High-resolution sea ice intelligent detection method and device, electronic equipment and storage medium

Abstract

The application provides a high-resolution sea ice intelligent detection method and device, electronic equipment and storage medium, and belongs to the technical field of image processing. In the method, a sea ice detection prompt word generated according to an initial sea ice detection result obtained by an optimization threshold algorithm is used to guide a SAM model to perform sea ice detection on a pretreated visible light remote sensing image, and then a target sea ice detection result is obtained. The above process greatly reduces the number of labeled samples, and can realize sea ice detection of the pretreated visible light remote sensing image by the SAM model in a zero sample condition.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and in particular to a high-resolution intelligent sea ice detection method, apparatus, electronic device, and storage medium. Background Technology

[0002] Traditional sea ice detection algorithms mainly use visible light and mid-infrared channels to calculate the NDSI (Normalized Difference Snow Index) threshold for extraction and identification.

[0003] For example, in MODIS data, snow-covered sea ice has similar reflectance to snow. The method for identifying snow-covered sea ice areas is used, defining the NDSI (Normalized Difference of Infrared) snow index: NDSI = (R4 - R6) / (R4 + R6). This requires using both visible and mid-infrared channels. Here, R4 and R6 are the reflectances of MODIS band 4 (0.58~0.68µm) and band 6 (1.55~1.64µm), respectively. For sea ice with snow cover, the snow index is used to distinguish between snow and ice. When NDSI > NDSITH, it is identified as a sea ice area. NDSITH is the threshold for distinguishing between snow and ice, a value that filters out sea ice, seawater, and fog; the NDSITH value ranges from -0.2 to 0.8.

[0004] Alternatively, there are image recognition methods based on deep learning, but traditional deep learning requires a large number of samples for training.

[0005] However, both of these algorithms have some limitations in current applications. Common meteorological remote sensing satellites such as MODIS and the Fengyun series have multiple visible and mid-infrared channels at 1000M resolution, facilitating sea ice identification. However, at 250M high resolution, only visible light channels exist, making effective extraction using the NDSI algorithm impossible. Furthermore, deep learning-based image recognition methods require a large number of labeled samples, which are often difficult to obtain in practical applications.

[0006] In summary, how to achieve high-resolution intelligent sea ice detection with zero samples has become an urgent technical problem to be solved. Summary of the Invention

[0007] In view of this, the purpose of the present invention is to provide a high-resolution intelligent sea ice detection method, apparatus, electronic device and storage medium to alleviate the technical problem that traditional technologies cannot achieve high-resolution intelligent sea ice detection in the case of zero samples.

[0008] In a first aspect, embodiments of the present invention provide a high-resolution intelligent sea ice detection method, comprising:

[0009] Acquire high-resolution visible light remote sensing images and preprocess the visible light remote sensing images to obtain preprocessed visible light remote sensing images;

[0010] An optimized thresholding algorithm was used to detect sea ice in the preprocessed visible light remote sensing image to obtain the initial sea ice detection results.

[0011] Generate sea ice detection prompts based on the initial sea ice detection results;

[0012] Using the SAM model adapted to high-resolution images, sea ice detection is performed on the preprocessed visible light remote sensing image according to the sea ice detection prompts to obtain intermediate sea ice detection results.

[0013] The intermediate sea ice detection results are post-processed to obtain the target sea ice detection results corresponding to the visible light remote sensing image.

[0014] Furthermore, the preprocessing includes noise reduction and enhancement processing.

