Submersible environment perception and obstacle avoidance method and system

Abstract

The invention relates to the field of underwater vehicles, and provides a submersible environment perception and obstacle avoidance method and system, and the method comprises the following steps: collecting and preprocessing underwater environment original multi-modal data of a current scene, and obtaining underwater environment multi-modal data; the underwater environment multi-modal data comprises underwater environment background global light intensity, underwater environment background color channel light intensity, underwater foul image pixel values and underwater foul image color channel pixel values; calculating the light transmittance of the current scene based on the pixel value of the color channel of the underwater dirty image and the light intensity of the color channel of the background of the underwater environment; calculating an underwater clear image pixel value according to the current scene light transmittance, the underwater dirty image pixel value and the underwater environment background global light intensity; the current scene is restored through the underwater clear image pixel value, and an underwater clear image of the current scene is obtained; and generating an optimal obstacle avoidance path of the submersible in real time based on an adaptive path planning algorithm and the underwater clear image. According to the invention, the obstacle identification and avoidance capability of the submersible in a complex underwater environment is improved, and smooth completion of tasks and safety of equipment are ensured.

Description

Technical Field

[0001] The present invention relates to the field of underwater vehicles, and particularly to a method and system for environmental perception and obstacle avoidance of submersibles. Background Art

[0002] With the increasing demand for underwater exploration and work, the need for autonomous navigation and intelligent control of submersibles is becoming increasingly urgent. When a submersible works in a complex underwater environment, it needs to effectively identify and avoid various obstacles to ensure the smooth completion of tasks and the safety of equipment. However, due to the poor light conditions in the underwater environment and large underwater visual interference, traditional obstacle recognition and avoidance methods are difficult to meet the requirements.

[0003] Existing obstacle recognition and avoidance methods generally use sensors to receive information about detected obstacles, including obstacle positions and obstacle attributes, and then perform circuitous obstacle avoidance through an obstacle avoidance module. However, when there are many obstacles in a turbid underwater environment, the circuitous space is limited, and the submersible needs to have strong recognition ability in turbid water, otherwise it cannot clearly identify the state of underwater obstacles. Summary of the Invention

[0004] In view of this, the present invention proposes a method and system for environmental perception and obstacle avoidance of submersibles. By collecting and processing multi-modal data of the underwater environment, calculating the light transmittance and clear image pixel values, and using an adaptive path planning algorithm to generate an optimal obstacle avoidance path, the technical effect of real-time identifying and avoiding obstacles in a complex underwater environment is achieved, and the problem of insufficient recognition ability of traditional methods in a turbid underwater environment is solved.

[0005] The technical solution of the present invention is implemented as follows: In the first aspect, the present invention provides a method for environmental perception and obstacle avoidance of submersibles, including the following steps:

[0006] S1, collecting and preprocessing the original multi-modal data of the underwater environment of the current scene to obtain multi-modal data of the underwater environment, where the multi-modal data of the underwater environment includes the global light intensity of the underwater environment background, the light intensity of the color channels of the underwater environment background, the pixel values of the underwater turbid image, and the pixel values of the color channels of the underwater turbid image;

[0007] S2, calculating the light transmittance of the current scene based on the pixel values of the color channels of the underwater turbid image and the light intensity of the color channels of the underwater environment background;

[0008] S3, calculating the pixel values of the underwater clear image according to the light transmittance of the current scene, the pixel values of the underwater turbid image, and the global light intensity of the underwater environment background;

[0009] S4, restoring the current scene through the pixel values of the underwater clear image to obtain the underwater clear image of the current scene;

[0010] S5. Based on the adaptive path planning algorithm and the underwater clear image, generate the optimal obstacle avoidance path of the submersible in real time.

