Oil spill monitoring method and system combined with image analysis processing technology

Abstract

The invention relates to the technical field of marine oil spill pollution monitoring, in particular to an oil spill monitoring method and system combined with an image analysis processing technology. The method comprises the following steps: S1, acquiring an initial oil spill diffusion path through remote sensing monitoring equipment, extracting a floating ice distribution image in a path coverage area, selecting a floating ice image at the tail end of the path as a reference image, and establishing an adjacent floating ice image matching model; s2, based on the adjacent floating ice image matching model, determining a minimum feature floating ice image through an adaptive pixel threshold algorithm; and establishing an oil-ice complex edge feature matching matrix. According to the method, the recognition precision is improved through multi-feature fusion and a self-adaptive threshold value, oil film permeation and streaming laws are analyzed through multi-scale analysis, a spatio-temporal evolution model is constructed to realize diffusion path closed-loop tracking, phase change nodes are accurately captured in combination with a dynamic verification formula, the environmental adaptability of a complex ice region is remarkably enhanced, and the method is suitable for popularization and application. And real-time early warning and ecological protection decision support is provided for polar region oil spill accidents.

Description

Technical Field

[0001] The present invention relates to the technical field of marine oil spill pollution monitoring, and in particular to an oil spill monitoring method and system combined with image analysis and processing technology. Background Art

[0002] With the continuous increase in the frequency of navigation in the Arctic route and the deepening of marine resource development activities, oil spill pollution monitoring in polar waters faces unprecedented technical challenges. This special environment has multiple physical field coupling effects such as dynamic displacement of floating ice, low-temperature solidification of oil film, and current disturbance. Traditional single-point monitoring methods are difficult to effectively capture the temporal and spatial evolution of oil-ice complexes. Especially in the ice edge area, the nonlinear interaction between the ice fragmentation and reorganization process and the oil spill diffusion path leads to significant deviations in the assessment of the scope of pollution.

[0003] The current monitoring system faces three major technical bottlenecks: First, the continuity of floating ice in remote sensing images is easily disturbed by polar day / polar night lighting conditions, resulting in inaccurate modeling boundary determination; second, there is a lack of quantitative evaluation standards for the interaction between oil film and floating ice, and the existing threshold method is difficult to adapt to the dynamic changes in ice porosity; third, the phase change process in low temperature environments accelerates the transformation of oil ice morphology, and traditional static models cannot provide real-time feedback on the periodic characteristics of ice aggregation and fragmentation. These defects seriously restrict the timeliness of pollution spread prediction and the scientific nature of disposal plans. Summary of the invention

[0004] In order to overcome the shortcomings of limited perception, delayed response and model lag, the present invention provides an oil spill monitoring method and system combined with image analysis and processing technology.

[0005] The technical implementation scheme of the present invention is: an oil spill monitoring method combined with image analysis and processing technology, comprising the following steps:

[0006] S1: Obtain the initial oil spill diffusion path through remote sensing monitoring equipment, extract the floating ice distribution image in the path coverage area, select the floating ice image at the end of the path as the reference image, and establish the adjacent floating ice image matching model;

[0007] S2: Based on the neighboring floating ice image matching model, the minimum feature floating ice image is determined by an adaptive pixel threshold algorithm; and an edge feature matching matrix of the oil-ice complex is established;

[0008] S3: Combined with the spatiotemporal sequence analysis, a cyclical model of the morphological evolution of the oil-ice complex is constructed. The cyclical model characterizes the periodic change process of the ice body from fragmentation to aggregation when the oil film wraps the floating ice;

[0009] S4: A dynamic verification formula is used to quantitatively verify the key phase change nodes in the cycle model, wherein the key phase change nodes include three characteristic stages of the small, medium and large states of the projected area of ​​the oil-ice complex.

[0010] Preferably, the method of acquiring the initial oil spill diffusion path through remote sensing monitoring equipment, extracting the floating ice distribution image within the path coverage area, selecting the floating ice image at the end of the path as the reference image, and establishing a matching model for adjacent floating ice images includes:

[0011] If the floating ice is distributed continuously in the floating ice distribution image within the path coverage area, then the adjacent monitoring path closest to the initial oil spill diffusion path is obtained until the floating ice is obtained to be distributed discontinuously;

[0012] The continuous distribution means that the ice floes are uniform in size and distributed at equal distances.