[0015] Furthermore, an optimized thresholding algorithm is used to perform sea ice detection on the preprocessed visible light remote sensing image, including:

[0016] The formula is calculated based on the ratio of the labeled pixel samples to the reflectance in the visible light band. Calculate the visible light reflectance ratio of the labeled pixel samples, where, This represents the ratio of visible light reflectance of the labeled pixel sample. This represents the blue light reflectance of the labeled pixel sample. This represents the green band reflectance of the labeled pixel sample;

[0017] The formula is calculated based on the labeled pixel samples and morphological gradient features. Calculate the morphological gradient features of the labeled pixel samples, where, This represents the morphological gradient features of the labeled pixel samples. This represents the grayscale value of the labeled pixel sample. This represents the grayscale gradient in the X direction of the labeled pixel sample. This represents the gray-level gradient in the Y direction of the labeled pixel sample;

[0018] The sea ice detection threshold is calculated based on the visible light band reflectance ratio of the labeled pixel samples and the morphological gradient features of the labeled pixel samples.

[0019] The comprehensive feature value of each pixel in the preprocessed visible light remote sensing image is calculated, and whether each pixel is sea ice is determined based on the comprehensive feature value of each pixel and the sea ice detection threshold, thereby obtaining the initial sea ice detection result.

[0020] Further, the comprehensive feature value of each pixel in the preprocessed visible light remote sensing image is calculated, including:

[0021] Calculate the visible light band reflectance ratio and morphological gradient features of each pixel;

[0022] The comprehensive feature value of each pixel is calculated based on the visible light band reflectance ratio and the morphological gradient features of each pixel.

[0023] Furthermore, based on the initial sea ice detection results, sea ice detection prompts are generated, including:

[0024] The initial sea ice detection results are converted into geometric coordinates or binary masks;

[0025] The geometric coordinates or the binary mask are used as prompts for sea ice detection to indicate the segmentation location of the SAM model.

[0026] Furthermore, the SAM model includes: an image encoder, a prompt information encoder, and a segmentation mask decoder, wherein the segmentation mask decoder is connected to the image encoder and the prompt information encoder.

[0027] Furthermore, the post-processing includes noise reduction and edge smoothing.

[0028] Secondly, embodiments of the present invention also provide a high-resolution intelligent sea ice detection device, comprising:

[0029] An acquisition and preprocessing unit is used to acquire high-resolution visible light remote sensing images and preprocess the visible light remote sensing images to obtain preprocessed visible light remote sensing images.

[0030] The first sea ice detection unit is used to perform sea ice detection on the preprocessed visible light remote sensing image using an optimized threshold algorithm to obtain an initial sea ice detection result.

[0031] The generation unit is used to generate sea ice detection prompt words based on the initial sea ice detection results;

[0032] The second sea ice detection unit is used to perform sea ice detection on the preprocessed visible light remote sensing image according to the sea ice detection prompt words using a SAM model adapted to high-resolution images, and to obtain intermediate sea ice detection results.

[0033] The post-processing unit is used to post-process the intermediate sea ice detection results to obtain the target sea ice detection results corresponding to the visible light remote sensing image.

[0034] Thirdly, embodiments of the present invention also provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the method described in any of the first aspects above.

[0035] Fourthly, embodiments of the present invention also provide a computer-readable storage medium storing machine-executable instructions, which, when invoked and executed by a processor, cause the processor to perform the method described in any of the first aspects above.

[0036] In this embodiment of the invention, a high-resolution intelligent sea ice detection method is provided, comprising: acquiring a high-resolution visible light remote sensing image and preprocessing the visible light remote sensing image to obtain a preprocessed visible light remote sensing image; performing sea ice detection on the preprocessed visible light remote sensing image using an optimized threshold algorithm to obtain an initial sea ice detection result; generating sea ice detection prompts based on the initial sea ice detection result; using a SAM model adapted to the high-resolution image to perform sea ice detection on the preprocessed visible light remote sensing image according to the sea ice detection prompts to obtain an intermediate sea ice detection result; and performing post-processing on the intermediate sea ice detection result to obtain a target sea ice detection result corresponding to the visible light remote sensing image. As described above, in the high-resolution intelligent sea ice detection method of the present invention, the sea ice detection prompt words generated based on the initial sea ice detection results obtained by the optimized threshold algorithm guide the SAM model to perform sea ice detection on the preprocessed visible light remote sensing image, thereby obtaining the target sea ice detection result. The above process greatly reduces the number of labeled samples, enabling the SAM model to perform sea ice detection on the preprocessed visible light remote sensing image with zero samples, thus alleviating the technical problem that traditional technologies cannot achieve high-resolution intelligent sea ice detection with zero samples. Attached Figure Description

