A method and system for evaluating deep-sea polymetallic nodule resources

Abstract

The present application relates to the technical field of deep-sea polymetallic nodule resource evaluation method and system, and provides a deep-sea polymetallic nodule resource evaluation method and system, which effectively solves the technical problem that the overlapping area between the polymetallic nodules is easy to be covered by silt and is located at the edge of the polymetallic nodules, is integrated with the seabed background, and cannot be completely filled by the inflation algorithm, resulting in low accuracy of the final calculation coverage.

Description

Technical Field

[0001] This invention relates to the technical field of assessment methods and systems for deep-sea polymetallic nodule resources, specifically, to an assessment method and system for deep-sea polymetallic nodule resources. Background Technology

[0002] Polymetallic nodule resource parameters include abundance, coverage, and grade. The polymetallic nodule coverage is the percentage of the area occupied by polymetallic nodules per unit area, which is one of the important parameters for evaluating polymetallic nodule resources. To explore polymetallic nodule resources on the seabed, AUVs are generally used to take optical images at depth on the seabed. The optical images are then processed and analyzed to determine the reserves and distribution of polymetallic nodule resources in a specific area.

[0003] Due to issues such as blurry images, color distortion, and inaccurate actual coverage rates caused by seabed sediment covering deep-sea polymetallic nodules captured by optical cameras, invention patent CN118505522A discloses an image processing-based method for assessing deep-sea polymetallic nodule resources. This method employs a binarized morphological dilation algorithm to compensate for the impact of seabed sediment cover on the polymetallic nodule coverage rate.

[0004] However, in the existing technology, the binary morphological dilation algorithm can only correct the part of the polymetallic nodule surface covered by mud and sand. In areas where polymetallic nodules are densely distributed, there is often a phenomenon of polymetallic nodules in close contact or even overlapping at the edges. The overlapping areas between polymetallic nodules are easily covered by mud and sand and are located at the edge of the polymetallic nodules and blend with the seabed background. They cannot be completely filled by the dilation algorithm, resulting in low accuracy of the final calculated coverage. Summary of the Invention

[0005] One of the objectives of this invention is to propose an evaluation method for deep-sea polymetallic nodule resources, which effectively solves the technical problem that the overlapping areas between polymetallic nodules are easily covered by sediment and are located at the edge of the polymetallic nodules and blend into the seabed background, making it impossible to completely fill them by the dilation algorithm, resulting in low accuracy of the final calculated coverage.

[0006] The technical solution of the present invention is as follows:

[0007] A method for assessing deep-sea polymetallic nodule resources includes the following steps:

[0008] S100: Acquire images of polymetallic nodules on the seabed in the target area;

[0009] S200: Preprocessing of seabed polymetallic nodule images based on the Sea-thru algorithm;

[0010] S300: Binarization of preprocessed seabed polymetallic nodule images based on threshold segmentation algorithm;

[0011] S400: The binarized image of the seabed polymetallic nodule is obtained by performing image filling once using a morphological reconstruction algorithm;

[0012] S500: The height of nodule individuals protruding from the seabed surface is obtained using three-dimensional reconstruction technology, and the particle size of nodule individuals is calculated based on the height of the nodule individuals;

[0013] S600: Based on particle size and the position of nodule individuals on the binarized image, determine the overlap of adjacent nodule individuals, and selectively perform secondary image filling on the binarized image according to the overlap of adjacent nodule individuals.

[0014] S700: Calculate the coverage rate by statistically analyzing the pixel ratio of the target region in the binarized image after secondary image filling.

[0015] Furthermore, step S500 includes:

[0016] S510: Using 3D reconstruction technology to obtain the height of individual nodules protruding from the seabed surface ;

[0017] S520: Calculate the width of the individual tuberculosis tuberculosis tuberculosis based on the binarized image. ;

[0018] S530: Judgment and Size;

[0019] like Less than or equal to The particle size r of a tuberculous individual is: ;

[0020] like Greater than The particle size of individual tuberculous tubers for:

[0021] ;

[0022] S540: Determine the overlap of adjacent nodule individuals based on particle size and the position of nodule individuals on the binarized image.

[0023] Furthermore, step S540 includes:

[0024] S541: Analyze the center pixel of each individual nodule based on binarized image;

[0025] S542: Construct a two-dimensional coordinate system on the binarized image, obtain the coordinate value of the center pixel of each nodule individual in the two-dimensional coordinate system, and calculate the distance s between the center pixels of two adjacent nodule individuals based on the coordinate values.