[0011] On the basis of the above technical solutions, preferably, step S2 includes:

[0012] The calculation formula for the light transmittance of the current scene is:

[0013]

[0014] where α(a) is the light transmittance at position a in the current scene, β is the light transmittance offset, Ω(a) is the local patch domain centered at position a, b is a random position in Ω(a), (r, g, b) are the red, green, and blue color channels, A c (b) is the pixel value of the underwater turbid image color channel at position b, G c is the light intensity of the underwater environment background color channel, c is any one of the red, green, and blue color channels, exp(·) is the natural exponential function, γ c is the attenuation coefficient corresponding to color channel c, and d(b) is the distance from position b to the submersible.

[0015] On the basis of the above technical solutions, preferably, step S3 includes:

[0016] The calculation formula for the pixel value of the underwater clear image is:

[0017]

[0018] where D(a) is the pixel value of the underwater clear image at position a, A(a) is the pixel value of the underwater turbid image at position a, G is the atmospheric light intensity, G' is the global light intensity of the underwater environment background, k c is the color adjustment parameter.

[0019] On the basis of the above technical solutions, preferably, step S4 includes:

[0020] Combine the pixel values of the underwater clear images at all positions of the current scene to obtain an underwater clear meta-image, perform global color adjustment and calibration on the underwater clear meta-image, and apply a contrast enhancement algorithm and Gaussian filtering to obtain an underwater clear image.

[0021] On the basis of the above technical solutions, preferably, step S5 includes:

[0022] S51. Use an image recognition algorithm to recognize the underwater clear image, obtain the position information and shape information of the obstacles in the underwater clear image, mark the target points of the submersible in the underwater clear image, mark the area where the obstacles are located in the underwater clear image as a risk area through a segmentation algorithm, and convert the underwater clear image marked with obstacle and target point information into a two-dimensional environmental map, where the two-dimensional environmental map includes free space, obstacles, risk areas, and target points;

[0023] S52. Collect dynamic and changing underwater environmental multimodal data in real time to update the underwater clear image and generate a dynamic two-dimensional environmental map;

[0024] S53. Generate a preliminary path from the submersible to the target point according to the two-dimensional environmental map and the path planning algorithm, smooth the preliminary path, and correct and optimize the path in real time based on the dynamic two-dimensional environmental map to obtain a dynamically optimized path;

[0025] S54. Evaluate the dynamically optimized path, including path length, safety, and energy consumption.

[0026] Based on the above technical solutions, preferably, step S5 further includes:

[0027] S55. Convert the dynamically optimized path into a motion control instruction for the submersible, monitor the execution of the submersible in real time, and when it is detected that the submersible path deviates, re-plan the path and update the motion control instruction.

[0028] Based on the above technical solutions, preferably, step S1 includes:

[0029] Collect the original underwater environmental multimodal data in real time through various sensors, and preprocess the original underwater environmental multimodal data, where the preprocessing includes image preprocessing, light intensity data calibration, and data synchronization and fusion;

[0030] The image preprocessing includes adjusting the color balance and histogram equalization of the original image to reduce the influence of uneven illumination, and using a denoising algorithm to remove the noise in the original image;

[0031] The light intensity data calibration includes calibrating and normalizing the ambient light intensity data to ensure that the ambient light intensity data and the pixel values of the optical image are in the same order of magnitude;

[0032] The data synchronization and fusion includes synchronizing multiple data to a unified time axis to ensure that all data is at the same timestamp.

[0033] In a second aspect, the present invention also provides a submersible environmental perception and obstacle avoidance system, where the system includes,

[0034] A data acquisition module, configured to acquire and preprocess the original multi-modal data of the underwater environment in the current scene to obtain the multi-modal data of the underwater environment, where the multi-modal data of the underwater environment includes the global light intensity of the underwater environment background, the light intensity of the color channels of the underwater environment background, the pixel values of the underwater turbid image, and the pixel values of the color channels of the underwater turbid image;

[0035] A light transmittance calculation module, configured to calculate the light transmittance of the current scene based on the pixel values of the color channels of the underwater turbid image and the light intensity of the color channels of the underwater environment background;

[0036] A pixel value calculation module, configured to calculate the pixel values of the underwater clear image according to the light transmittance of the current scene, the pixel values of the underwater turbid image, and the global light intensity of the underwater environment background;

[0037] An image restoration module, configured to restore the current scene through the pixel values of the underwater clear image to obtain the underwater clear image of the current scene;

[0038] A path generation module, configured to generate the optimal obstacle avoidance path of the submersible in real time based on the adaptive path planning algorithm and the underwater clear image.