[0037] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0038] Figure 1 A flowchart illustrating a high-resolution intelligent sea ice detection method provided in an embodiment of the present invention;

[0039] Figure 2 A schematic diagram of a high-resolution intelligent sea ice detection device provided in an embodiment of the present invention;

[0040] Figure 3 This is a schematic diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0041] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0042] Traditional technologies cannot achieve high-resolution intelligent detection of sea ice with zero samples.

[0043] Based on this, in the high-resolution intelligent sea ice detection method of the present invention, the sea ice detection prompt words generated by the initial sea ice detection results obtained by the optimized threshold algorithm guide the SAM model to perform sea ice detection on the preprocessed visible light remote sensing image, thereby obtaining the target sea ice detection result. The above process greatly reduces the number of labeled samples, and enables the SAM model to detect sea ice in the preprocessed visible light remote sensing image with zero samples.

[0044] To facilitate understanding of this embodiment, a high-resolution intelligent sea ice detection method disclosed in this embodiment of the invention will first be described in detail.

[0045] Example 1:

[0046] According to an embodiment of the present invention, an embodiment of a high-resolution intelligent sea ice detection method is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0047] Figure 1 This is a flowchart of a high-resolution intelligent sea ice detection method according to an embodiment of the present invention, such as... Figure 1 As shown, the method includes the following steps:

[0048] Step S102: Acquire a high-resolution visible light remote sensing image and preprocess the visible light remote sensing image to obtain a preprocessed visible light remote sensing image.

[0049] Specifically, the aforementioned high-resolution visible light remote sensing images can be 250M resolution visible light remote sensing images acquired by meteorological remote sensing satellites.

[0050] Step S104: Use an optimized threshold algorithm to perform sea ice detection on the preprocessed visible light remote sensing image to obtain the initial sea ice detection result;

[0051] Step S106: Generate sea ice detection prompts based on the initial sea ice detection results;

[0052] Step S108: Use the SAM model adapted to the high-resolution image to perform sea ice detection on the preprocessed visible light remote sensing image according to the sea ice detection prompt words to obtain the intermediate sea ice detection result.

[0053] Specifically, SAM (Segment Anything Model) is a deep learning-based image segmentation model with zero-shot transfer capability, allowing it to perform segmentation inference directly without specific scene training data. Cue words are input information that guides the SAM model to focus on a specific region or target; these can take the form of point coordinates, bounding boxes, text descriptions, or image masks. Through these cue words, the model can clearly define the segmentation target, improving its targeting and accuracy.

[0054] Step S110: Post-process the intermediate sea ice detection results to obtain the target sea ice detection results corresponding to the visible light remote sensing image.

[0055] In this embodiment of the invention, a high-resolution intelligent sea ice detection method is provided, comprising: acquiring a high-resolution visible light remote sensing image and preprocessing the visible light remote sensing image to obtain a preprocessed visible light remote sensing image; performing sea ice detection on the preprocessed visible light remote sensing image using an optimized threshold algorithm to obtain an initial sea ice detection result; generating sea ice detection prompts based on the initial sea ice detection result; using a SAM model adapted to the high-resolution image to perform sea ice detection on the preprocessed visible light remote sensing image according to the sea ice detection prompts to obtain an intermediate sea ice detection result; and performing post-processing on the intermediate sea ice detection result to obtain a target sea ice detection result corresponding to the visible light remote sensing image. As described above, in the high-resolution intelligent sea ice detection method of the present invention, the sea ice detection prompt words generated based on the initial sea ice detection results obtained by the optimized threshold algorithm guide the SAM model to perform sea ice detection on the preprocessed visible light remote sensing image, thereby obtaining the target sea ice detection result. The above process greatly reduces the number of labeled samples, enabling the SAM model to perform sea ice detection on the preprocessed visible light remote sensing image with zero samples, thus alleviating the technical problem that traditional technologies cannot achieve high-resolution intelligent sea ice detection with zero samples.