[0013] Preferably, the determining the minimum feature ice floe image by an adaptive pixel threshold algorithm based on the neighboring ice floe image matching model comprises:

[0014] Obtaining K floating ice distribution images through an adaptive threshold calculation formula, where K is greater than or equal to 3, and the K floating ice distribution images include three features, namely, floating ice edge recognition, floating ice edge texture, and floating ice edge complexity;

[0015] Taking the image with the highest edge recognition degree of the ice floes as the benchmark, the edge features of the ice floes distribution image with the highest edge recognition degree are extracted from the K ice floes distribution images, and then the edge features of the ice floes distribution image with the clearest edge texture are fused with the edge features of the ice floes distribution image with the most complex edge; the fused edge features are defined as the first edge features;

[0016] Based on the first edge feature, the first edge feature is sorted at equal intervals according to the feature quantity of the first edge feature;

[0017] The adaptive threshold calculation formula is as follows:

[0018]

[0019] in, is the adaptive threshold, is the average gray value of the local area of ​​the image, is the local contrast, is the edge direction gradient modulus, , is the adjustment coefficient.

[0020] Preferably, the K floating ice distribution images include three features, namely, floating ice edge recognition, floating ice edge texture and floating ice edge complexity, including:

[0021] Edge recognition: quantified by analyzing edge continuity and gradient direction consistency;

[0022] Edge texture: multi-scale texture features and frequency domain energy distribution, distinguishing between regular arrangement of floating ice crystals and blurred texture caused by oil film coverage;

[0023] Edge complexity: Quantifying contour irregularities based on fractal geometry and local curvature variations.

[0024] Preferably, the step of establishing an oil-ice complex edge feature matching matrix includes:

[0025] Based on the minimum characteristic floating ice image, screening out the floating ice distribution image that meets the second edge characteristic standard;

[0026] At the same time, the floating ice distribution images that meet the third edge feature standard are screened out;

[0027] Based on the first edge feature, the second edge feature and the third edge feature, edge matching is performed on the ice floe distribution image after the oil spill and the ice floe are combined;

[0028] The second edge feature standard is: among the three features, at least one sub-feature threshold is strictly higher than the first edge feature construction standard, and the remaining sub-feature thresholds are not lower than the first edge feature standard;

[0029] The third edge feature standard is: among the three features, at least one sub-feature threshold is strictly lower than the first edge feature construction standard, and the remaining sub-feature thresholds are not higher than the first edge feature standard;

[0030] The construction method of the second edge feature and the third edge feature follows the fusion algorithm framework of the first edge feature, and adjusts the adjustment coefficient in the adaptive threshold calculation formula according to the sub-feature threshold difference , .

[0031] Preferably, the edge matching of the floating ice distribution image after the oil spill and the floating ice are combined based on the first edge feature, the second edge feature and the third edge feature includes:

[0032] The first edge feature is divided by the second edge feature as a first ratio;

[0033] The value of the third edge feature divided by the second edge feature is taken as the second ratio;

[0034] The deviation between the first ratio and the second ratio is used as the edge matching amount of the floating ice distribution image after the oil spill and the floating ice are combined.

[0035] Preferably, the construction of a morphological evolution cycle model of oil-ice complexes in combination with spatiotemporal sequence analysis includes:

[0036] Based on the neighboring floating ice distribution images and edge matching information, N floating ice distribution images are obtained, and the floating ice distribution images are sorted from large to small or from small to large according to their sizes;

[0037] Based on the sorting results, the image sequence that reaches the complete morphological evolution cycle is obtained, and the key phase change nodes in the image sequence process of the complete morphological evolution cycle of the oil-ice complex are verified using the dynamic verification formula;

[0038] The complete morphological evolution cycle refers to the period of N1 consecutive time intervals between the oil-ice complex and the Within a certain period of time, at least one cycle of morphological change from fragmentation to aggregation or from aggregation to fragmentation is completed.

[0039] Preferably, the use of a dynamic verification formula to quantitatively verify key phase change nodes in the cycle model includes:

[0040] During each verification process, a simulation key phase change node is generated. Specifically, when the verification parameter reaches a preset threshold, an initial parameter set for the next phase change node is automatically generated.

[0041] Preferably, the use of a dynamic verification formula to quantitatively verify the key phase change nodes in the cycle model includes: the dynamic verification formula is as follows,

[0042]

[0043] in, To verify the value, is the rate of change of the projected area of ​​the oil-ice complex per unit time, is the ratio of the total edge length to the area of ​​the region, is the spatial variation intensity of the temperature field on the floating ice surface, To adjust the sensitivity of the temperature gradient, is the projected area of ​​the oil-ice complex in the image, is the time interval of the image sequence, is the total edge length of the oil-ice complex.

