055] In the above formula, D is the pixel distance from the ship point to the sea ice mass center, (u s ,vs ) is the pixel coordinate of the ship, (u t ,v t ) is the pixel coordinate of the sea ice centroid, Br is the azimuth angle between the sea ice and the ship, px size is the ratio of the pixel distance to the actual distance, (S lon ,S lat ) is the latitude and longitude coordinate of the target ship, (T lon ,T lat ) is the latitude and longitude coordinate of the target sea ice centroid.

[0056] 3.2, preliminary matching of sea ice

[0057] After calculating the latitude and longitude coordinates of the sea ice centroid at two time points, the sea ice is preliminarily matched; the preliminary matching of sea ice refers to taking each block of sea ice in the second binary image Img1 as a template image, taking each block of sea ice in the first binary image Img0 as a target image, and calculating the similarity of each block of sea ice in the template image and the target image respectively by using a template matching algorithm, and matching the sea ice with a similarity less than a preset similarity threshold in the two images. The preliminary matching result of the sea ice is shown in Figure 2 . Wherein, the similarity is calculated according to the following formula:

[0058] R(u,v) = ∑ u`,v` |I(u+u`,v+v`)-T(u`,v`)| (7)

[0059] In the above formula, R(u,v) is the similarity variable, the value of 0 indicates that the two images are completely the same, and the larger the value, the smaller the similarity. R(u`,v`) is the template image, I(u,v) is the target image, and I(u+u`,v+v`) is the same size sub-image matched with the template image.

[0060] 3.3, correcting and matching sea ice

[0061] After the preliminary matching of sea ice, the sea ice is also corrected and matched; the correcting and matching of sea ice refers to calculating the moving distance D t of each block of sea ice within the time interval according to the calculated latitude and longitude coordinates of the sea ice centroid and the time interval, comparing the moving distance of the sea ice with the sea ice moving distance threshold set according to the time interval, retaining the matched sea ice with a moving distance less than the sea ice moving distance threshold, removing the redundant matching result, and then splicing the retained first binary image and the second binary image side by side, with Img0 on the left side and Img1 on the right side. The sea ice centroid of the matched sea ice is connected respectively to obtain a plurality of matched sea ice line segments, and the average slope of all line segments is calculated as the reference slope. When a block of sea ice is matched to multiple objects, the redundant matching is deleted according to the reference slope. The correcting and matching result of the sea ice is shown in Figure 3as shown.

[0062] 3.4, calculating sea ice movement parameters

[0063] Finally, the sea ice flow direction and sea ice flow speed are calculated according to the longitude and latitude coordinates of the matched sea ice centroids at two time points, specifically, as shown, the sea ice flow direction and sea ice flow speed are calculated according to the matched sea ice longitude and latitude coordinates ICE(lon0, lat0) at t0 and sea ice longitude and latitude coordinates ICE(lon1, lat1) at t1, and a dynamic ice area navigation scene is further constructed, which can help the driver obtain sea ice movement information. Wherein, the arrow direction represents the sea ice flow direction, and the text marks the sea ice flow direction and flow speed, and the sea ice flow direction and sea ice flow speed are calculated according to the following formula respectively: Figure 4

[0064] a = lat1 - lat0

[0065] b = lon1 - lon0

[0066]

[0067]

[0068] In the above formula, ICE speed is the sea ice flow speed, a is the latitude difference of the sea ice at two time points, b is the longitude difference of the sea ice at two time points, and t1-t0 is the time interval of the first marine radar image and the second marine radar image; ICE ang is the sea ice flow direction.