[0056] The above provides a brief overview of the high-resolution intelligent sea ice detection method of the present invention. The specific details involved are described in detail below.

[0057] In an optional embodiment of the present invention, the preprocessing includes: noise reduction processing and enhancement processing.

[0058] The above preprocessing steps are used to improve the accuracy of subsequent processing.

[0059] In an optional embodiment of the present invention, an optimized thresholding algorithm is used to detect sea ice in the preprocessed visible light remote sensing image, specifically including the following steps:

[0060] (1) Calculate the formula based on the labeled pixel samples and the ratio of reflectance in the visible light band. Calculate the visible light reflectance ratio of the labeled pixel samples, where, This represents the ratio of visible light reflectance of the labeled pixel sample. This represents the blue light reflectance of the labeled pixel sample. This represents the green band reflectance of the labeled pixel sample;

[0061] Specifically, the pixel samples labeled above (labeling refers to the type of sea ice and non-sea ice) are a set of typical pixel samples used to calculate the sea ice detection threshold T. These pixels need to be manually labeled in advance or identified through other reliable methods as a benchmark for threshold calculation.

[0062] (2) Calculate the formula based on the labeled pixel samples and morphological gradient features. Calculate the morphological gradient features of the labeled pixel samples, where, This represents the morphological gradient features of the labeled pixel samples. This represents the grayscale value of the labeled pixel sample. This represents the grayscale gradient in the X direction of the labeled pixel sample. This represents the gray-level gradient in the Y direction of the labeled pixel sample;

[0063] (3) Calculate the sea ice detection threshold based on the visible light band reflectance ratio of the labeled pixel samples and the morphological gradient features of the labeled pixel samples;

[0064] Specifically, the formula for calculating the sea ice detection threshold is as follows: ,in, Indicates the sea ice detection threshold. This indicates the weight corresponding to the ratio of reflectivity in the visible light band. This represents the ratio of visible light reflectance of the labeled pixel sample. This represents the morphological gradient features of the labeled pixel samples.

[0065] In this invention, the traditional thresholding algorithm is optimized by combining the spectral characteristics and morphological features of sea ice to generate a sea ice detection threshold suitable for 250M resolution.

[0066] (4) Calculate the comprehensive feature value of each pixel in the preprocessed visible light remote sensing image, and determine whether each pixel is sea ice based on the comprehensive feature value of each pixel and the sea ice detection threshold, so as to obtain the initial sea ice detection result.

[0067] Specifically, calculating the comprehensive feature value of each pixel in the preprocessed visible light remote sensing image includes the following steps:

[0068] (41) Calculate the visible light band reflectance ratio and morphological gradient features of each pixel;

[0069] The process of calculating the visible light band reflectance ratio of each pixel is similar to the process of calculating the visible light band reflectance ratio of the labeled pixel sample, and will not be repeated here. Similarly, the process of calculating the morphological gradient features of each pixel is similar to the process of calculating the morphological gradient features of the labeled pixel sample, and will not be repeated here.

[0070] (42) Calculate the comprehensive feature value of each pixel based on the visible light band reflectance ratio and the morphological gradient features of each pixel.

[0071] The process of calculating the comprehensive feature value of each pixel is similar to the process of calculating the sea ice detection threshold, and it is also obtained by weighting, so it will not be described in detail here.

[0072] In an optional embodiment of the present invention, generating sea ice detection prompts based on initial sea ice detection results specifically includes the following steps:

[0073] (1) Convert the initial sea ice detection results into geometric coordinates or binary masks;

[0074] (2) Use geometric coordinates or binary masks as sea ice detection prompts to indicate the segmentation location of the SAM model.