[0044] Preferably, the oil spill monitoring system combined with image analysis and processing technology comprises:

[0045] The image acquisition and modeling module uses multispectral remote sensing imaging equipment to obtain the oil spill diffusion path and ice floe distribution data, dynamically identifies the distribution state through the ice floe continuity algorithm, and automatically tracks the adjacent paths until the non-continuous ice floe image is captured as the modeling boundary;

[0046] The feature extraction module uses adaptive pixel thresholds to screen ice floes images, integrates edge recognition, texture clarity and complexity three-dimensional analysis technology, builds a multi-level feature matching matrix to achieve dynamic comparison of oil-ice complexes, and quantifies the oil film-ice interaction effect through feature ratios;

[0047] The dynamic modeling module builds a morphological evolution model based on the spatiotemporal sequence, and combines the temperature field gradient factor to present the periodic changes from ice fragmentation to aggregation, covering the evolution characteristics from small to large states;

[0048] The verification and optimization module uses dynamic verification formulas to evaluate phase change nodes, triggers morphological reconstruction through area change rate thresholds, and automatically generates new phase change nodes and optimizes parameters when the prediction deviation exceeds the limit, forming a parameter self-optimization closed-loop system.

[0049] Beneficial effects: The present invention, by constructing a multimodal feature fusion model and a dynamic adaptive threshold mechanism, first breaks through the fuzzy edge recognition problem caused by the complex morphology of floating ice in traditional oil spill monitoring, and realizes high-precision segmentation of the oil film-floating ice interface. On this basis, through hierarchical feature extraction and relative ratio analysis, a scale-related oil-ice interaction quantitative model is established to accurately analyze the composite diffusion law of microscopic infiltration and macroscopic bypass. Then, based on the spatiotemporal sequence drive, a morphological evolution cycle model is constructed, and through the dynamic verification formula, the diffusion rate mutation and phase change nodes are dynamically captured to achieve full-cycle closed-loop tracking of the oil spill path. Finally, through the adaptive phase change state transition mechanism and threshold trigger logic, an environmentally responsive monitoring strategy is formed to systematically improve the robustness and timeliness of oil spill monitoring in complex ice areas in the Arctic, and provide intelligent decision-making support for polar ecological security. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 A flowchart of an oil spill monitoring method combined with image analysis and processing technology of the present invention;

[0051] Figure 2 The present invention is a structural schematic diagram of an oil spill monitoring system combined with image analysis and processing technology. DETAILED DESCRIPTION

[0052] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0053] Embodiment 1: A method for oil spill monitoring using a combined image analysis and processing technique, such as Figure 1 As shown, the following steps are included:

[0054] S1: Obtain the initial oil spill diffusion path through remote sensing monitoring equipment, extract the floating ice distribution image in the path coverage area, select the floating ice image at the end of the path as the reference image, and establish the adjacent floating ice image matching model;

[0055] S2: Based on the neighboring floating ice image matching model, the minimum feature floating ice image is determined by an adaptive pixel threshold algorithm; and an edge feature matching matrix of the oil-ice complex is established;

[0056] S3: Combined with the spatiotemporal sequence analysis, a cyclical model of the morphological evolution of the oil-ice complex is constructed. The cyclical model characterizes the periodic change process of the ice body from fragmentation to aggregation when the oil film wraps the floating ice;

[0057] S4: A dynamic verification formula is used to quantitatively verify the key phase change nodes in the cycle model, wherein the key phase change nodes include three characteristic stages of the small, medium and large states of the projected area of ​​the oil-ice complex.

[0058] The small state means that the oil-ice complex projected area A is less than or equal to A1;

[0059] The medium state means that the oil-ice complex projected area A is greater than A1 and less than or equal to A2;

[0060] The large state refers to the oil-ice complex projected area greater than A2;

[0061] A1 and A2 are preset thresholds.

[0062] The initial oil spill diffusion path is obtained through remote sensing monitoring equipment, the floating ice distribution image in the path coverage area is extracted, the floating ice image at the end of the path is selected as the reference image, and the adjacent floating ice image matching model is established, including:

[0063] If the floating ice is distributed continuously in the floating ice distribution image within the path coverage area, then the adjacent monitoring path closest to the initial oil spill diffusion path is obtained until the floating ice is obtained to be distributed discontinuously;

[0064] The continuous distribution means that the ice floes are uniform in size and distributed at equal distances.