[0039] In a third aspect, the present invention further provides an electronic device, which is characterized by including: at least one processor, at least one memory, a communication interface, and a bus;

[0040] Wherein, the processor, the memory, and the communication interface complete communication with each other through the bus, the memory stores program instructions executable by the processor, and the processor calls the program instructions to implement the steps of a method for submersible environment perception and obstacle avoidance.

[0041] In a fourth aspect, the present invention further provides a computer-readable storage medium, which is characterized in that the computer-readable storage medium stores computer instructions, and the computer instructions enable a computer to implement the steps of a method for submersible environment perception and obstacle avoidance.

[0042] The method and system for submersible environment perception and obstacle avoidance of the present invention have the following

[0043] Advantages over the prior art:

[0044] (1) By collecting and processing the multi-modal data of the underwater environment, calculating the light transmittance and the pixel values of the clear image, thereby generating the underwater clear image, improving the obstacle recognition and avoidance ability of the submersible in a complex underwater environment, and ensuring the smooth completion of the task and the safety of the equipment;

[0045] (2) Calculate the current scene light transmittance based on the pixel values of the underwater turbid image color channels and the light intensity of the underwater environment background color channels, accurately evaluate the light transmission situation of the underwater environment, and provide a basis for image restoration;

[0046] (3) Calculate the pixel values of the underwater clear image according to the current scene light transmittance, the pixel values of the underwater turbid image, and the global light intensity of the underwater environment background, generate a clearer underwater image, improve the accuracy of obstacle recognition, restore the current scene, obtain the underwater clear image of the current scene, restore the real image of the underwater environment, and improve the environmental perception ability of the submersible;

[0047] (4) Based on the adaptive path planning algorithm and the underwater clear image, generate the optimal obstacle avoidance path of the submersible in real time, convert the dynamic optimization path into the motion control instruction of the submersible, and monitor the execution situation of the submersible in real time. When it is detected that the path of the submersible deviates, re-plan the path and update the motion control instruction, realize the dynamic adjustment of the path planning of the submersible, improve the real-time performance and accuracy of obstacle avoidance, ensure that the submersible can correct the path deviation in time during the execution process, and ensure the smooth progress of the task. Description of the Drawings

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0049] Figure 1 It is a flowchart of a method for submersible environmental perception and obstacle avoidance of the present invention;

[0050] Figure 2 It is a structural diagram of a submersible environmental perception and obstacle avoidance system of the present invention. Detailed Embodiments

[0051] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0052] Please refer to Figure 1 , the present invention provides a method for submersible environmental perception and obstacle avoidance, including the following steps:

[0053] S1. Collect the original multi-modal data of the underwater environment in the current scene and perform preprocessing to obtain the multi-modal data of the underwater environment, where the multi-modal data of the underwater environment includes the global light intensity of the underwater environment background, the light intensity of the color channels of the underwater environment background, the pixel values of the underwater turbid image, and the pixel values of the color channels of the underwater turbid image;

[0054] S2. Calculate the light transmittance of the current scene based on the pixel values of the color channels of the underwater turbid image and the light intensity of the color channels of the underwater environment background;

[0055] S3. Calculate the pixel values of the underwater clear image according to the light transmittance of the current scene, the pixel values of the underwater turbid image, and the global light intensity of the underwater environment background;

[0056] S4. Restore the current scene through the pixel values of the underwater clear image to obtain the underwater clear image of the current scene;

[0057] S5. Based on the adaptive path planning algorithm and the underwater clear image, generate the optimal obstacle avoidance path of the submersible in real time.