[0026] ;

[0027] In the formula, For the first The center pixel of a tuberculous individual axis coordinate values, For the first The center pixel of a tuberculous individual axis coordinate values, For the first The center pixel of a tuberculous individual axis coordinate values, For the first The center pixel of a tuberculous individual The coordinate values ​​of the axes, where the first axis is the first coordinate value. The first tuberculosis individual and the first Adjacent tuberculous individuals;

[0028] S543: Obtain the particle size of two individual nodules. ,like If two adjacent tuberculous individuals overlap, then it is determined that they are: ;

[0029] In the formula, The overlapping state parameters of individual tuberculous tubers;

[0030] like If the two adjacent tuberculous individuals do not overlap, then it can be determined that there is no overlap. .

[0031] Furthermore, in step S543, when there is overlap between two adjacent nodule individuals, the pixels corresponding to the two adjacent nodule individuals are connected point-to-point to obtain the first... The first tuberculosis individual The pixel and the The first tuberculosis individual The connecting line between pixels:

[0032] ;

[0033] ;

[0034] ;

[0035] In the formula, For any point on the connecting line corresponding axis coordinate values, For any point on the connecting line corresponding axis coordinate values, For the first The first tuberculosis individual 1 pixel axis coordinate values, For the first The first tuberculosis individual 1 pixel axis coordinate values, For the first The first tuberculosis individual 1 pixel axis coordinate values, For the first The first tuberculosis individual 1 pixel Axis coordinate values;

[0036] To complete the secondary image filling, adjust all pixels involved in the connecting lines between adjacent nodule individuals to 1.

[0037] Furthermore, step S540 includes:

[0038] S544: Edge pixels of each nodule individual based on binarized image analysis;

[0039] Calculate the distance between the edge pixels and the center pixel of each nodule individual. And filter out the maximum distance :

[0040] ;

[0041] In the formula, For tuberculosis individuals edge pixels axis coordinate values, The center pixel of the tuberculous individual axis coordinate values, For tuberculosis individuals edge pixels axis coordinate values, The center pixel of the tuberculous individual Axis coordinate values;

[0042] S545: Determine the maximum distance Particle size relative to individual tuberculosis cells Size;

[0043] like Greater than or equal to Then the original pixel combination shape of the nodule individual is maintained;

[0044] like Less than Then, with the center pixel as the center, Draw a circle with a radius of 1, and obtain the circular region:

[0045] ;

[0046] ;

[0047] In the formula, Any point within the circular region of axis coordinate values, Any point within the circular region of Axis coordinate values;

[0048] S546: Adjust the pixels of the pixels involved in the circular area to 1 to complete the secondary image filling.

[0049] Furthermore, step S540 includes:

[0050] S547: Based on the completion of secondary image filling, the coverage rate of tuberculous individuals is calculated using the following formula:

[0051] ;

[0052] In the formula, For the first The pixel area of ​​an individual nodule. The length of the image of polymetallic nodules on the seabed. This represents the width of the image of polymetallic nodules on the seabed.

[0053] Furthermore, step S200 includes:

[0054] S210: Constructing an underwater imaging physical model:

[0055] ;

[0056] In the formula, The final signal captured by the camera module, The attenuated direct signal, The blue-green color blocking light caused by backscattering For a clear, unattenuated, direct signal. As background light, For any channel in RGB, The actual position of each pixel and its distance from the camera module. , All are attenuation coefficients;

[0057] S220: Utilizes the structure-of-motion (SOG) algorithm to estimate the 3D structure from multiple images of polymetallic nodules on the seabed, obtaining the actual position of each pixel and its distance from the camera module. ;

[0058] S230: The blue-green shielding light caused by backscattering is expressed by the following formula. ,according to The images of seabed polymetallic nodules are divided into... Take the darkest area from each of the three regions at different distances. Each pixel is fitted with the parameters in the following formula using the least squares method, and the following formula is applied to all pixels. get :

[0059] ;

[0060] S240: Will from Removing the attenuated direct signal from the middle yields the signal. Estimate the attenuation coefficient ,Will Expressed according to the following formula:

[0061] ;

[0062] In the formula, All are fitted parameters;

[0063] S250: Based on estimated attenuation coefficient The fitting parameters in the formula are fitted using the nonlinear least squares method to obtain the final attenuation coefficient. :

[0064] S260: The results obtained from the above steps , , Substitute the data into an underwater imaging physical model to reconstruct a clear image of polymetallic nodules on the seabed.