[0065] A further explanation is that when floating ice is distributed continuously, the oil film will be blocked by the ice body to form a pseudo diffusion path. The core logic of using progressive path expansion is that continuous floating ice will change the surface tension distribution of the oil spill; monitoring requirements: it is necessary to track the ice gap area (discontinuous area) to obtain the real diffusion path; if it is changed to fixed range monitoring, the key transition area will be missed, and random sampling cannot guarantee the continuity of the path.

[0066] Based on the neighboring floating ice image matching model, determining the minimum feature floating ice image by using an adaptive pixel threshold algorithm, including:

[0067] Obtaining K floating ice distribution images through an adaptive threshold calculation formula, where K is greater than or equal to 3, and the K floating ice distribution images include three features, namely, floating ice edge recognition, floating ice edge texture, and floating ice edge complexity;

[0068] Taking the image with the highest edge recognition degree of the ice floes as the benchmark, the edge features of the ice floes distribution image with the highest edge recognition degree are extracted from the K ice floes distribution images, and then the edge features of the ice floes distribution image with the clearest edge texture are fused with the edge features of the ice floes distribution image with the most complex edge; the fused edge features are defined as the first edge features;

[0069] Based on the first edge feature, the first edge feature is sorted at equal intervals according to the feature quantity of the first edge feature;

[0070] The adaptive threshold calculation formula is as follows:

[0071]

[0072] in, is the adaptive threshold, is the average gray value of the local area of ​​the image, is the local contrast, is the edge direction gradient modulus, , is the adjustment coefficient.

[0073] Further explanation is that accurate identification of oil spill edges is achieved through multimodal feature fusion. First, three core features are extracted from the floating ice remote sensing image: edge recognition (quantifying contour continuity), edge texture (characterizing surface material properties) and edge complexity (describing morphological irregularities). Then, based on the adaptive threshold model, the local grayscale mean ( ), relative contrast ( ) and the edge direction gradient modulus ( ) is combined to dynamically calculate the pixel segmentation threshold. High noise interference is suppressed by nonlinear logarithmic transformation, and edge consistency is improved by coupling directional constraints to achieve adaptive detection of the oil film-floating ice interface.

[0074] Further explanation is that the adaptive threshold The calculation formula is divided into three parts: the baseline grayscale term establishes the brightness benchmark, the logarithmic contrast term enhances the sensitivity of weak signals, and the directional constraint term suppresses broken edges. The baseline grayscale term uses the local mean Reflects the light and dark characteristics of the area; the logarithmic comparison term is ln(1+ ) realizes nonlinear mapping of contrast relative intensity ( is the local standard deviation), when When <0.3, the approximate linear response preserves details. >0.5 when compressing overexposure risk; direction constraint item ( )middle, The discreteness of the edge direction is quantified (in the range of 0-π), and β adjusts the direction consistency weight to effectively suppress the misjudgment caused by ice cracks. The formula constructs a dynamic segmentation boundary with scene adaptability by coupling the three physical quantities of brightness, contrast, and direction.

[0075] The K floating ice distribution images include three features, namely, floating ice edge recognition, floating ice edge texture and floating ice edge complexity, including:

[0076] Edge recognition: quantified by analyzing edge continuity and gradient direction consistency;

[0077] Edge texture: multi-scale texture features and frequency domain energy distribution, distinguishing between regular arrangement of floating ice crystals and blurred texture caused by oil film coverage;

[0078] Edge complexity: Quantifying contour irregularities based on fractal geometry and local curvature variations.

[0079] Establish the edge feature matching matrix of the oil-ice complex, including:

[0080] Based on the minimum characteristic floating ice image, screening out the floating ice distribution image that meets the second edge characteristic standard;

[0081] At the same time, the floating ice distribution images that meet the third edge feature standard are screened out;

[0082] Based on the first edge feature, the second edge feature and the third edge feature, edge matching is performed on the ice floe distribution image after the oil spill and the ice floe are combined;

[0083] The second edge feature standard is: among the three features, at least one sub-feature threshold is strictly higher than the first edge feature construction standard, and the remaining sub-feature thresholds are not lower than the first edge feature standard;

[0084] The third edge feature standard is: among the three features, at least one sub-feature threshold is strictly lower than the first edge feature construction standard, and the remaining sub-feature thresholds are not higher than the first edge feature standard;

[0085] The construction method of the second edge feature and the third edge feature follows the fusion algorithm framework of the first edge feature, and adjusts the adjustment coefficient in the adaptive threshold calculation formula according to the sub-feature threshold difference , .