[0069] 4, sea ice collision probability calculation steps: in the dynamic ice area navigation scene, the closest encounter distance and the closest encounter time of a piece of sea ice and the target ship are calculated according to the distance and azimuth angle of the piece of sea ice and the target ship, the azimuth change of the piece of sea ice relative to the target ship is calculated according to the azimuth angle of the piece of sea ice and the target ship at two time points, and the closest encounter distance, the closest encounter time, the azimuth change, and the distance of the piece of sea ice and the target ship are normalized respectively to obtain the normalized closest encounter distance, the closest encounter time, the azimuth change, and the distance of the piece of sea ice and the target ship, and then the sea ice collision probability variable is obtained, the closest encounter distance weight, the closest encounter time weight, the azimuth change weight, and the distance weight of the piece of sea ice and the target ship are set according to the closest encounter distance, the closest encounter time, the azimuth change, and the distance of the piece of sea ice and the target ship, and then the sea ice collision probability weight is obtained, and the sea ice collision probability of the piece of sea ice is calculated according to the sea ice collision probability variable and the sea ice collision probability weight. This step belongs to the evaluation and early warning of sea ice risk in Figure 1 .

[0070] ​Specifically, the closest point of approach distance and the closest point of approach time between the target ship and the ice are calculated according to the distance and the azimuth angle between the target ship and the ice, respectively, and the closest point of approach distance and the closest point of approach time are calculated according to the following formulae:

[0071] DCPA i = D i × sin (φ i - Br i - π) (10)

[0072]

[0073] In the above formulae, i is the ice number; DCPA i is the closest point of approach distance of the ith ice; D i is the distance between the ith ice and the target ship; φ i is the direction of relative motion between the ith ice and the target ship; Br i is the azimuth angle of the ith ice relative to the target ship; TCPA i is the closest point of approach time of the ith ice.

[0074] The azimuth change of the ice relative to the target ship, i.e., the change of the compass azimuth of the ice, is calculated according to the azimuth angles of the ice relative to the target ship at two time points, and the change of the compass azimuth of the ice is calculated according to the following formula:

[0075] VCD i = |Br t - Br t-1 | (12)

[0076] In the above formula, VCD i is the azimuth change of the ith ice relative to the target ship; Br t is the azimuth angle of the ice relative to the target ship at time t; and Br t-1 is the azimuth angle of the ice relative to the target ship at time t-1.

[0077] Then, the closest point of approach distance, the closest point of approach time, the distance between the ice and the target ship, and the azimuth change of the ice relative to the target ship are normalized to obtain the normalized closest point of approach distance r DCPA , the normalized closest point of approach time r TCPA , the normalized distance between the ice and the target ship r D , and the normalized azimuth change of the ith ice relative to the target ship r VCD , and further to obtain the normalized ice collision probability variable r R and the normalized ice collision probability variable r R , and the expressions are as follows:

[0078] r R= [r DCPA ,r TCPA ,r VCD ,r D ] T (13)

[0079] In the above formula, T is the vector transpose.

[0080] Further, the closest encounter distance weight w DCPA , the closest encounter time weight w TCPA , the distance between the sea ice and the target ship weight w D , and the azimuth change weight w VCD of the sea ice relative to the target ship are set according to the closest encounter distance, the closest encounter time, and the distance between the sea ice and the target ship, so as to obtain the sea ice collision probability weight w R , wherein the sea ice collision probability weight w R is expressed as follows:

[0081] w R = [w DCPA , w TCPA , w VCD , w D ] (14)

[0082] Finally, the sea ice collision probability P i of the sea ice is calculated according to the sea ice collision probability variable and the sea ice collision probability weight, and the sea ice collision probability P i is calculated according to the following formula:

[0083] P i = w R × r R (15)

[0084] In the above formula, w R is the sea ice collision probability weight, and r R is the normalized sea ice collision probability variable.

[0085] 5、The sea ice collision risk index calculation step: according to the longitude and latitude coordinates of the centroid of a piece of sea ice and the minimum circumscribed rectangle of the piece of sea ice, the image matrix in which the piece of sea ice is located is obtained, the sea ice area of the piece of sea ice is calculated according to the image matrix in which the piece of sea ice is located, the sea ice area and the ship ice resistance level in the navigation information are normalized respectively to obtain the normalized sea ice area and the normalized ship ice resistance level, and then the normalized sea ice collision consequence variable is obtained, and the sea ice area weight and the ship ice resistance level weight are set according to the sea ice area and the ship ice resistance level, and then the sea ice collision consequence weight is obtained, the sea ice collision consequence of the piece of sea ice is calculated according to the sea ice collision consequence variable and the sea ice collision consequence weight, and the sea ice collision risk index of the piece of sea ice is calculated according to the calculated sea ice collision probability and the sea ice collision consequence of the piece of sea ice. This step belongs to the evaluation and early warning of sea ice risk in Figure 1 .