[0065] Furthermore, step S200 also includes:

[0066] S270: Image enhancement of clear images of polymetallic nodules on the seabed using a white balance algorithm.

[0067] Furthermore, step S300 includes:

[0068] S310: Color image is processed using the following formula Convert to grayscale image :

[0069] ;

[0070] in, For color images The corresponding position above The pixel value of the upper red channel, For color images The corresponding position above The pixel value of the upper green channel, For color images The corresponding position above The pixel value of the upper blue channel;

[0071] S320: Yes On each pixel Calculate the mean of the domain in the domain. and standard deviation :

[0072] ;

[0073] ;

[0074] According to the mean and standard deviation Calculate the threshold for each pixel :

[0075] ;

[0076] In the formula, The dynamic range of the standard deviation. To correct the parameters;

[0077] S330: Classify pixels based on a threshold value for each pixel.

[0078] If the pixel value is greater than the threshold If so, mark it as 1;

[0079] If the pixel value is less than the threshold If it is, then mark it as 0;

[0080] Step S400 includes:

[0081] S410: Repeatedly dilate the image of polymetallic nodules on the seabed until convergence. The expression for the dilation process is as follows:

[0082] ;

[0083] In the formula, The image is repeatedly expanded until it converges. Images of polymetallic seabeds. The convolution size is... Template image;

[0084] Among them, iteration Next, until .

[0085] Another objective of this invention is to provide an evaluation system for deep-sea polymetallic nodule resources, comprising:

[0086] Camera module: Used to acquire images of polymetallic nodules on the seabed in the target area;

[0087] Image processing module: Communicatively connected to the camera module, the image processing module preprocesses the seabed polymetallic nodule image based on the Sea-thru algorithm, and performs binarization processing on the preprocessed seabed polymetallic nodule image;

[0088] First filling module: The binarized image of the seabed polymetallic nodule is filled by a morphological reconstruction algorithm to obtain a binarized image.

[0089] The second filling module is connected in communication with the first filling module. The second filling module determines the overlap of adjacent nodule individuals based on the particle size and the position of the nodule individuals on the binarized image, and selectively performs secondary image filling on the binarized image according to the overlap of adjacent nodule individuals.

[0090] Calculation module: Calculates the coverage rate based on the binarized image after secondary image filling.

[0091] The beneficial effects of this invention are as follows:

[0092] 1. This invention utilizes a morphological reconstruction algorithm to perform a first image filling on images of polymetallic nodules on the seabed, correcting pixels in areas where the surface of individual nodules is covered by mud and sand, thus improving image accuracy. In addition, the position of individual nodules on the binarized image is used to determine the overlap between adjacent nodules. When adjacent nodules overlap, a second image filling is performed on the binarized image to supplement pixels in areas where the edges of the nodules' surfaces are covered due to overlap, further improving the accuracy of coverage.

[0093] 2. Due to the uneven illumination often present in images of polymetallic nodules on the seabed, global thresholding segmentation can divide areas of varying brightness into two separate parts, making accurate segmentation difficult. Therefore, this invention employs a local segmentation method in step S300, calculating a corresponding threshold T for each pixel and then binarizing it based on the threshold T to achieve the final segmentation. This binarization method effectively avoids the problem of inaccurate image segmentation caused by uneven illumination. Attached Figure Description

[0094] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0095] Figure 1 This is a flowchart of the present invention;

[0096] Figure 2 This is a schematic diagram showing that two adjacent nodule individuals do not overlap in this invention;

[0097] Figure 3 This is a schematic diagram showing the overlap of two adjacent nodule individuals in this invention;

[0098] Figure 4 This is a schematic diagram illustrating the filling of connecting lines between two adjacent nodule individuals in this invention;

[0099] Figure 5 This is a schematic diagram of a single combined entity performing circle fitting and filling in this invention. Detailed Implementation

[0100] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. 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 of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0101] Example 1

[0102] like Figure 1-5 As shown, a method for assessing deep-sea polymetallic nodule resources involves first acquiring images of the target area's seabed polymetallic nodules using a camera module, or acquiring multiple frames of seabed polymetallic nodule images. These multiple frames are then stitched together to assess parameters over a wider area.

