A method and system for waste management data analysis and optimization based on a cloud computing platform

Abstract

This invention relates to the field of waste identification technology, specifically to a waste processing data analysis and optimization method and system under a cloud computing platform. The invention obtains the chromatic similarity coefficients of suspected waste regions between different stable frame images; combines the positional features of suspected waste regions between different stable frame images to obtain multiple groups of similar suspected waste regions between different stable frame images; obtains the floating passivity degree of each group of similar suspected waste regions between different adjacent stable frame images, and filters out groups of confirmed waste regions; based on the morphological features, texture features, and corresponding floating passivity degree of different confirmed waste regions in each group of confirmed waste regions, obtains the waste matching priority of each confirmed waste region in each group of confirmed waste regions; filters out the waste to be matched regions in each group of confirmed waste regions, and identifies the waste. This invention improves waste identification efficiency by obtaining accurate matching regions for waste.

Description

Technical Field

[0001] This invention relates to the field of waste identification technology, specifically to a waste treatment data analysis and optimization method and system based on a cloud computing platform. Background Technology

[0002] In marine plastic waste recycling projects, to improve recycling efficiency, underwater robots equipped with cameras are first used to conduct preliminary reconnaissance of the quantity and distribution of plastic waste in the seabed area. Then, a cloud computing platform identifies the waste based on the reconnaissance images and formulates a waste recycling plan. Therefore, accurate identification of waste is necessary to improve recycling efficiency.

[0003] In existing technologies, template matching algorithms are used to identify and classify marine debris affected by seabed debris detection. However, due to interference from fish in the ocean, there are many objects that need to be identified through template matching, which consumes a lot of redundant platform cloud computing resources. Furthermore, the complex seabed environment causes severe distortion of debris areas in some frames, affecting the accuracy of template matching and resulting in poor identification efficiency of debris images. Summary of the Invention

[0004] To address the technical problems of severe distortion in waste areas, which affects template matching accuracy and image recognition efficiency, the present invention aims to provide a waste treatment data analysis and optimization method and system based on a cloud computing platform. The specific technical solution adopted is as follows:

[0005] This invention proposes a waste management data analysis and optimization method under a cloud computing platform, the method comprising:

[0006] Acquire video frame images of low-velocity sea areas containing debris at every moment, along with water flow velocity data;

[0007] Based on the grayscale distribution of pixels between different video frame images, stable frame images are selected, and multiple suspected garbage regions in each stable frame image are obtained; based on the color difference of the suspected garbage regions between different stable frame images, the color similarity coefficient of the suspected garbage regions between different stable frame images is obtained.

[0008] Based on the location characteristics of suspected garbage areas between different stable frame images and the color similarity coefficient, multiple groups of similar suspected garbage areas between different stable frame images are obtained; based on the difference and change characteristics of each group of similar suspected garbage areas between different adjacent stable frame images and the water flow velocity data distribution of stable frame images, the floating passivity of each group of similar suspected garbage areas between different adjacent stable frame images is obtained, and garbage-identified area groups are selected.

[0009] Based on the morphological and textural features of different waste-identified areas in each group of waste-identified areas, as well as the corresponding floating passivity, the waste matching priority of each waste-identified area in each group of waste-identified areas is obtained; based on the waste matching priority of each waste-identified area in different groups of waste-identified areas, waste-to-match areas in each group of waste-identified areas are filtered out.

[0010] Waste is identified based on the area to be matched.

[0011] Furthermore, based on the grayscale distribution of pixels between different video frame images, the image stability of each video frame image is obtained;

[0012] If the image stability of each video frame is greater than the preset stability threshold, the corresponding video frame image is taken as a stable frame image.

[0013] Furthermore, the method for obtaining the image stability includes:

[0014] For any video frame image, count the number of pixels corresponding to each gray level, and construct a gray level representation curve with gray level size as the horizontal axis and the number of corresponding pixels as the vertical axis.

[0015] The average correlation coefficient between the grayscale performance curves of each video frame and other different video frames is obtained as the grayscale correlation of each video frame relative to other video frames.

[0016] Obtain the maximum value of the corresponding gray-level correlation in all video frame images as the gray-level reference correlation; obtain the ratio of the corresponding gray-level correlation to the gray-level reference correlation for each video frame image as the image stability of each video frame image.

[0017] Furthermore, the method for obtaining the suspected garbage area includes:

[0018] K-means clustering is performed on all pixels based on the grayscale values ​​of pixels in each stable frame image to obtain multiple pixel cluster regions, which are used as suspected garbage regions.

[0019] Furthermore, the method for obtaining the chromaticity similarity coefficient includes:

[0020] Obtain the parameter values ​​of each pixel in the suspected garbage region in each stable frame image in the LAB space;

[0021] For any suspected garbage region in any stable frame image, take the target region as the target region. Count the number of pixels with the same parameter value between different suspected garbage regions and the target region in each other stable frame image, and use this as the number of overlapping pixels.