[0086] Further explanation is that the traditional method uses a single threshold screening, which cannot distinguish the difference in the effect of oil film on ice floes of different sizes, resulting in confusion between the microscopic penetration (small ice floe gaps) and the macroscopic bypass (large ice floe blocking) effects, and distortion of the pollution diffusion model. The second edge feature (high threshold): strengthens the capture of local details of oil film infiltration into tiny ice crevices to avoid interference from the edges of large ice floes; the third edge feature (low threshold): relaxes the screening conditions to capture the path deflection characteristics of oil film bypassing large ice floes; synergy: by comparing high / low threshold features, quantify the scale interaction effect between oil film and ice floes, and accurately model the fractal law of the diffusion path. The traditional adaptive algorithm parameters are fixed (such as the Otsu method), and the edge mis-segmentation rate surges under the complex polar lighting and temperature fields (such as ice surface reflections are misjudged as oil film). Dynamic parameter adjustment: Adaptively adjust α (contrast sensitivity) and β (directional constraint weight) according to the difference in sub-feature thresholds. For example: High threshold scenario (second feature): increase α to enhance weak signal capture, and reduce β to tolerate complex edge directions; Low threshold scenario (third feature): reduce α to suppress noise, and increase β to enhance edge consistency; Environmental adaptability: The same algorithm framework is compatible with different ice conditions to avoid the computational overhead of multi-model switching. The absolute threshold method is interfered by environmental noise (such as surface texture of floating ice and snow cover), which leads to the accumulation of edge matching deviations, especially in long-term monitoring. The relative ratio eliminates system errors: Taking the second feature as the benchmark, the relative deviation of the first and third features is calculated to suppress global interference such as uneven lighting and equipment resolution differences;

[0087] Improved noise resistance: By fusing multi-scale features with deviation, the robustness to local anomalies (such as temporary frozen oil film) is enhanced.

[0088] Based on the first edge feature, the second edge feature and the third edge feature, edge matching is performed on the floating ice distribution image after the oil spill and the floating ice are combined, including:

[0089] The first edge feature is divided by the second edge feature as a first ratio;

[0090] The value of the third edge feature divided by the second edge feature is taken as the second ratio;

[0091] The deviation between the first ratio and the second ratio is used as the edge matching amount of the floating ice distribution image after the oil spill and the floating ice are combined.

[0092] Further explanation is that normalization eliminates system deviation: the ratio calculation is based on the second feature, suppressing global interference factors such as equipment resolution and light intensity, and improving cross-time / space data consistency; strengthening local feature comparison: the first ratio (first / second) highlights the penetration advantage of the oil film on small floating ice, and the second ratio (third / second) reflects the bypass effect of the oil film on large floating ice, quantifying the scale difference through relative values. Bidirectional interactive modeling: the deviation integrates the difference between the two ratios and captures the oil film penetration-circulation coupling effect at the same time, for example: the deviation is positive: penetration dominates, and the oil film quickly invades the ice gap; the deviation is negative: bypass dominates, and the oil film path is deflected by large floating ice; nonlinear response: the deviation is sensitive to small changes and can provide early warning of sudden changes in diffusion trends (such as oil film breaking through ice barriers).

[0093] Combined with the spatiotemporal sequence analysis, a cyclic model of the morphological evolution of the oil-ice complex is constructed, including:

[0094] Based on the neighboring floating ice distribution images and edge matching information, N floating ice distribution images are obtained, and the floating ice distribution images are sorted from large to small or from small to large according to their sizes;

[0095] Based on the sorting results, the image sequence that reaches the complete morphological evolution cycle is obtained, and the key phase change nodes in the image sequence process of the complete morphological evolution cycle of the oil-ice complex are verified using the dynamic verification formula;

[0096] The complete morphological evolution cycle refers to the period of N1 consecutive time intervals between the oil-ice complex and the Within a certain period of time, at least one cycle of morphological change from fragmentation to aggregation or from aggregation to fragmentation is completed.