[0103] Specifically, the steps include the following:

[0104] S100: Acquire images of polymetallic nodules on the seabed in the target area;

[0105] S200: Preprocessing of seabed polymetallic nodule images based on the Sea-thru algorithm;

[0106] Step S200 includes:

[0107] S210: Constructing an underwater imaging physical model:

[0108] ;

[0109] In the formula, The final signal captured by the camera module, The attenuated direct signal, The blue-green color blocking light caused by backscattering For a clear, unattenuated, direct signal. As background light, For any channel in RGB, The actual position of each pixel and its distance from the camera module. , All are attenuation coefficients;

[0110] S220: Utilizes the structure-of-motion (SOG) algorithm to estimate the 3D structure from multiple images of polymetallic nodules on the seabed, obtaining the actual position of each pixel and its distance from the camera module. ;

[0111] S230: The blue-green shielding light caused by backscattering is expressed by the following formula. ,according to The images of seabed polymetallic nodules are divided into... Take the darkest area from each of the three regions at different distances. Each pixel is fitted with the parameters in the following formula using the least squares method, and the following formula is applied to all pixels. get :

[0112] ;

[0113] S240: Will from Removing the attenuated direct signal from the middle yields the signal. Estimate the attenuation coefficient ,Will Expressed according to the following formula:

[0114] ;

[0115] In the formula, All are fitted parameters;

[0116] S250: Based on estimated attenuation coefficient The fitting parameters in the formula are fitted using the nonlinear least squares method to obtain the final attenuation coefficient. :

[0117] S260: The results obtained from the above steps , , Substitute the data into an underwater imaging physical model to reconstruct a clear image of polymetallic nodules on the seabed.

[0118] Step S200 also includes:

[0119] S270: Image enhancement of clear images of polymetallic nodules on the seabed using a white balance algorithm.

[0120] S300: Binarization of preprocessed seabed polymetallic nodule images based on threshold segmentation algorithm;

[0121] Step S300 includes:

[0122] S310: Color image is processed using the following formula Convert to grayscale image :

[0123] ;

[0124] in, For color images The corresponding position above The pixel value of the upper red channel, For color images The corresponding position above The pixel value of the upper green channel, For color images The corresponding position above The pixel value of the upper blue channel;

[0125] S320: Yes On each pixel Calculate the mean of the domain in the domain. and standard deviation :

[0126] ;

[0127] ;

[0128] According to the mean and standard deviation Calculate the threshold for each pixel :

[0129] ;

[0130] In the formula, The dynamic range of the standard deviation. To correct the parameters;

[0131] S330: Classify pixels based on a threshold value for each pixel.

[0132] If the pixel value is greater than the threshold If so, mark it as 1;

[0133] If the pixel value is less than the threshold If it is, then mark it as 0;

[0134] One image filling is to fill in the mud and sand pixels on the surface of individual nodules.

[0135] S400: The binarized image of the seabed polymetallic nodule image is obtained by performing image filling once using a morphological reconstruction algorithm.

[0136] In this invention, a morphological reconstruction algorithm is first used to preprocess the polymetallic nodules. Based on the morphological characteristics of the polymetallic nodules, the buried or adhered parts can be reconstructed, and the initial and effective differentiation can be performed on the initial image, thereby reducing the difficulty of subsequent three-dimensional reconstruction and processing of nodule overlap.

[0137] Step S400 includes:

[0138] S410: Repeatedly dilate the image of polymetallic nodules on the seabed until convergence. The expression for the dilation process is as follows:

[0139] ;

[0140] In the formula, The image is repeatedly expanded until it converges. Images of polymetallic seabeds. The convolution size is... Template image;

[0141] Among them, iteration Next, until .

[0142] After one image filling, the white holes on the black target area corresponding to most nodule individuals were filled. The error in the statistical analysis of the nodule individual coverage rate caused by the seabed sediment coverage has been reduced to a low level. One image filling is only to fill the sediment on the surface of the nodule individuals. The edges of the nodule individuals and the overlapping areas cannot be filled according to the actual situation.

[0143] In this embodiment, three-dimensional reconstruction technology is first used to obtain the height of the nodule individual protruding from the seabed surface. The particle size of the nodule individual is calculated based on its height. Then, the overlap of adjacent nodule individuals is determined based on the particle size and the position of the nodule individual on the binarized image. Based on the overlap of adjacent nodule individuals, secondary image filling is selectively performed on the binarized image.