[0086] Specifically, the cv2.minAreaRect function of Python language is called to obtain the minimum circumscribed rectangle of the sea ice, the image matrix in which the piece of sea ice is located is obtained according to the longitude and latitude coordinates of the centroid of a piece of sea ice and the minimum circumscribed rectangle of the piece of sea ice, and the sea ice area Area of the piece of sea ice is calculated according to the image matrix in which the piece of sea ice is located. The sea ice area Area is calculated according to the following formula:

[0087] Area = ∑ I:src(I)≠0 1 (16)

[0088] In the formula, I is the image matrix in which the sea ice is located.

[0089] Then, the sea ice area and the ship ice resistance level in the navigation information are normalized respectively to obtain the normalized sea ice area r Area and the normalized ship ice resistance level r Ship , and then the normalized sea ice collision consequence variable r C is obtained. The sea ice collision consequence variable r C is expressed as follows:

[0090]

[0091] In the formula, T is the vector transpose.

[0092] Then, the sea ice area weight w Area and the ship ice resistance level weight w Ship are set according to the sea ice area and the ship ice resistance level, and then the sea ice collision consequence weight w C is obtained. The sea ice collision consequence weight w C is expressed as follows:

[0093] w C = [w Area , w Ship(18)

[0094] Then, based on the sea ice collision consequence variables and sea ice collision consequence weights, the sea ice collision consequence C of this sea ice block is calculated. i Consequences of sea ice collision C i Calculate according to the following formula:

[0095] C i =w C ×r C (19)

[0096] In the above formula, w C For the weighting of the consequences of sea ice collisions, r C For the consequences of sea ice collisions, variables are used.

[0097] Finally, based on the calculated probability of sea ice collision and the consequences of such a collision, the sea ice collision risk index (ICRI) is calculated. i Calculate according to the following formula:

[0098] ICRI i =P i ×C i (20)

[0099] In the above formula, P i Let C be the probability of sea ice collision. i The consequences of sea ice collisions.

[0100] 6. Sea ice risk warning procedures, such as... Figure 1 The sea ice risk visualization warning shown is as follows: the sea ice collision risk index of each piece of sea ice is compared with the preset risk thresholds of different levels, and the sea ice risk warnings of different levels are given according to the comparison results.

[0101] Specifically, the calculated sea ice collision risk index for each piece of sea ice is first compared with a preset first risk threshold and a second risk threshold. When the sea ice collision risk index of a certain piece of sea ice is less than or equal to the first risk threshold, that is, when the ICRI... i When the sea ice collision risk index is ≤0.33, a low-risk green alert is issued for the sea ice block. A lower risk green alert is issued when the sea ice collision risk index is greater than the first risk threshold and less than or equal to the second risk threshold, i.e., when 0.33... <ICRI i A medium-risk yellow alert is issued for the sea ice block when its sea ice collision risk index is ≤0.66. A lower alert (ICRI) is issued when the sea ice collision risk index exceeds the second risk threshold. i When the risk level is greater than 0.66, a high-risk red alert is issued for the sea ice block. It should be noted that the different risk thresholds and alert colors described above are merely illustrative examples and not the only possible interpretations. Specific threshold values ​​and alert colors can be adjusted according to actual needs.