[0075] Specifically, sea ice contours or masks are extracted from the initial sea ice detection results and converted into geometric coordinates or binary masks that the SAM model can recognize. The geometric coordinates or binary masks are used as sea ice detection prompts to inform the SAM model of the segmentation location and guide it to optimize sea ice edge details by combining image features (i.e., guide the model to perform more refined sea ice edge segmentation and extraction).

[0076] In an optional embodiment of the present invention, post-processing includes: noise reduction and edge smoothing.

[0077] Specifically, noise removal and edge smoothing are performed on the intermediate sea ice detection results. Parameters may be adjusted to suit the 250M high resolution characteristics to ensure high accuracy of the output target sea ice detection results and adapt to the needs of high-resolution scenarios.

[0078] In an optional embodiment of the present invention, the SAM model includes: an image encoder, a cue message encoder, and a segmentation mask decoder, wherein the segmentation mask decoder is connected to the image encoder and the cue message encoder.

[0079] Specifically, the input size of the image encoder was adjusted to adapt to 250M resolution images, and the receptive field of the feature extraction layer was optimized to enhance the ability to capture sea ice details (such as texture and edges) at high resolutions. A threshold algorithm was optimized based on sea ice spectral and morphological features to generate more accurate prompts, improve the ability to describe sea ice boundaries, and guide the SAM model to focus more on sea ice regions.

[0080] Image encoder: Adopting the Vision Transformer structure, the image is divided into blocks and mapped to embedding vectors. Global features are extracted through a multi-layer Transformer encoder to capture the semantic relationships and spatial location information of objects in the image.

[0081] The cue information encoder converts geometric cues (points, boxes) into coordinate embeddings; if the cues are textual, a vector is generated using a text encoder; and then fused with image features to clearly define the segmentation target.

[0082] Segmentation mask decoder: Based on the Transformer decoder, it combines global image features with local information from prompt words to generate a refined segmentation mask, optimizing edge details and target integrity.

[0083] The high-resolution intelligent sea ice detection method of the present invention has the following advantages:

[0084] (1) Zero-shot requirement: Compared with deep learning algorithms, the number of labeled samples required by this invention is greatly reduced, thus lowering the cost of data acquisition and processing. As the first basic model dedicated to image segmentation, SAM has demonstrated powerful zero-shot transfer capabilities. This means that it can perform segmentation inference directly without any training data from actual use cases, greatly expanding its application scope. Combined with the prompt word generation technology of the SAM large model, high-precision sea ice extraction can be achieved;

[0085] (2) High-precision extraction: The SAM model adopts a Transformer-based architecture, which includes three main parts: an image encoder, a cue information encoder, and a segmentation mask decoder. This architecture enables the model to efficiently process image data and generate high-quality segmentation results;

[0086] (3) High resolution support: It is specially optimized for 250M high resolution channels, filling the gap of traditional NDSI algorithm at high resolution.

[0087] The operational sea ice monitoring algorithm of this invention is expected to achieve high-precision and high-efficiency sea ice extraction at a high resolution of 250M with few samples (SAM is the basic segmentation model, which has zero-sample transfer capability after large-scale pre-training, and does not require a large number of sea ice labeled samples. This invention combines an optimized threshold algorithm to generate prompt words to guide segmentation, greatly reducing data dependence and enabling applications with few samples), providing a new technical means for sea ice monitoring by meteorological remote sensing satellites and improving the accuracy and timeliness of sea ice monitoring.

[0088] Example 2:

[0089] This invention also provides a high-resolution intelligent sea ice detection device, which is mainly used to execute the high-resolution intelligent sea ice detection method provided in Embodiment 1 of this invention. The high-resolution intelligent sea ice detection device provided in this invention will be described in detail below.

[0090] Figure 2 This is a schematic diagram of a high-resolution intelligent sea ice detection device according to an embodiment of the present invention, as shown below. Figure 2 As shown, the device mainly includes: an acquisition and preprocessing unit 10, a first sea ice detection unit 20, a generation unit 30, a second sea ice detection unit 40, and a post-processing unit 50, wherein:

[0091] The acquisition and preprocessing unit is used to acquire high-resolution visible light remote sensing images and preprocess the visible light remote sensing images to obtain preprocessed visible light remote sensing images.