[0097] A further explanation is that the time dimension of oil film diffusion is mapped by size sorting. The small→large sorting corresponds to the ice floe aggregation process, and the large→small sorting reflects the ice body fragmentation stage. When the image sequence presents a complete size change cycle (such as small→large→small or large→small→large), the cycle verification is triggered to capture the periodic characteristics of oil film diffusion. The three key phase change nodes of small state (initial infiltration), medium state (stable diffusion), and large state (path saturation) are located in the sorting sequence, and the transition parameters between each node are quantified through the dynamic verification formula. "Neighboring ice floe distribution image": refers to the ice floe distribution image whose straight-line distance to the initial oil spill diffusion path is less than a preset threshold. The preset threshold is calibrated by the average size of ice floes and the resolution of remote sensing images. Verification refers to the preliminary screening of the key phase change nodes (small state, medium state, and large state) in the oil-ice complex morphological evolution cycle model through the dynamic verification formula. The specific steps are: input the spatiotemporal sequence image data; calculate the verification value ;like The value is within the preset range (such as 0.8≤ ≤1.2), determine that the node is valid; remove abnormal nodes to ensure that the model conforms to the actual diffusion law.

[0098] Dynamic verification formulas are used to quantitatively verify key phase change nodes in the cycle model, including:

[0099] During each verification process, a simulation key phase change node is generated. Specifically, when the verification parameter reaches a preset threshold, an initial parameter set for the next phase change node is automatically generated.

[0100] A further explanation is that the oil film diffusion process is discretized into a phase change node chain, and each node defines unique morphological and dynamic constraints; only when the current node feature quantity (such as edge matching degree, projection area change rate) continues to meet the threshold requirements, the next node is allowed to be generated to ensure that the evolution process conforms to physical laws; each time a new node is entered, the subsequent threshold is adaptively adjusted according to the verification result of the previous node to form a learning optimization closed loop. Each phase change node defines a unique verification standard (such as morphological stability coefficient ≥ 0.8, diffusion rate fluctuation < 15%) to ensure the physical rationality of the stage evolution; only when all parameters of the current node continue to meet the threshold, the system generates the next node parameter set (such as increasing the projection area threshold of the medium form to 3 times the initial value), forcing the evolution process to follow the laws of oil film diffusion dynamics; if the verification fails, the backtracking mechanism is triggered to recalibrate the parameters of the previous node to avoid error accumulation. Quantitative verification is to optimize the parameters of the nodes after preliminary verification, which specifically includes: analysis Value parameters (such as The gradient, deviation); when the parameter exceeds the threshold (such as >0.5 m 2 / h), automatically generate the initial parameters of the next node; feed the new parameters back to the model, and iterate and optimize until the error is less than the preset threshold.

[0101] The dynamic verification formula is used to quantitatively verify the key phase change nodes in the cycle model, including: The dynamic verification formula is as follows,

[0102]

[0103] in, To verify the value, is the rate of change of the projected area of ​​the oil-ice complex per unit time, is the ratio of the total edge length to the area of ​​the region, is the spatial variation intensity of the temperature field on the floating ice surface, To adjust the sensitivity of the temperature gradient, is the projected area of ​​the oil-ice complex in the image, is the time interval of the image sequence, is the total edge length of the oil-ice complex.

[0104] Further explanation is that the dynamic change factor ( ): quantifies the oil film diffusion rate, reflecting the intensity of the morphological evolution dominated by the aggregation / fragmentation of floating ice; the morphological complexity factor ( ): Characterizes the edge irregularity of the oil-ice mixture. The larger the value, the more significant the contour fractal characteristics caused by oil film infiltration; Temperature constraint factor ( ): verification weight for suppressing abnormal surface temperature gradient areas, When the temperature increases, the exponential decay is strengthened, reducing the interference of local freeze-thaw mutations; when the oil film spreads stably, Keep it moderate and positive. Reduced temperature gradient due to edge regularity Approaching zero, verifying the value It shows stable characteristics; when phase change occurs (such as oil film breaking through the barrier of floating ice), Sudden increase, Rising from the broken edges, The oil film absorbs heat and generates a gradient. The pulse peak appears in the value to mark the phase change node; the projection area calculation method is specifically: adaptive threshold segmentation is performed on the image to extract the oil-ice complex area; the total number of pixels in the connected area is counted and multiplied by the single-pixel actual area calibration value; the edge length calculation method is specifically: multi-scale gradient detection is performed on the oil-ice complex area; the number of continuous edge pixels is counted and multiplied by the single-pixel actual length calibration value.