[0058] Specifically, a method for submersible environmental perception and obstacle avoidance in this embodiment improves the obstacle recognition and avoidance capabilities of the submersible in complex underwater environments by collecting and processing multi-modal data of the underwater environment, calculating the light transmittance and the pixel values of the clear image, thereby ensuring the smooth completion of the task and the safety of the equipment.

[0059] Step S1 includes:

[0060] Collect the original multi-modal data of the underwater environment in real time through various sensors, and perform preprocessing on the original multi-modal data of the underwater environment. The preprocessing includes image preprocessing, light intensity data calibration, and data synchronization and fusion;

[0061] The image preprocessing includes adjusting the color balance and histogram equalization of the original image to reduce the influence of uneven illumination, and using a denoising algorithm to remove the noise in the original image;

[0062] The light intensity data calibration includes calibrating and normalizing the ambient light intensity data to ensure that the ambient light intensity data and the pixel values of the optical image are in the same magnitude;

[0063] The data synchronization and fusion includes synchronizing multiple data to a unified time axis to ensure that all data is at the same timestamp.

[0064] Specifically, step S1 of this embodiment ensures the comprehensiveness and real-time nature of the data by using multiple sensors to collect the original multi-modal data of the underwater environment in real time, providing a rich information basis for data processing and analysis, and enhancing the system's perception ability of the underwater environment.

[0065] By adjusting the color balance and histogram equalization, the image quality problems caused by uneven illumination are reduced, making the colors of the image more uniform and natural, which helps to improve the visual effect of the image.

[0066] Removing the noise in the image improves the clarity and quality of the image, reduces the interference of noise on subsequent image processing and analysis, and ensures higher image processing accuracy.

[0067] By calibrating and normalizing the ambient light intensity data, it is ensured that the light intensity data and the pixel values of the optical image are on the same order of magnitude. The magnitude differences between different data sources are eliminated, making the data more consistent and comparable, and providing a data basis for transmittance calculation and image restoration.

[0068] By synchronizing multiple data to a unified time axis, it is ensured that all data have the same timestamp. The problem of time asynchronization between different data sources is solved, enabling multi-modal data to be fused and analyzed at the same time point, and improving the accuracy and consistency of data processing.

[0069] Step S1 of this embodiment significantly improves the quality and consistency of underwater environment data through preprocessing steps, including image preprocessing, light intensity data calibration, and data synchronization and fusion. It ensures the accuracy and reliability of subsequent transmittance calculation and image restoration, thereby improving the perception and obstacle avoidance capabilities of the submersible in complex underwater environments.

[0070] Step S2 includes:

[0071] The calculation formula for the current scene transmittance is:

[0072]

[0073] where α(a) is the transmittance at position a in the current scene, β is the transmittance offset, Ω(a) is the local patch domain centered at position a, b is a random position in Ω(a), (r, g, b) are the red, green, and blue color channels, A c (b) is the pixel value of the underwater turbid image color channel at position b, G c is the light intensity of the underwater environment background color channel, c is any one of the red, green, and blue color channels, exp(·) is the natural exponential function, γ c is the attenuation coefficient corresponding to color channel c, and d(b) is the distance from position b to the submersible.

[0074] Specifically, in the underwater environment, the absorption and scattering of light will not only cause a decrease in contrast but also color distortion. Considering that the underwater light absorption and scattering are not uniform, this embodiment introduces a specific parameter γ of the underwater environment c , γ cis the attenuation coefficient corresponding to the color channel c, representing the absorption and scattering effects of underwater light. d(b) is the distance from position b to the submersible, which can be measured by sensors.

[0075] In this embodiment, the attenuation coefficient γ is added c , which is used to adapt to the different absorption and scattering degrees of different colors (such as red, green, and blue light in this embodiment) underwater.

[0076] By correcting the transmittance estimation through exp(-γ c ·d(b)), it can more accurately reflect the absorption and scattering of underwater ambient light. By considering the characteristics of underwater light propagation, the improved calculation formula for the current scene transmittance has higher accuracy.