[0092] The first sea ice detection unit is used to perform sea ice detection on the preprocessed visible light remote sensing image using an optimized threshold algorithm to obtain the initial sea ice detection result.

[0093] The generation unit is used to generate sea ice detection prompts based on the initial sea ice detection results;

[0094] The second sea ice detection unit is used to perform sea ice detection on the preprocessed visible light remote sensing image according to the sea ice detection prompt words using the SAM model adapted to the high-resolution image, and obtain intermediate sea ice detection results.

[0095] The post-processing unit is used to post-process the intermediate sea ice detection results to obtain the target sea ice detection results corresponding to the visible light remote sensing image.

[0096] In this embodiment of the invention, a high-resolution intelligent sea ice detection device is provided, comprising: acquiring a high-resolution visible light remote sensing image and preprocessing the visible light remote sensing image to obtain a preprocessed visible light remote sensing image; performing sea ice detection on the preprocessed visible light remote sensing image using an optimized threshold algorithm to obtain an initial sea ice detection result; generating sea ice detection prompts based on the initial sea ice detection result; using a SAM model adapted to the high-resolution image to perform sea ice detection on the preprocessed visible light remote sensing image according to the sea ice detection prompts to obtain an intermediate sea ice detection result; and performing post-processing on the intermediate sea ice detection result to obtain a target sea ice detection result corresponding to the visible light remote sensing image. As described above, in the high-resolution intelligent sea ice detection device of the present invention, the sea ice detection prompt words generated based on the initial sea ice detection results obtained by the optimized threshold algorithm guide the SAM model to perform sea ice detection on the preprocessed visible light remote sensing image, thereby obtaining the target sea ice detection result. The above process greatly reduces the number of labeled samples, enabling the SAM model to perform sea ice detection on the preprocessed visible light remote sensing image with zero samples, thus alleviating the technical problem that traditional technologies cannot achieve high-resolution intelligent sea ice detection with zero samples.

[0097] Optionally, preprocessing includes noise reduction and enhancement.

[0098] Optionally, the first sea ice detection unit is also used to: calculate a formula based on the ratio of labeled pixel samples to visible light band reflectance. Calculate the visible light reflectance ratio of the labeled pixel samples, where, This represents the ratio of visible light reflectance of the labeled pixel sample. This represents the blue light reflectance of the labeled pixel sample. This represents the green band reflectance of the labeled pixel sample; the formula is calculated based on the labeled pixel sample and morphological gradient features. Calculate the morphological gradient features of the labeled pixel samples, where, This represents the morphological gradient features of the labeled pixel samples. This represents the grayscale value of the labeled pixel sample. This represents the grayscale gradient in the X direction of the labeled pixel sample. The Y-axis grayscale gradient of the labeled pixel sample is represented; the sea ice detection threshold is calculated based on the visible light band reflectance ratio and morphological gradient features of the labeled pixel sample; the comprehensive feature value of each pixel in the preprocessed visible light remote sensing image is calculated, and whether each pixel is sea ice is determined based on the comprehensive feature value of each pixel and the sea ice detection threshold, thus obtaining the initial sea ice detection result.

[0099] Optionally, the first sea ice detection unit is also used to: calculate the visible light band reflectance ratio of each pixel and the morphological gradient features of each pixel; and calculate the comprehensive feature value of each pixel based on the visible light band reflectance ratio and the morphological gradient features of each pixel.

[0100] Optionally, the generation unit is also used to: convert the initial sea ice detection results into geometric coordinates or binary masks; and use the geometric coordinates or binary masks as sea ice detection cue words to indicate the segmentation location of the SAM model.