[0144] S500: The height of nodule individuals protruding from the seabed surface is obtained using three-dimensional reconstruction technology, and the particle size of nodule individuals is calculated based on the height of the nodule individuals;

[0145] Step S500 includes:

[0146] S510: Using 3D reconstruction technology to obtain the height of individual nodules protruding from the seabed surface ;

[0147] S520: Calculating the width of individual tuberculosis cells based on binarized images ;

[0148] Width of individual tuberculosis The calculation can be performed as follows: edge detection and contour extraction are performed on the tuberculosis individual, and finally the centroid of the figure is calculated. With the centroid as the endpoint, rays are extended to both sides, and the intersection points with the edges of the tuberculosis individual are extracted to obtain a line segment. The above steps are repeated to extend rays in different directions to obtain several line segments. The length of the line segments is calculated, and the longest length is taken as the width. .

[0149] S530: At the height of obtaining tuberculous individuals and width Then, a circle fitting method is used to determine... and Size;

[0150] like Less than or equal to This indicates that the center of the nodule is above the seabed level, and the particle size r of the nodule is: ;

[0151] like Greater than This indicates that the center of the tuberculous individual is below the seabed level, at which point the following conditions are met: , and All values ​​are known; particle size of individual tuberculous tubers. for:

[0152] ;

[0153] S540: The particle size of the tuberculous individual was calculated. Then, the overlap of adjacent nodule individuals is determined based on the particle size and the position of the nodule individuals on the binarized image.

[0154] Step S540, determining the overlap of adjacent nodule individuals based on particle size and the position of the nodule individual in the binarized image, specifically includes:

[0155] S541: Analyze the center pixel of each individual nodule based on binarized image;

[0156] S542: Construct a two-dimensional coordinate system on the binarized image, obtain the coordinate value of the center pixel of each nodule individual in the two-dimensional coordinate system, and calculate the distance s between the center pixels of two adjacent nodule individuals based on the coordinate values.

[0157] ;

[0158] In the formula, For the first The center pixel of a tuberculous individual axis coordinate values, For the first The center pixel of a tuberculous individual axis coordinate values, For the first The center pixel of a tuberculous individual axis coordinate values, For the first The center pixel of a tuberculous individual The coordinate values ​​of the axes, where the first axis is the first coordinate value. The first tuberculosis individual and the first Adjacent tuberculous individuals;

[0159] S543: Obtain the particle size of two individual nodules. ,like If two adjacent tuberculous individuals overlap, then it is determined that they are: ;

[0160] In the formula, The overlapping state parameters of individual tuberculous tubers;

[0161] like If the two adjacent tuberculous individuals do not overlap, then it can be determined that there is no overlap. .

[0162] In the above embodiments, the center pixel can be analyzed using the above method of finding the centroid, such as... Figure 3 As shown, Two adjacent tuberculous individuals may overlap, such as Figure 2 As shown, If two adjacent tuberculous individuals do not overlap, the above judgment method can only estimate the overlap state of two adjacent tuberculous individuals. Alternatively, the threshold for s can be set to... , here These are all weights, which can be adjusted according to the actual situation, or the optimal values ​​can be trained through a neural network to improve the accuracy of judging the overlap between two adjacent tuberculous individuals.

[0163] Based on binarized image analysis, the overlap between adjacent nodules is further determined by the distance between the central pixels (based on the distance between the central pixels and the difference in particle size). For nodules with overlap, a secondary image filling is performed. When performing the secondary image filling, the lines between the pixels are used as a guide, thereby determining the accuracy.

[0164] For cases where adjacent tuberculosis individuals overlap, this embodiment also provides two implementation methods to fill and optimize the images of overlapping tuberculosis individuals. One implementation method is as follows:

[0165] In step S543, when there is overlap between two adjacent nodule individuals, the pixels corresponding to the two adjacent nodule individuals are connected point-to-point to obtain the first... The first tuberculosis individual The pixel and the The first tuberculosis individual The connecting line between pixels:

[0166] ;

[0167] ;

[0168] ;

[0169] In the formula, For any point on the connecting line corresponding axis coordinate values, For any point on the connecting line corresponding axis coordinate values, For the first The first tuberculosis individual 1 pixel axis coordinate values, For the first The first tuberculosis individual 1 pixel axis coordinate values, For the first The first tuberculosis individual 1 pixel axis coordinate values, For the first The first tuberculosis individual 1 pixel Axis coordinate values;

[0170] To complete the secondary image filling, adjust all pixels involved in the connecting lines between adjacent nodule individuals to 1.