[0077] By calculating the transmittance of position a in the current scene through the formula, the light transmission situation of the underwater environment can be accurately evaluated. Transmittance is a key parameter reflecting the light propagation characteristics of the underwater environment, and accurate transmittance calculation provides a necessary basis for image restoration.

[0078] By considering the attenuation coefficient of the color channel and the distance from position b to the submersible, this embodiment can adapt to complex underwater environments. The introduction of these parameters enables the transmittance calculation to cope with different underwater conditions, such as water turbidity and light attenuation, improving the robustness of the system in different environments.

[0079] Step S2 of this embodiment ensures the accuracy and reliability of the transmittance through a detailed transmittance calculation formula and parameter settings, provides a solid foundation for image restoration and path planning, and significantly improves the perception ability of the system in complex underwater environments.

[0080] Step S3 includes:

[0081] The calculation formula for the pixel value of the underwater clear image is:

[0082]

[0083] Among them, D(a) is the pixel value of the underwater clear image at position a, A(a) is the pixel value of the underwater turbid image at position a, G is the atmospheric light intensity, G' is the global light intensity of the underwater environment background, and k c is the color adjustment parameter.

[0084] Specifically, in step S3 of this embodiment, k c is the color adjustment parameter, which is used to correct color distortion; G' is the global light intensity of the underwater environment background, which is the adjusted background light color and is used to correct color distortion in the underwater environment.

[0085] In the underwater environment, color distortion is a significant problem. By introducing the color adjustment coefficient k cTogether with the adjusted background light G', it can effectively correct color distortion, making the colors of the restored image more natural and closer to the real scene.

[0086] The newly added color balance and adjustment steps can correct the common color drift and distortion problems in the underwater environment, making the colors of the restored image more vivid.

[0087] By calculating the pixel values of the underwater clear image in step S3, a clear image can be restored from the turbid underwater image. This improves the visibility and quality of the image, enabling the submersible to more accurately perceive and identify details and obstacles in the underwater environment.

[0088] Taking into account the pixel values of the turbid underwater image at position a, the atmospheric light intensity, the global light intensity of the underwater environment background, and the color adjustment parameters, it ensures the accuracy and authenticity of the restored image, and can better reflect the actual situation of the underwater environment.

[0089] By considering the atmospheric light intensity and the global light intensity of the underwater environment background, step S3 can effectively correct the image distortion caused by changes in lighting conditions, ensuring that the brightness and contrast of the restored image are more balanced and consistent, and improving the overall quality of the image.

[0090] The step S3 of this embodiment significantly improves the restoration effect of the underwater clear image through a detailed pixel value calculation formula and comprehensive consideration of multiple factors, provides high-quality image data for subsequent image analysis and path planning, and ensures the obstacle avoidance ability of the submersible in a complex underwater environment.

[0091] Step S4 includes:

[0092] Combining the pixel values of the underwater clear images at all positions in the current scene to obtain an underwater clear meta-image, performing global color adjustment and calibration on the underwater clear meta-image, and applying a contrast enhancement algorithm and Gaussian filtering to obtain an underwater clear image.

[0093] Specifically, in step S4 of this embodiment, by combining the pixel values of the underwater clear images at all positions in the current scene, a complete underwater clear meta-image is generated, ensuring that the image data of the entire scene is comprehensively integrated and providing a complete basis for the image processing in S5.

[0094] Performing global color adjustment and calibration on the underwater clear meta-image can unify the color performance of the image, eliminate the color inconsistency problems caused by different lighting conditions and sensor characteristics, improve the color restoration degree of the image, and make the image more natural and real.

[0095] By applying a contrast enhancement algorithm, the contrast of the image is enhanced, making the details in the image clearer and more prominent, which helps to enhance the visual effect of the image, improve the visibility of obstacles and environmental features, and thus enhance the perception ability of the submersible.

[0096] Gaussian filtering can effectively remove the noise in the image while retaining the edge details of the image, thereby improving the smoothness and clarity of the image, reducing the interference of noise on image analysis and recognition, and ensuring higher-quality image data.