[0102] The application also relates to an ice region navigation dynamic sea ice risk early warning system corresponding to the ice region navigation dynamic sea ice risk early warning method, which can be understood as a system for realizing the method, and comprises, which are connected in sequence, a marine radar image information acquisition module, a sea ice echo extraction module, a dynamic ice region navigation scene construction module, a sea ice collision probability calculation module, a sea ice collision risk index calculation module and a sea ice risk early warning module, the dynamic ice region navigation scene construction module comprises a longitude and latitude coordinate calculation module and a motion parameter calculation module which are directly or indirectly connected, the sea ice echo extraction module is connected with the longitude and latitude coordinate calculation module, and the motion parameter calculation module is connected with the sea ice collision probability calculation module. Specifically,

[0103] The marine radar image information acquisition module acquires navigation information of a target ship, first marine radar image information at a first time point and second marine radar image information at a second time point, and calculates a time interval between the two time points;

[0104] The sea ice echo extraction module extracts a region of interest displaying sea ice echoes in the first marine radar image information and the second marine radar image information through a mask matrix respectively, fills pixels of the region of interest of sea ice echoes and pixels of other regions except the region of interest of sea ice echoes with different colors respectively, and obtains first and second binary images of sea ice echoes at the two time points;

[0105] The longitude and latitude coordinate calculation module extracts pixel coordinates of the center of each piece of sea ice in the first and second binary images respectively, extracts pixel coordinates of the target ship according to longitude and latitude coordinates of the target ship in the navigation information, calculates distances and azimuth angles of each piece of sea ice and the target ship according to the pixel coordinates of the center of each piece of sea ice and the pixel coordinates of the target ship respectively, and calculates longitude and latitude coordinates of the center of each piece of sea ice at the two time points according to the distances, the azimuth angles and the pixel coordinates of the target ship;

[0106] The motion parameter calculation module calculates a sea ice flow direction and a sea ice flow speed according to the longitude and latitude coordinates of the center of each piece of sea ice at the two time points, and further constructs a dynamic ice region navigation scene;

[0107] The sea ice collision probability calculation module calculates the closest encounter distance and the closest encounter time between the target ship and the sea ice according to the distance and the azimuth angle between the target ship and the sea ice in the dynamic ice navigation scene, calculates the azimuth change of the sea ice relative to the target ship according to the azimuth angle between the target ship and the sea ice at two time points, and performs normalization processing on the closest encounter distance, the closest encounter time, the azimuth change, and the distance between the target ship and the sea ice, respectively, to obtain the normalized closest encounter distance, the normalized closest encounter time, the normalized azimuth change, and the normalized distance between the target ship and the sea ice, and then obtain the sea ice collision probability variable. The closest encounter distance weight, the closest encounter time weight, the azimuth change weight, and the distance weight between the target ship and the sea ice are set according to the closest encounter distance, the closest encounter time, the azimuth change, and the distance between the target ship and the sea ice, and then the sea ice collision probability weight is obtained. The sea ice collision probability of the sea ice is calculated according to the sea ice collision probability variable and the sea ice collision probability weight.

[0108] The sea ice collision risk index calculation module obtains the image matrix in which the sea ice is located according to the latitude and longitude coordinates of the sea ice centroid and the minimum circumscribed rectangle of the sea ice, calculates the sea ice area of the sea ice according to the image matrix in which the sea ice is located, performs normalization processing on the sea ice area and the ship ice resistance level in the navigation information respectively to obtain the normalized sea ice area and the normalized ship ice resistance level, and then obtains the normalized sea ice collision consequence variable. The sea ice area weight and the ship ice resistance level weight are set according to the sea ice area and the ship ice resistance level, and then the sea ice collision consequence weight is obtained. The sea ice collision consequence of the sea ice is calculated according to the sea ice collision consequence variable and the sea ice collision consequence weight. The sea ice collision risk index of the sea ice is calculated according to the sea ice collision probability and the sea ice collision consequence of the sea ice.

[0109] The sea ice risk early warning module compares the sea ice collision risk index of each sea ice with the preset risk threshold of different levels respectively, and performs corresponding sea ice risk early warning of different levels on the sea ice according to the comparison result.

[0110] Preferably, the navigation information includes a radar range, a distance scale circle interval, a ship position, a ship heading, a ground heading, and a ground speed.