[0171] Because two adjacent nodule individuals may overlap below the seabed surface, the black areas of two actually overlapping nodule individuals in the binarized image may not actually overlap. Figure 4 As shown, the shaded area in the image represents the filled area. Adjusting the pixels on the connecting lines allows two adjacent nodule individuals to be directly filled into one nodule individual. This filled area not only compensates for the mud and sand covering the edges between two adjacent nodule individuals but also compensates for the mud and sand covering the other edges of the non-overlapping areas between the two nodule individuals. Therefore, the above secondary image filling method does not form a more accurate binarized image in terms of shape, but rather a binarized image with a more accurate number of pixels. This is necessary to obtain a more accurate coverage rate in the subsequent process. The binarized image obtained through the above steps is only used as a reference for coverage rate calculation and cannot be used as a reference for the actual shape and distribution of nodule individuals.

[0172] The above embodiments mainly fill the blank areas between two adjacent nodule individuals that overlap, avoiding the problem of low coverage caused by mud and sand covering the overlapping edges of adjacent nodule individuals.

[0173] As the next step after step S543, S547: Based on the completion of the secondary image filling, calculate the coverage of tuberculous individuals using the following formula:

[0174] ;

[0175] In the formula, For the first The pixel area of ​​an individual nodule. The length of the image of polymetallic nodules on the seabed. This represents the width of the image of polymetallic nodules on the seabed.

[0176] Using any of the above secondary image filling methods can make the pixel area of ​​the combined individuals on the binarized image more accurate, and make the coverage of all tuberculous individuals in the target area more accurate.

[0177] S600: Based on particle size and the position of nodule individuals on the binarized image, determine the overlap of adjacent nodule individuals, and selectively perform secondary image filling on the binarized image according to the overlap of adjacent nodule individuals.

[0178] S700: Calculate the coverage rate by statistically analyzing the pixel ratio of the target region in the binarized image after secondary image filling.

[0179] Example 2

[0180] Based on Example 1, another implementation method for filling and optimizing overlapping tuberculosis individual images is as follows: the pixel distribution of tuberculosis individuals on the binarized image is simulated by using circle fitting. Steps S541-S543 can be replaced by the following steps S544-S546 and serve as a prerequisite for S547.

[0181] S544: Edge pixels of each nodule individual based on binarized image analysis;

[0182] Calculate the distance between the edge pixels and the center pixel of each nodule individual. And filter out the maximum distance :

[0183] ;

[0184] In the formula, For tuberculosis individuals edge pixels axis coordinate values, The center pixel of the tuberculous individual axis coordinate values, For tuberculosis individuals edge pixels axis coordinate values, The center pixel of the tuberculous individual Axis coordinate values;

[0185] S545: Determine the maximum distance Particle size relative to individual tuberculosis cells Size;

[0186] like Greater than or equal to Then the original pixel combination shape of the nodule individual is maintained;

[0187] like Less than Then, with the center pixel as the center, Draw a circle with a radius of 1, and obtain the circular region:

[0188] ;

[0189] ;

[0190] In the formula, Any point within the circular region of axis coordinate values, Any point within the circular region of Axis coordinate values;

[0191] S546: Adjust the pixels of the pixels involved in the circular area to 1 to complete the secondary image filling.

[0192] The above embodiments involve circle fitting for potentially overlapping tuberculous individuals. In actual binarized images, the regions corresponding to tuberculous individuals may not be regular. The calculation of the tuberculous individual's... Afterwards, comparison and The size, if Greater than or equal to This indicates that most of the nodules are above the seabed surface and do not require additional filling. Less than This indicates that most of the nodule individuals are below the seabed surface, so the central pixel is taken as the center. Draw a circular region with a radius, such as Figure 5 As shown, the shaded area is the filled area. The pixels of the pixels in the circular area are adjusted to 1, and the pixels in the circular area are filled.

[0193] The above-mentioned circle fitting secondary filling method has the same effect as the connecting line filling method and achieves the optimization of the number of pixels of a single nodule. However, this secondary filling method does not form a more accurate binarized image in terms of shape, but rather a binarized image with a more accurate number of pixels. This is necessary to obtain a more accurate coverage rate in the subsequent process. The binarized image obtained through the above steps is only used as a reference for coverage rate calculation and cannot be used as a reference for the actual shape and distribution of the nodule.

[0194] In this invention, after processing the polymetallic nodules based on a morphological reconstruction algorithm, the subsequent processing no longer considers the morphology but converts them into pixels. This avoids interference from external data during the processing and eliminates the need for back-and-forth conversion between pixels and images, reducing the difficulty of data processing and further improving computational efficiency.