[0097] Step S4 of this embodiment significantly improves the quality and clarity of the underwater clear image through detailed image processing steps, including global color adjustment, contrast enhancement, and Gaussian filtering, providing high-quality image data for the image analysis and path planning in S5.

[0098] Step S5 includes:

[0099] S51, Use an image recognition algorithm to recognize the underwater clear image to obtain the position information and shape information of the obstacles in the underwater clear image, mark the target point of the submersible in the underwater clear image, mark the area where the obstacles are located in the underwater clear image as a risk area through a segmentation algorithm, and convert the underwater clear image marked with obstacle and target point information into a two-dimensional environmental map, where the two-dimensional environmental map includes free space, obstacles, risk areas, and target points;

[0100] S52, Real-time collect multi-modal data of the dynamically changing underwater environment to update the underwater clear image and generate a dynamic two-dimensional environmental map;

[0101] S53, Generate a preliminary path for the submersible to the target point according to the two-dimensional environmental map and the path planning algorithm, smooth the preliminary path, and correct and optimize the path in real time based on the dynamic two-dimensional environmental map to obtain a dynamically optimized path;

[0102] S54, Evaluate the dynamically optimized path, including path length, safety, and energy consumption.

[0103] S55, Convert the dynamically optimized path into a motion control instruction for the submersible, and monitor the execution of the submersible in real time. When it is detected that the submersible path deviates, re-plan the path and update the motion control instruction.

[0104] Specifically, step S5 of this embodiment describes the detailed process of generating the optimal obstacle avoidance path of the submersible based on the underwater clear image.

[0105] First, by using an image recognition algorithm to identify underwater clear images, the position information and shape information of obstacles can be accurately obtained, ensuring that the submersible can identify potential obstacles in the underwater environment and mark the target points of the submersible in the image. The area where the obstacle is located is marked as a risk area through a segmentation algorithm, and a two-dimensional environmental map including free space, obstacles, risk areas, and target points is generated. The two-dimensional environmental map provides detailed environmental information for path planning, ensuring the accuracy and safety of path planning.

[0106] Second, real-time collect multi-modal data of the dynamically changing underwater environment to update the underwater clear image and generate a dynamic two-dimensional environmental map, ensuring that the submersible can respond to environmental changes in a timely manner and maintain an accurate perception of the current environment. The dynamically updated environmental map improves the real-time performance and adaptability of path planning, enabling the submersible to work stably in complex and changing underwater environments.

[0107] Third, generate a preliminary path from the submersible to the target point according to the two-dimensional environmental map and the path planning algorithm, and smooth the preliminary path to reduce sharp turns and unnecessary detours in the path, improving the smoothness and driving efficiency of the path.

[0108] Fourth, based on the dynamic two-dimensional environmental map, correct and optimize the path in real time to obtain a dynamically optimized path, thereby ensuring that path planning can adapt to the dynamic changes of the environment, and adjust the path in real time to avoid newly emerging obstacles and risk areas. The dynamically optimized path improves the safety and effectiveness of the path, ensuring that the submersible can reach the target point smoothly.

[0109] Finally, evaluate the dynamically optimized path, including path length, safety, and energy consumption, select the optimal path to balance path length, driving safety, and energy consumption, improve the rationality of path planning, ensure the efficient operation of the submersible, convert the dynamically optimized path into the motion control instructions of the submersible, and monitor the execution of the submersible in real time. When it is detected that the submersible path deviates, re-plan the path and update the motion control instructions to achieve timely correction of path deviation, ensure that the submersible travels according to the optimal path, and improve the safety of the system.

[0110] The steps S5 of this embodiment significantly improve the path planning and obstacle avoidance capabilities of the submersible through detailed obstacle recognition, dynamic environmental data update, path generation and optimization, path evaluation, and path execution and monitoring. It ensures that the submersible can navigate and perform tasks safely and efficiently in complex and dynamic underwater environments, and improves reliability.