[0111] Preferably, the dynamic ice region navigation scene construction module further comprises a preliminary matching sea ice sub-module, the latitude and longitude coordinate calculation module is connected to the motion parameter calculation module through the preliminary matching sea ice sub-module, at this time, the latitude and longitude coordinate calculation module is indirectly connected to the motion parameter calculation module; after the latitude and longitude coordinates of the sea ice centroids of the two time points are calculated by the latitude and longitude coordinate calculation module, the preliminary matching sea ice sub-module takes each piece of sea ice in the second binary image as a template image and each piece of sea ice in the first binary image as a target image, and calculates the similarity of each piece of sea ice in the template image and the target image respectively by using a template matching algorithm, and matches the sea ice in the two images with a similarity less than a preset similarity threshold; then, the motion parameter calculation module calculates the sea ice flow direction and the sea ice flow speed according to the latitude and longitude coordinates of the matched sea ice centroids of the two time points, and further constructs the dynamic ice region navigation scene.

[0112] Preferably, the dynamic ice region navigation scene construction module further comprises a preliminary matching sea ice sub-module, the latitude and longitude coordinate calculation module is connected to the motion parameter calculation module through the preliminary matching sea ice sub-module, at this time, the latitude and longitude coordinate calculation module is indirectly connected to the motion parameter calculation module; after the latitude and longitude coordinates of the sea ice centroids of the two time points are calculated by the latitude and longitude coordinate calculation module, the preliminary matching sea ice sub-module takes each piece of sea ice in the second binary image as a template image and each piece of sea ice in the first binary image as a target image, and calculates the similarity of each piece of sea ice in the template image and the target image respectively by using a template matching algorithm, and matches the sea ice in the two images with a similarity less than a preset similarity threshold; then, the motion parameter calculation module calculates the sea ice flow direction and the sea ice flow speed according to the latitude and longitude coordinates of the matched sea ice centroids of the two time points, and further constructs the dynamic ice region navigation scene.

[0113] Preferably, in the sea ice risk early warning module, comparing the sea ice collision risk index value of each piece of sea ice with the preset risk threshold values of different levels comprises:

[0114] When the sea ice collision risk index value of a piece of sea ice is less than or equal to the first risk threshold value, a low-risk green early warning is given to the piece of sea ice; when the sea ice collision risk index value of the piece of sea ice is greater than the first risk threshold value and less than or equal to the second risk threshold value, a medium-risk yellow early warning is given to the piece of sea ice; and when the sea ice collision risk index value of the piece of sea ice is greater than the second risk threshold value, a high-risk red early warning is given to the piece of sea ice.

[0115] The ice region navigation dynamic sea ice risk early warning method and system provided by the application can extract sea ice echo information based on the navigation information of a ship and through two-time point navigation radar images, construct a dynamic ice region navigation scene, and in the dynamic ice region navigation scene, calculate the sea ice collision probability and the sea ice collision consequence by using a specific calculation method, and then calculate the sea ice collision risk index, so as to visually warn the risk of each piece of sea ice relative to the ship, and greatly improve the accuracy of near-field sea ice perception.

[0116] It should be noted that the above specific embodiments can enable those skilled in the art to more fully understand the present application, but in no way limit the present application. Therefore, although the present application has been described in detail with reference to the drawings and examples, those skilled in the art should understand that the present application can still be modified or replaced by equivalents, in short, all technical solutions and improvements that do not deviate from the spirit and scope of the present application should be covered in the protection scope of the patent of the present application.