[0101] Optionally, the SAM model includes: an image encoder, a cue message encoder, and a segmentation mask decoder, with the segmentation mask decoder connected to the image encoder and the cue message encoder.

[0102] Optionally, post-processing includes noise reduction and edge smoothing.

[0103] The device provided in this embodiment of the invention has the same implementation principle and technical effect as the aforementioned method embodiment. For the sake of brevity, any parts not mentioned in the device embodiment can be referred to the corresponding content in the aforementioned method embodiment.

[0104] like Figure 3 As shown in the embodiment of this application, an electronic device 600 includes a processor 601, a memory 602, and a bus. The memory 602 stores machine-readable instructions executable by the processor 601. When the electronic device is running, the processor 601 communicates with the memory 602 via the bus, and the processor 601 executes the machine-readable instructions to perform the steps of the control method of the braking system described above.

[0105] Specifically, the memory 602 and processor 601 can be general-purpose memory and processor, without any specific limitations. When the processor 601 runs the computer program stored in the memory 602, it can execute the control method of the braking system described above.

[0106] The processor 601 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of the processor 601 or by instructions in software form. The processor 601 may be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly manifested as execution by a hardware decoding processor, or execution by a combination of hardware and software modules in the decoding processor. The software module can reside in a mature storage medium in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory 602, and processor 601 reads the information from memory 602 and, in conjunction with its hardware, completes the steps of the above method.

[0107] Corresponding to the control method of the above-described braking system, this application embodiment also provides a computer-readable storage medium storing machine-executable instructions. When the machine-executable instructions are invoked and executed by a processor, the machine-executable instructions cause the processor to perform the steps of the control method of the above-described braking system.

[0108] The control device for the braking system provided in this application embodiment can be specific hardware on the device or software or firmware installed on the device. The implementation principle and technical effects of the device provided in this application embodiment are the same as those in the foregoing method embodiments. For the sake of brevity, any parts not mentioned in the device embodiment can be referred to the corresponding content in the foregoing method embodiments. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can all be referred to the corresponding processes in the above method embodiments, and will not be repeated here.

[0109] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the displayed or discussed mutual couplings, direct couplings, or communication connections may be through some communication interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms.

[0110] For example, the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram and / or flowchart, and combinations of blocks in block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0111] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0112] In addition, the functional units in the embodiments provided in this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0113] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause an electronic device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the control method of the braking system described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0114] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. In addition, the terms "first", "second", "third", etc. are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0115] Finally, it should be noted that the above-described embodiments are merely specific implementations of this application, used to illustrate the technical solutions of this application, and not to limit them. The protection scope of this application is not limited thereto. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some of the technical features, within the scope of the technology disclosed in this application; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application. All should be covered within the protection scope of this application. Therefore, the protection scope of this application should be determined by the protection scope of the claims.

Claims

1. A high-resolution sea ice intelligent detection method, characterized in that, The method comprises the following steps: obtaining a high-resolution visible light remote sensing image, and preprocessing the visible light remote sensing image to obtain a preprocessed visible light remote sensing image; performing sea ice detection on the preprocessed visible light remote sensing image by using an optimized threshold algorithm to obtain an initial sea ice detection result; generating a sea ice detection prompt word according to the initial sea ice detection result; performing sea ice detection on the preprocessed visible light remote sensing image according to the sea ice detection prompt word by using a SAM model adapted to high-resolution images to obtain an intermediate sea ice detection result; performing post-processing on the intermediate sea ice detection result to obtain a target sea ice detection result corresponding to the visible light remote sensing image; wherein performing sea ice detection on the preprocessed visible light remote sensing image by using the optimized threshold algorithm comprises: According to the annotated pixel samples and the visible band reflectance ratio calculation formula calculating the visible band reflectance ratio of the annotated pixel samples, wherein, the visible band reflectance ratio of the annotated pixel samples is represented as, the blue light band reflectance of the annotated pixel samples is represented as, the green light band reflectance of the annotated pixel samples is represented as, calculating morphological gradient features of the annotated pixel samples based on the annotated pixel samples and the morphological gradient features calculating morphological gradient features of the annotated pixel samples, representing morphological gradient features of the annotated pixel samples, representing a gray scale of the annotated pixel sample, representing an X-direction gray scale gradient of the annotated pixel sample, representing a Y-direction gray scale gradient of the annotated pixel sample; calculating a sea ice detection threshold according to the visible light band reflectance ratio of the labeled pixel sample and the morphological gradient feature of the labeled pixel sample; calculating the comprehensive feature value of each pixel point in the preprocessed visible light remote sensing image, and determining whether each pixel point is sea ice according to the comprehensive feature value of each pixel point and the sea ice detection threshold, thereby obtaining the initial sea ice detection result.