[0105] Embodiment 2: Based on embodiment 1, Figure 2 As shown, an oil spill monitoring system combined with image analysis and processing technology includes:

[0106] The image acquisition and modeling module uses multispectral remote sensing imaging equipment to obtain the oil spill diffusion path and ice floe distribution data, dynamically identifies the distribution state through the ice floe continuity algorithm, and automatically tracks the adjacent paths until the non-continuous ice floe image is captured as the modeling boundary;

[0107] The feature extraction module uses adaptive pixel thresholds to screen ice floes images, integrates edge recognition, texture clarity and complexity three-dimensional analysis technology, builds a multi-level feature matching matrix to achieve dynamic comparison of oil-ice complexes, and quantifies the oil film-ice interaction effect through feature ratios;

[0108] The dynamic modeling module builds a morphological evolution model based on the spatiotemporal sequence, and combines the temperature field gradient factor to present the periodic changes from ice fragmentation to aggregation, covering the evolution characteristics from small to large states;

[0109] The verification and optimization module uses dynamic verification formulas to evaluate phase change nodes, triggers morphological reconstruction through area change rate thresholds, and automatically generates new phase change nodes and optimizes parameters when the prediction deviation exceeds the limit, forming a parameter self-optimization closed-loop system.

[0110] The above is a detailed introduction to the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea. At the same time, for general technical personnel in this field, according to the idea of ​​the present application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. An oil spill monitoring method combined with image analysis and processing technology, characterized in that: The following steps are involved: S1: Obtain the initial oil spill diffusion path through remote sensing monitoring equipment, extract the floating ice distribution image in the path coverage area, select the floating ice image at the end of the path as the reference image, and establish the adjacent floating ice image matching model; S2: Based on the neighboring floating ice image matching model, the minimum feature floating ice image is determined by an adaptive pixel threshold algorithm; and an edge feature matching matrix of the oil-ice complex is established; S3: Combined with the spatiotemporal sequence analysis, a cyclical model of the morphological evolution of the oil-ice complex is constructed. The cyclical model characterizes the periodic change process of the ice body from fragmentation to aggregation when the oil film wraps the floating ice; S4: A dynamic verification formula is used to quantitatively verify the key phase change nodes in the cycle model, wherein the key phase change nodes include three characteristic stages of the small, medium and large states of the projected area of ​​the oil-ice complex.

2. The oil spill monitoring method combined with image analysis and processing technology according to claim 1 is characterized in that: The method of obtaining the initial oil spill diffusion path through remote sensing monitoring equipment, extracting the floating ice distribution image in the path coverage area, selecting the floating ice image at the end of the path as the reference image, and establishing the adjacent floating ice image matching model includes: If the floating ice is distributed continuously in the floating ice distribution image within the path coverage area, then the adjacent monitoring path closest to the initial oil spill diffusion path is obtained until the floating ice is obtained to be distributed discontinuously; The continuous distribution means that the ice floes are uniform in size and distributed at equal distances.

3. The oil spill monitoring method combined with image analysis and processing technology according to claim 1 is characterized in that: The method of determining the minimum feature ice floe image by using an adaptive pixel threshold algorithm based on the adjacent ice floe image matching model includes: Obtaining K floating ice distribution images through an adaptive threshold calculation formula, where K is greater than or equal to 3, and the K floating ice distribution images include three features, namely, floating ice edge recognition, floating ice edge texture, and floating ice edge complexity; Taking the image with the highest edge recognition degree of the ice floes as the benchmark, the edge features of the ice floes distribution image with the highest edge recognition degree are extracted from the K ice floes distribution images, and then the edge features of the ice floes distribution image with the clearest edge texture are fused with the edge features of the ice floes distribution image with the most complex edge; the fused edge features are defined as the first edge features; Based on the first edge feature, the first edge feature is sorted at equal intervals according to the feature quantity of the first edge feature; The adaptive threshold calculation formula is as follows: ； in, is the adaptive threshold, is the average gray value of the local area of ​​the image, is the local contrast, is the edge direction gradient modulus, , is the adjustment coefficient.

4. The oil spill monitoring method combined with image analysis and processing technology according to claim 3 is characterized in that: The K floating ice distribution images include three features, namely, floating ice edge recognition, floating ice edge texture and floating ice edge complexity, including: Edge recognition: quantified by analyzing edge continuity and gradient direction consistency; Edge texture: multi-scale texture features and frequency domain energy distribution, distinguishing between regular arrangement of floating ice crystals and blurred texture caused by oil film coverage; Edge complexity: Quantifying contour irregularities based on fractal geometry and local curvature variations.