Claims

1. A method for dynamic sea ice risk warning for ice navigation, characterized in that, The method comprises the following steps: a marine radar image information acquisition step: acquiring sailing information of a target ship, first marine radar image information at a first time point, and second marine radar image information at a second time point, and calculating a time interval between the two time points; a sea ice echo extraction step: extracting a region of interest showing sea ice echoes from the first marine radar image information and the second marine radar image information respectively by using a mask matrix, and filling pixels in the region of interest of the sea ice echoes and pixels in other regions except the region of interest of the sea ice echoes with different colors respectively to obtain first and second binary images of sea ice echoes at the two time points; a dynamic ice area sailing scene construction step: extracting pixel coordinates of each sea ice centroid in the first and second binary images respectively, extracting pixel coordinates of the target ship according to latitude and longitude coordinates of the target ship in the sailing information, calculating distances and azimuth angles of each sea ice and the target ship according to the pixel coordinates of the sea ice centroid and the pixel coordinates of the target ship respectively, and calculating latitude and longitude coordinates of the sea ice centroid at the two time points according to the distances, the azimuth angles and the pixel coordinates of the target ship; then calculating a sea ice flow direction and a sea ice flow speed according to the latitude and longitude coordinates of the sea ice centroid at the two time points, and constructing a dynamic ice area sailing scene; a sea ice collision probability calculation step: in the dynamic ice area sailing scene, calculating a closest encounter distance and a closest encounter time of a certain piece of sea ice and the target ship according to the distance and the azimuth angle of the certain piece of sea ice and the target ship, calculating a change in the azimuth of the sea ice relative to the target ship according to the azimuth angles of the certain piece of sea ice and the target ship at the two time points, and performing normalization processing on the closest encounter distance, the closest encounter time, the change in the azimuth and the distance of the sea ice and the target ship respectively to obtain normalized closest encounter distance, closest encounter time, change in the azimuth and distance of the sea ice and the target ship, and then obtain a sea ice collision probability variable, setting a closest encounter distance weight, a closest encounter time weight, a change in the azimuth weight and a distance weight of the sea ice and the target ship according to the closest encounter distance, the closest encounter time, the change in the azimuth and the distance of the sea ice and the target ship, and then obtaining a sea ice collision probability weight, and calculating a sea ice collision probability of the certain piece of sea ice according to the sea ice collision probability variable and the sea ice collision probability weight; a sea ice collision risk index calculation step: obtaining an image matrix in which the certain piece of sea ice is located according to the latitude and longitude coordinates of the centroid of the certain piece of sea ice and a minimum circumscribed rectangle of the certain piece of sea ice, calculating a sea ice area of the certain piece of sea ice according to the image matrix in which the certain piece of sea ice is located, performing normalization processing on the sea ice area and a ship ice resistance level in the sailing information respectively to obtain normalized sea ice area and ship ice resistance level, and then obtaining a normalized sea ice collision consequence variable, setting a sea ice area weight and a ship ice resistance level weight according to the sea ice area and the ship ice resistance level, and then obtaining a sea ice collision consequence weight, calculating a sea ice collision consequence of the certain piece of sea ice according to the sea ice collision consequence variable and the sea ice collision consequence weight, and calculating a sea ice collision risk index of the certain piece of sea ice according to the sea ice collision probability and the sea ice collision consequence of the certain piece of sea ice. The sea ice risk early warning step comprises: comparing the sea ice collision risk index of each piece of sea ice with preset risk thresholds of different levels respectively, and performing corresponding sea ice risk early warning of different levels on the sea ice according to the comparison results.

2. The ice navigation dynamic sea ice risk warning method according to claim 1, characterized in that, In the step of obtaining the marine radar image information, the navigation information includes radar range, distance scale circle interval, ship position, ship heading, ground heading and ground speed.

3. The ice navigation dynamic sea ice risk warning method according to claim 1, characterized in that, In the step of constructing the dynamic ice area navigation scene, after the longitude and latitude coordinates of the sea ice centroids at the two time points are calculated, the sea ice is preliminarily matched; the preliminary matching of the sea ice refers to: taking each piece of sea ice in the second binary image as a template image, taking each piece of sea ice in the first binary image as a target image, and calculating the similarity of each piece of sea ice in the template image and the target image respectively by using a template matching algorithm, and matching the sea ice with a similarity less than a preset similarity threshold in the two images; the sea ice flow direction and the sea ice flow speed are calculated according to the longitude and latitude coordinates of the matched sea ice centroids at the two time points.