[0111] Please refer to Figure 2 , the present invention also provides a submersible environmental perception and obstacle avoidance system, the system includes,

[0112] A data acquisition module, configured to acquire and preprocess the original underwater environmental multi-modal data of the current scene to obtain underwater environmental multi-modal data, where the underwater environmental multi-modal data includes the global light intensity of the underwater environmental background, the light intensity of the color channels of the underwater environmental background, the pixel values of the underwater turbid image, and the pixel values of the color channels of the underwater turbid image;

[0113] A light transmittance calculation module, configured to calculate the light transmittance of the current scene based on the pixel values of the color channels of the underwater turbid image and the light intensity of the color channels of the underwater environmental background;

[0114] A pixel value calculation module, configured to calculate the pixel values of the underwater clear image according to the light transmittance of the current scene, the pixel values of the underwater turbid image, and the global light intensity of the underwater environmental background;

[0115] An image restoration module, configured to restore the current scene through the pixel values of the underwater clear image to obtain the underwater clear image of the current scene;

[0116] A path generation module, configured to generate an optimal obstacle avoidance path for the submersible in real time based on the adaptive path planning algorithm and the underwater clear image.

[0117] Specifically, the submersible environmental perception and obstacle avoidance system of this embodiment significantly improves the environmental perception and obstacle avoidance capabilities of the submersible through efficient data acquisition and preprocessing, accurate light transmittance calculation, high-quality image restoration, and real-time path generation and optimization, ensures the robustness of the system, and improves the safety and efficiency of task execution.

[0118] The present invention also discloses an electronic device, including: at least one processor, at least one memory communication interface, and a bus: wherein, the processor, the memory, and the communication interface complete communication with each other through the bus; the memory stores program instructions executable by the processor, and the processor calls the program instructions to implement a method for submersible environmental perception and obstacle avoidance.

[0119] The present invention also discloses a computer-readable storage medium, where the computer-readable storage medium stores computer instructions, and the computer instructions enable the computer to implement all or part of the steps of the method for submersible environmental perception and obstacle avoidance described in the embodiments of the present invention. The storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory ROM, random access memory RAM, magnetic disks, or optical discs that can store program codes.

[0120] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for environmental perception and obstacle avoidance of a submersible, characterized in that, It includes the following steps: S1. Collect the original underwater environment multi-modal data of the current scene and perform preprocessing to obtain the underwater environment multi-modal data, where the underwater environment multi-modal data includes the global light intensity of the underwater environment background, the light intensity of the color channels of the underwater environment background, the pixel values of the underwater turbid image, and the pixel values of the color channels of the underwater turbid image; S2. Calculate the light transmittance of the current scene based on the pixel values of the color channels of the underwater turbid image and the light intensity of the color channels of the underwater environment background; S3. Calculate the pixel values of the underwater clear image according to the light transmittance of the current scene, the pixel values of the underwater turbid image, and the global light intensity of the underwater environment background; S4. Restore the current scene through the pixel values of the underwater clear image to obtain the underwater clear image of the current scene; S5. Based on the adaptive path planning algorithm and the underwater clear image, generate the optimal obstacle avoidance path of the submersible in real time.

2. The method for environmental perception and obstacle avoidance of a submersible according to claim 1, characterized in that, Step S2 includes: The calculation formula for the light transmittance of the current scene is: Among them, α(a) is the light transmittance at position a in the current scene, β is the light transmittance offset, Ω(a) is the local patch domain centered at position a, b is a random position in Ω(a), (r, g, b) are the red, green, and blue color channels, A c (b) is the pixel value of the underwater turbid image color channel at position b, G c is the light intensity of the underwater environment background color channel, c is any one of the red, green, and blue color channels, exp(·) is the natural exponential function, γ c is the attenuation coefficient corresponding to color channel c, and d(b) is the distance from position b to the submersible.

3. A method for environmental perception and obstacle avoidance of a submersible according to claim 2, characterized in that Step S3 includes: The calculation formula for the pixel values of the underwater clear image is: Among them, D(a) is the pixel value of the clear underwater image at position a, A(a) is the pixel value of the turbid underwater image at position a, G is the atmospheric light intensity, G' is the global light intensity of the underwater environmental background, and k c is the color adjustment parameter.