[0195] Example 3

[0196] An evaluation system for deep-sea polymetallic nodule resources, comprising:

[0197] Camera module: Used to acquire images of polymetallic nodules on the seabed in the target area;

[0198] Image processing module: Communicatively connected to the camera module, the image processing module preprocesses the seabed polymetallic nodule image based on the Sea-thru algorithm, and performs binarization processing on the preprocessed seabed polymetallic nodule image;

[0199] First filling module: The binarized image of the seabed polymetallic nodule is filled by a morphological reconstruction algorithm to obtain a binarized image.

[0200] The second filling module is connected in communication with the first filling module. The second filling module determines the overlap of adjacent nodule individuals based on the particle size and the position of the nodule individuals on the binarized image, and selectively performs secondary image filling on the binarized image according to the overlap of adjacent nodule individuals.

[0201] Calculation module: Calculates the coverage rate based on the binarized image after secondary image filling.

[0202] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for evaluating deep-sea polymetallic nodule resources, characterized in that, Includes the following steps: S100: Acquire images of polymetallic nodules on the seabed in the target area; S200: Preprocessing of seabed polymetallic nodule images based on the Sea-thru algorithm; S300: Binarization of preprocessed seabed polymetallic nodule images based on threshold segmentation algorithm; S310: Color image is processed using the following formula Convert to grayscale image : ； in, For color images The corresponding position above The pixel value of the upper red channel, For color images The corresponding position above The pixel value of the upper green channel, For color images Corresponding position The pixel value of the upper blue channel; S320: Yes On each pixel Calculate the mean of the domain in the domain. and standard deviation : ； ； According to the mean and standard deviation Calculate the threshold for each pixel : ； In the formula, The dynamic range of the standard deviation. To correct the parameters; S330: Classify pixels based on a threshold value for each pixel. If the pixel value is greater than the threshold If so, mark it as 1; If the pixel value is less than the threshold If it is, then mark it as 0; S400: The binarized image of the seabed polymetallic nodule is obtained by performing image filling once using a morphological reconstruction algorithm; S500: The height of nodule individuals protruding from the seabed surface is obtained using three-dimensional reconstruction technology, and the particle size of nodule individuals is calculated based on the height of the nodule individuals; S600: Based on particle size and the position of nodule individuals on the binarized image, determine the overlap of adjacent nodule individuals, and selectively perform secondary image filling on the binarized image according to the overlap of adjacent nodule individuals. Step S600 includes: S610: Analysis of the center pixel of each individual nodule based on binarized image; S620: Construct a two-dimensional coordinate system on the binarized image, obtain the coordinate value of the center pixel of each nodule individual in the two-dimensional coordinate system, and calculate the distance s between the center pixels of two adjacent nodule individuals based on the coordinate values. ； In the formula, For the first The center pixel of a tuberculous individual axis coordinate values, For the first The center pixel of a tuberculous individual axis coordinate values, For the first The center pixel of a tuberculous individual axis coordinate values, For the first The center pixel of a tuberculous individual The coordinate values ​​of the axes, where the first axis is the first coordinate value. The first tuberculosis individual and the first Adjacent tuberculous individuals; S630: Obtain the particle size of two individual nodules. , ,like If two adjacent tuberculous individuals overlap, then it is determined that they are: ; In the formula, The overlapping state parameters of individual tuberculous tubers; like If the two adjacent tuberculous individuals do not overlap, then it can be determined that there is no overlap. ; In step S630, when there is overlap between two adjacent nodule individuals, the pixels corresponding to the two adjacent nodule individuals are connected point-to-point to obtain the first... The first tuberculosis individual The pixel and the The first tuberculosis individual The connecting line between pixels: ； ； ； In the formula, For any point on the connecting line corresponding axis coordinate values, For any point on the connecting line corresponding axis coordinate values, For the first The first tuberculosis individual 1 pixel axis coordinate values, For the first The first tuberculosis individual 1 pixel axis coordinate values, For the first The first tuberculosis individual 1 pixel axis coordinate values, For the first The first tuberculosis individual 1 pixel Axis coordinate values; To complete the secondary image filling, adjust all pixels involved in the connecting lines between adjacent nodule individuals to 1. S700: Calculate the coverage rate by statistically analyzing the pixel ratio of the target region in the binarized image after secondary image filling.

2. The method for evaluating deep-sea polymetallic nodule resources according to claim 1, characterized in that, Step S500 includes: S510: Using 3D reconstruction technology to obtain the height of individual nodules protruding from the seabed surface ; S520: Calculate the width of the individual tuberculosis tuberculosis tuberculosis based on the binarized image. ; S530: Judgment and Size; like Less than or equal to The particle size of individual tuberculous tubers for: ; like Greater than The particle size of individual tuberculous tubers for: .