4. The ice navigation dynamic sea ice risk warning method according to claim 3, characterized in that, In the step of constructing the dynamic ice area navigation scene, after the sea ice is preliminarily matched, the matched sea ice is corrected; the correction of the matched sea ice refers to: calculating the moving distance of a piece of sea ice within a time interval according to the longitude and latitude coordinates of the centroid of the sea ice and the time interval, comparing the moving distance of the sea ice with a sea ice moving distance threshold set according to the time interval, retaining the matched sea ice with a moving distance less than the sea ice moving distance threshold, removing the redundant matching results, then splicing the retained first binary image and the second binary image side by side, connecting the sea ice centroids of the matched sea ice respectively, obtaining a plurality of matched sea ice line segments, and calculating the average slope of all the line segments, taking the average slope as a reference slope, and deleting the redundant matching according to the reference slope when a piece of sea ice is matched to multiple objects; the sea ice flow direction and the sea ice flow speed are calculated according to the longitude and latitude coordinates of the matched sea ice centroids at the two time points.

5. The ice navigation dynamic sea ice risk warning method according to one of claims 1 to 4, characterized in that, In the step of early warning the sea ice risk, the comparison of the sea ice collision risk index of each piece of sea ice with preset risk thresholds of different levels respectively comprises: When the sea ice collision risk index value of a piece of sea ice is less than or equal to a first risk threshold, the piece of sea ice is given a low-risk green early warning; when the sea ice collision risk index value of the piece of sea ice is greater than the first risk threshold and less than or equal to a second risk threshold, the piece of sea ice is given a medium-risk yellow early warning; and when the sea ice collision risk index value of the piece of sea ice is greater than the second risk threshold, the piece of sea ice is given a high-risk red early warning.

6. A dynamic sea ice risk warning system for ice navigation, characterized in that, The system comprises sequentially connected marine radar image information obtaining module, sea ice echo extracting module, dynamic ice area navigation scene constructing module, sea ice collision probability calculating module, sea ice collision risk index calculating module and sea ice risk early warning module; the dynamic ice area navigation scene constructing module comprises directly or indirectly connected longitude and latitude coordinate calculating module and motion parameter calculating module; the sea ice echo extracting module is connected with the longitude and latitude coordinate calculating module; and the motion parameter calculating module is connected with the sea ice collision probability calculating module. The navigation radar image information acquisition module acquires the navigation information of the target ship, the first navigation radar image information at the first time point, and the second navigation radar image information at the second time point, and calculates the time interval between the two time points; The sea ice echo extraction module extracts the region of interest showing the sea ice echo in the first navigation radar image information and the second navigation radar image information respectively through a mask matrix, and fills the pixels of the region of interest of the sea ice echo and the pixels of other regions except the region of interest of the sea ice echo with different colors respectively to obtain the first binary image and the second binary image of the sea ice echo at the two time points; The latitude and longitude coordinate calculation module extracts the pixel coordinates of each sea ice centroid in the first binary image and the second binary image respectively, and extracts the pixel coordinates of the target ship according to the latitude and longitude coordinates of the target ship in the navigation information, calculates the distance and azimuth angle of each sea ice and the target ship according to the pixel coordinates of the sea ice centroid and the pixel coordinates of the target ship respectively, and calculates the latitude and longitude coordinates of the sea ice centroid at the two time points according to the distance, the azimuth angle and the pixel coordinates of the target ship; The motion parameter calculation module calculates the sea ice flow direction and the sea ice flow speed according to the latitude and longitude coordinates of the sea ice centroid at the two time points, and further constructs a dynamic ice area navigation scene; The sea ice collision probability calculation module calculates the closest encounter distance and the closest encounter time of a certain piece of sea ice and the target ship according to the distance and the azimuth angle of the sea ice and the target ship in the dynamic ice area navigation scene, calculates the change of the azimuth of the sea ice relative to the target ship according to the azimuth angle of the sea ice and the target ship at the two time points, and normalizes the closest encounter distance, the closest encounter time, the change of the azimuth and the distance between the sea ice and the target ship respectively to obtain the normalized closest encounter distance, the closest encounter time, the change of the azimuth and the distance between the sea ice and the target ship, and further obtain the sea ice collision probability variable, sets the closest encounter distance weight, the closest encounter time weight, the change of the azimuth weight and the distance between the sea ice and the target ship weight according to the closest encounter distance, the closest encounter time, the change of the azimuth and the distance between the sea ice and the target ship, and further obtains the sea ice collision probability weight, and calculates the sea ice collision probability of the piece of sea ice according to the sea ice collision probability variable and the sea ice collision probability weight; The sea ice collision risk index calculation module obtains the image matrix where the piece of sea ice is located according to the latitude and longitude coordinates of the centroid of the piece of sea ice and the minimum circumscribed rectangle of the piece of sea ice, calculates the sea ice area of the piece of sea ice according to the image matrix where the piece of sea ice is located, normalizes the sea ice area and the ship ice resistance grade in the navigation information respectively to obtain the normalized sea ice area and the ship ice resistance grade, and further obtains the normalized sea ice collision consequence variable, sets the sea ice area weight and the ship ice resistance grade weight according to the sea ice area and the ship ice resistance grade, and further obtains the sea ice collision consequence weight, calculates the sea ice collision consequence of the piece of sea ice according to the sea ice collision consequence variable and the sea ice collision consequence weight, and calculates the sea ice collision risk index of the piece of sea ice according to the sea ice collision probability and the sea ice collision consequence of the piece of sea ice. The sea ice risk early warning module compares the sea ice collision risk index of each piece of sea ice with preset risk thresholds of different levels respectively, and performs corresponding sea ice risk early warning of different levels on the sea ice according to the comparison result.