2. The method of claim 1, wherein, The preprocessing comprises denoising processing and enhancement processing.

3. The method of claim 1, wherein, Calculating the comprehensive feature value of each pixel point in the preprocessed visible light remote sensing image comprises: calculating the visible light band reflectance ratio of each pixel point and the morphological gradient feature of each pixel point; calculating the comprehensive feature value of each pixel point according to the visible light band reflectance ratio of each pixel point and the morphological gradient feature of each pixel point.

4. The method of claim 1, wherein, Generating a sea ice detection prompt word according to the initial sea ice detection result comprises: converting the initial sea ice detection result into geometric coordinates or a binary mask; using the geometric coordinates or the binary mask as the sea ice detection prompt word to prompt the SAM model to segment the position.

5. The method of claim 1, wherein, The SAM model comprises a picture encoder, a prompt information encoder, and a segmentation mask decoder, wherein the segmentation mask decoder is connected with the picture encoder and the prompt information encoder.

6. The method of claim 1, wherein, The post-processing comprises denoising processing and edge smoothing processing.

7. A high-resolution sea ice intelligent detection device, characterized in that, The method comprises the following steps: an acquisition and preprocessing unit is configured to obtain a high-resolution visible light remote sensing image, and preprocess the visible light remote sensing image to obtain a preprocessed visible light remote sensing image; a first sea ice detection unit is configured to perform sea ice detection on the preprocessed visible light remote sensing image by using an optimized threshold algorithm to obtain an initial sea ice detection result; a generation unit is configured to generate a sea ice detection prompt word according to the initial sea ice detection result; a second sea ice detection unit is configured to perform sea ice detection on the preprocessed visible light remote sensing image according to the sea ice detection prompt word by using a SAM model adapted to high-resolution images to obtain an intermediate sea ice detection result; a post-processing unit is configured to perform post-processing on the intermediate sea ice detection result to obtain a target sea ice detection result corresponding to the visible light remote sensing image. The first sea ice detection unit is further configured to calculate a visible band reflectivity ratio of the labeled pixel sample according to a calculation formula The visible band reflectivity ratio of the labeled pixel sample is calculated, wherein The visible band reflectivity ratio of the labeled pixel sample is represented as The blue band reflectivity of the labeled pixel sample is represented as The green band reflectivity of the labeled pixel sample is represented as The morphological gradient feature of the labeled pixel sample is calculated, wherein The morphological gradient feature of the labeled pixel sample is represented as The gray scale of the labeled pixel sample is represented as The X-direction gray scale gradient of the labeled pixel sample is represented as The Y-direction gray scale gradient of the labeled pixel sample is represented as The sea ice detection threshold is calculated according to the visible band reflectivity ratio of the labeled pixel sample and the morphological gradient feature of the labeled pixel sample; the comprehensive feature value of each pixel point in the preprocessed visible light remote sensing image is calculated, and whether each pixel point is sea ice is determined according to the comprehensive feature value of each pixel point and the sea ice detection threshold, thereby obtaining the initial sea ice detection result.

8. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the computer program to realize the steps of the method in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer readable storage medium stores machine executable instructions which, when invoked and run by the processor, cause the processor to run the method of any one of claims 1 to 6.

Images