4. The method for environmental perception and obstacle avoidance of a submersible according to claim 3, characterized in that, Step S4 includes: Combine the pixel values of the underwater clear image at all positions of the current scene to obtain the underwater clear meta-image, perform global color adjustment and calibration on the underwater clear meta-image, and apply the contrast enhancement algorithm and Gaussian filtering to obtain the underwater clear image.

5. A method for environmental perception and obstacle avoidance of a submersible, as described in claim 4, characterized in that Step S5 includes: S51. Use the image recognition algorithm to recognize the underwater clear image, obtain the position information and shape information of the obstacles in the underwater clear image, mark the target point of the submersible in the underwater clear image, mark the area where the obstacles are located in the underwater clear image as the risk area through the segmentation algorithm, and convert the underwater clear image marked with obstacle and target point information into a two-dimensional environment map, where the two-dimensional environment map includes free space, obstacles, risk areas, and target points; S52. Collect the dynamically changing underwater environment multi-modal data in real time to update the underwater clear image and generate a dynamic two-dimensional environment map; S53. Generate the preliminary path of the submersible to the target point according to the two-dimensional environment map and the path planning algorithm, smooth the preliminary path, and correct and optimize the path in real time based on the dynamic two-dimensional environment map to obtain the dynamically optimized path; S54. Evaluate the dynamically optimized path, including path length, safety, and energy consumption.

6. The method for environmental perception and obstacle avoidance of a submersible according to claim 5, characterized in that, Step S5 further includes: S55. Convert the dynamically optimized path into the motion control instructions of the submersible, monitor the execution of the submersible in real time, and when it is detected that the path of the submersible deviates, re-plan the path and update the motion control instructions.

7. A method for environmental perception and obstacle avoidance of a submersible, as described in claim 1, characterized in that Step S1 includes: Collect the original underwater environment multi-modal data in real time through various sensors, and perform preprocessing on the original underwater environment multi-modal data. The preprocessing includes image preprocessing, light intensity data calibration, and data synchronization and fusion; The image preprocessing includes adjusting the color balance and histogram equalization of the original image to reduce the influence of uneven illumination, and using the denoising algorithm to remove the noise in the original image; The light intensity data calibration includes calibrating and normalizing the ambient light intensity data to ensure that the ambient light intensity data and the pixel values of the optical image are in the same order of magnitude; The data synchronization and fusion include synchronizing multiple types of data to a unified timeline to ensure that all data has the same timestamp.

8. An underwater vehicle environment perception and obstacle avoidance system, characterized in that, The system includes a data acquisition module, configured to acquire and preprocess the original multi-modal data of the underwater environment in the current scene to obtain multi-modal data of the underwater environment, where the multi-modal data of the underwater environment includes the global light intensity of the underwater environment background, the light intensity of the color channels of the underwater environment background, the pixel values of the underwater turbid image, and the pixel values of the color channels of the underwater turbid image; a light transmittance calculation module, configured to calculate the light transmittance of the current scene based on the pixel values of the color channels of the underwater turbid image and the light intensity of the color channels of the underwater environment background; a pixel value calculation module, configured to calculate the pixel values of the underwater clear image according to the light transmittance of the current scene, the pixel values of the underwater turbid image, and the global light intensity of the underwater environment background; an image restoration module, configured to restore the current scene through the pixel values of the underwater clear image to obtain the underwater clear image of the current scene; a path generation module, configured to generate an optimal obstacle avoidance path of the submersible in real time based on the adaptive path planning algorithm and the underwater clear image.

9. An electronic device, characterized in that, including: at least one processor, at least one memory, a communication interface, and a bus; wherein, the processor, the memory, and the communication interface complete communication with each other through the bus, the memory stores program instructions executable by the processor, and the processor invokes the program instructions to implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions, and the computer instructions cause the computer to implement the method according to any one of claims 1 to 7.

Images