5. The oil spill monitoring method combined with image analysis and processing technology according to claim 1 is characterized in that: The method of establishing an oil-ice complex edge feature matching matrix includes: Based on the minimum characteristic floating ice image, screening out the floating ice distribution image that meets the second edge characteristic standard; At the same time, the floating ice distribution images that meet the third edge feature standard are screened out; Based on the first edge feature, the second edge feature and the third edge feature, edge matching is performed on the ice floe distribution image after the oil spill and the ice floe are combined; The second edge feature standard is: among the three features, at least one sub-feature threshold is strictly higher than the first edge feature construction standard, and the remaining sub-feature thresholds are not lower than the first edge feature standard; The third edge feature standard is: among the three features, at least one sub-feature threshold is strictly lower than the first edge feature construction standard, and the remaining sub-feature thresholds are not higher than the first edge feature standard; The construction method of the second edge feature and the third edge feature follows the fusion algorithm framework of the first edge feature, and adjusts the adjustment coefficient in the adaptive threshold calculation formula according to the sub-feature threshold difference , .

6. The oil spill monitoring method combined with image analysis and processing technology according to claim 5 is characterized in that: The edge matching of the floating ice distribution image after the oil spill and the floating ice are combined based on the first edge feature, the second edge feature, and the third edge feature includes: The first edge feature is divided by the second edge feature as a first ratio; The value of the third edge feature divided by the second edge feature is taken as the second ratio; The deviation between the first ratio and the second ratio is used as the edge matching amount of the floating ice distribution image after the oil spill and the floating ice are combined.

7. The oil spill monitoring method combined with image analysis and processing technology according to claim 6 is characterized in that: The above-mentioned construction of a cyclic model of the morphological evolution of the oil-ice complex by combining spatiotemporal sequence analysis includes: Based on the neighboring floating ice distribution images and edge matching information, N floating ice distribution images are obtained, and the floating ice distribution images are sorted from large to small or from small to large according to their sizes; Based on the sorting results, the image sequence that reaches the complete morphological evolution cycle is obtained, and the key phase change nodes in the image sequence process of the complete morphological evolution cycle of the oil-ice complex are verified using the dynamic verification formula; The complete morphological evolution cycle refers to the period of N1 consecutive time intervals between the oil-ice complex and the Within a certain period of time, at least one cycle of morphological change from fragmentation to aggregation or from aggregation to fragmentation is completed.

8. The oil spill monitoring method combined with image analysis and processing technology according to claim 7 is characterized in that: The dynamic verification formula is used to quantitatively verify the key phase change nodes in the cycle model, including: During each verification process, a simulation key phase change node is generated. Specifically, when the verification parameter reaches a preset threshold, an initial parameter set for the next phase change node is automatically generated.

9. The oil spill monitoring method combined with image analysis and processing technology according to claim 7 is characterized in that: The dynamic verification formula is used to quantitatively verify the key phase change nodes in the cycle model, including: the dynamic verification formula is as follows, ； in, To verify the value, is the rate of change of the projected area of ​​the oil-ice complex per unit time, is the ratio of the total edge length to the area of ​​the region, is the spatial variation intensity of the temperature field on the floating ice surface, To adjust the sensitivity of the temperature gradient, is the projected area of ​​the oil-ice complex in the image, is the time interval of the image sequence, is the total edge length of the oil-ice complex.

10. An oil spill monitoring system combined with image analysis and processing technology, used in an oil spill monitoring method combined with image analysis and processing technology as claimed in any one of claims 1 to 9, characterized in that: Included are: The image acquisition and modeling module uses multispectral remote sensing imaging equipment to obtain the oil spill diffusion path and ice floe distribution data, dynamically identifies the distribution state through the ice floe continuity algorithm, and automatically tracks the adjacent paths until the non-continuous ice floe image is captured as the modeling boundary; The feature extraction module uses adaptive pixel thresholds to screen ice floes images, integrates edge recognition, texture clarity and complexity three-dimensional analysis technology, builds a multi-level feature matching matrix to achieve dynamic comparison of oil-ice complexes, and quantifies the oil film-ice interaction effect through feature ratios; The dynamic modeling module builds a morphological evolution model based on the spatiotemporal sequence, and combines the temperature field gradient factor to present the periodic changes from ice fragmentation to aggregation, covering the evolution characteristics from small to large states; The verification and optimization module uses dynamic verification formulas to evaluate phase change nodes, triggers morphological reconstruction through area change rate thresholds, and automatically generates new phase change nodes and optimizes parameters when the prediction deviation exceeds the limit, forming a parameter self-optimization closed-loop system.

Images