3. The method for evaluating deep-sea polymetallic nodule resources according to claim 1, characterized in that, Step S600 also includes: S640: Analyze the edge pixels of each nodule individual based on binarized image analysis; Calculate the distance between the edge pixels and the center pixel of each nodule individual. And filter out the maximum distance : ； In the formula, For tuberculosis individuals edge pixels axis coordinate values, The center pixel of the tuberculous individual axis coordinate values, For tuberculosis individuals edge pixels axis coordinate values, The center pixel of the tuberculous individual Axis coordinate values; S650: Determine maximum distance Particle size relative to individual tuberculosis cells Size; like Greater than or equal to Then the original pixel combination shape of the nodule individual is maintained; like Less than Then, with the center pixel as the center, Draw a circle with a radius of 1, and obtain the circular region: ； ； In the formula, Any point within the circular region of axis coordinate values, Any point within the circular region of Axis coordinate values; S660: Adjust the pixels of the pixels involved in the circular area to 1 to complete the secondary image filling.

4. The method for evaluating deep-sea polymetallic nodule resources according to claim 3, characterized in that, Step S600 also includes: S670: Based on the completion of secondary image filling, the coverage rate of tuberculous individuals is calculated using the following formula: ； In the formula, For the first The pixel area of ​​an individual nodule. The length of the image of polymetallic nodules on the seabed. This represents the width of the image of polymetallic nodules on the seabed.

5. The method for evaluating deep-sea polymetallic nodule resources according to claim 4, characterized in that, Step S200 includes: S210: Constructing an underwater imaging physical model: ； In the formula, The final signal captured by the camera module, The attenuated direct signal, The blue-green color blocking light caused by backscattering For a clear, unattenuated, direct signal. As background light, For any channel in RGB, The actual position of each pixel and its distance from the camera module. , All are attenuation coefficients; S220: Utilizes the structure-of-motion (SOG) algorithm to estimate the 3D structure from multiple images of polymetallic nodules on the seabed, obtaining the actual position of each pixel and its distance from the camera module. ; S230: The blue-green shielding light caused by backscattering is expressed by the following formula. ,according to The images of seabed polymetallic nodules are divided into... Take the darkest area from each of the three regions at different distances. Each pixel is fitted with the parameters in the following formula using the least squares method, and the following formula is applied to all pixels. get : ； S240: Will from Removing the attenuated direct signal from the middle yields the signal. Estimate the attenuation coefficient ,Will Expressed according to the following formula: ； In the formula, All are fitted parameters; S250: Based on estimated attenuation coefficient The fitting parameters in the formula are fitted using the nonlinear least squares method to obtain the final attenuation coefficient. : S260: The results obtained from the above steps , , Substitute the data into an underwater imaging physical model to reconstruct a clear image of polymetallic nodules on the seabed.

6. The method for evaluating deep-sea polymetallic nodule resources according to claim 5, characterized in that, Step S200 also includes: S270: Image enhancement of clear images of polymetallic nodules on the seabed using a white balance algorithm.

7. The method for evaluating deep-sea polymetallic nodule resources according to claim 1, characterized in that, Step S300 includes: Step S400 includes: S410: Repeatedly dilate the image of polymetallic nodules on the seabed until convergence. The expression for the dilation process is as follows: ； In the formula, The image is repeatedly expanded until it converges. Images of polymetallic seabeds. The convolution size is... Template image; Among them, iteration Next, until .

8. An evaluation system for deep-sea polymetallic nodule resources, used to implement the method described in any one of claims 1-7, characterized in that: include: Camera module: Used to acquire images of polymetallic nodules on the seabed in the target area; Image processing module: Communicatively connected to the camera module, the image processing module preprocesses the seabed polymetallic nodule image based on the Sea-thru algorithm, and performs binarization processing on the preprocessed seabed polymetallic nodule image; First filling module: The binarized image of the seabed polymetallic nodule is filled by a morphological reconstruction algorithm to obtain a binarized image. The second filling module is connected in communication with the first filling module. The second filling module determines the overlap of adjacent nodule individuals based on the particle size and the position of the nodule individuals on the binarized image, and selectively performs secondary image filling on the binarized image according to the overlap of adjacent nodule individuals. Calculation module: Calculates the coverage rate based on the binarized image after secondary image filling.

Images