7. The ice regime navigation dynamic sea ice risk warning system according to claim 6, characterized in that, The navigation information includes radar range, distance scale circle interval, ship position, ship heading, ground heading and ground speed.

8. The ice regime navigation dynamic sea ice risk warning system according to claim 6, characterized in that, The dynamic ice area navigation scene construction module further comprises a preliminary matching sea ice submodule, the latitude and longitude coordinate calculation module is connected to the motion parameter calculation module through the preliminary matching sea ice submodule, after the latitude and longitude coordinates of the sea ice centroids of the two time points are calculated by the latitude and longitude coordinate calculation module, the preliminary matching sea ice submodule takes each piece of sea ice in the second binary image as a template image and each piece of sea ice in the first binary image as a target image, and calculates the similarity of each piece of sea ice in the template image and the target image respectively by using a template matching algorithm, and matches the sea ice with a similarity less than a preset similarity threshold in the two images; then the motion parameter calculation module calculates the sea ice flow direction and the sea ice flow speed according to the latitude and longitude coordinates of the matched sea ice centroids of the two time points.

9. The ice navigation dynamic sea ice risk warning system of claim 8, wherein, The dynamic ice area navigation scene construction module further comprises a preliminary matching sea ice submodule, the latitude and longitude coordinate calculation module is connected to the motion parameter calculation module through the preliminary matching sea ice submodule, after the latitude and longitude coordinates of the sea ice centroids of the two time points are calculated by the latitude and longitude coordinate calculation module, the preliminary matching sea ice submodule takes each piece of sea ice in the second binary image as a template image and each piece of sea ice in the first binary image as a target image, and calculates the similarity of each piece of sea ice in the template image and the target image respectively by using a template matching algorithm, and matches the sea ice with a similarity less than a preset similarity threshold in the two images; then the motion parameter calculation module calculates the sea ice flow direction and the sea ice flow speed according to the latitude and longitude coordinates of the matched sea ice centroids of the two time points.

10. The ice navigation dynamic sea ice risk warning system according to one of claims 6 to 9, characterized in that, The sea ice risk early warning module compares the sea ice collision risk index of each piece of sea ice with preset risk thresholds of different levels respectively, and performs corresponding sea ice risk early warning of different levels on the sea ice according to the comparison result. When the sea ice collision risk index value of a piece of sea ice is less than or equal to the first risk threshold, the piece of sea ice is given a low-risk green early warning; when the sea ice collision risk index value of the piece of sea ice is greater than the first risk threshold and less than or equal to the second risk threshold, the piece of sea ice is given a medium-risk yellow early warning; and when the sea ice collision risk index value of the piece of sea ice is greater than the second risk threshold, the piece of sea ice is given a high-risk red early warning.

Images