[0022] Obtain the overlapping pixel data between each suspected garbage region and the target region in each other stable frame image, and the ratio of the maximum number of overlapping pixels, as the chromatic similarity coefficient between each suspected garbage region and the target region in each other stable frame image; change the target region to obtain the chromatic similarity coefficient of the suspected garbage region between different stable frame images.

[0023] Furthermore, the method for obtaining the suspected areas of similar waste includes:

[0024] Obtain the geometric center of each suspected garbage region on each stable frame image; obtain the relative distance between the corresponding geometric centers of suspected garbage regions between different stable frame images, as the region distance;

[0025] Negative correlation mapping is performed on the regional distances. The product of the negative correlation mapping result and the chromatic similarity coefficient of the suspected garbage regions between the corresponding stable frame images is calculated and normalized to serve as the regional similarity of the suspected garbage regions between different stable frame images.

[0026] If the regional similarity of suspected garbage regions between different stable frame images is greater than a preset similarity threshold, the corresponding suspected garbage regions between different stable frame images are taken as similar suspected garbage regions, and the similar suspected garbage regions of each suspected garbage region are obtained to form multiple sets of similar suspected garbage regions.

[0027] Furthermore, the process of obtaining the degree of floating passivity of suspected garbage regions in each group among different adjacent stable frame images, and filtering out garbage-identified region groups, includes:

[0028] The number of difference pixels in each group of suspected garbage regions of the same type between adjacent stable frame images is obtained as the floating intensity parameter of the corresponding suspected garbage regions of the same type between adjacent stable frame images.

[0029] The DTW distance between the curve composed of the floating intensity parameters and the curve composed of the water flow velocity data at the corresponding time of each group of suspected garbage areas of the same type is obtained. The DTW distance is negatively correlated and normalized to be mapped as the floating passivity of each group of suspected garbage areas of the same type between different adjacent stable frame images.

[0030] If the degree of fluctuation of suspected waste areas in each group of similar waste exceeds the preset threshold, the suspected waste areas in the corresponding group of similar waste will be identified as waste areas.

[0031] Furthermore, the method for obtaining the garbage matching priority includes:

[0032] For each group of garbage-identified regions, the number of pixels in each garbage-identified region is obtained as a morphological feature; the morphological feature of each garbage-identified region is normalized, and the product of the normalization result and the corresponding floating passive intensity is calculated as the garbage target degree of each garbage-identified region in each group of garbage-identified regions.

[0033] The corner points of each garbage-defined region are obtained based on the corner detection algorithm; the average gradient of all pixels in the neighborhood of each corner point is obtained as the texture saliency value of each corner point; the average texture saliency value of all corner points in each garbage-defined region is obtained as the texture performance level of each defined region.

[0034] Obtain the product of the garbage target degree and texture representation level of each garbage-defined region in each group of garbage-defined regions, and perform normalization mapping as the garbage matching priority of each region in each group of garbage-defined regions.

[0035] Furthermore, the method for obtaining the garbage region to be matched includes:

[0036] Select the waste region with the highest waste matching priority value among all waste regions in each group of waste regions, and use the corresponding waste region as the waste region to be matched.

[0037] The present invention also proposes a waste management data analysis and optimization system under a cloud computing platform, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements any of the steps of the waste management data analysis and optimization method under a cloud computing platform.

[0038] The present invention has the following beneficial effects:

[0039] This invention selects stable frame images based on the grayscale distribution of pixels between different video frames. This allows for the analysis of frames with stable image quality, reducing the complexity and errors in subsequent processing. It also identifies multiple potential waste regions within each stable frame image, helping to narrow down the processing scope and improve recognition efficiency. Furthermore, based on the color differences between these potential waste regions in different stable frames, it obtains a chromatic similarity coefficient, reflecting the color consistency of waste across different frames and aiding in the identification of objects with similar color features. Finally, based on the positional characteristics and chromatic similarity coefficients of these potential waste regions in different stable frames, it obtains multiple groups of similar potential waste regions, helping to reduce redundant information and improve recognition accuracy. Finally, it identifies multiple groups of similar waste regions between adjacent stable frames. By analyzing the variation characteristics of suspected waste areas and the distribution of water flow velocity data in stable frame images, the floating passivity of each group of similar suspected waste areas in different adjacent stable frame images is obtained. This allows for the selection of waste-identified regions, reflecting the stability and dynamic changes of waste in video frame images. Groups of waste-identified regions with high stability and low dynamic changes are selected. Based on the morphological and texture features and corresponding floating passivity of different waste-identified regions within each group, the waste matching priority of each waste-identified region in each group is obtained. This allows for the selection of waste-to-match regions within each group, providing a more comprehensive evaluation of the characteristics of each waste-identified region, thereby more accurately identifying the type and attributes of waste and selecting waste-to-match regions with high matching priority. This invention improves waste identification efficiency by obtaining accurate matching regions for waste. Attached Figure Description

[0040] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 A flowchart illustrating a waste management data analysis and optimization method based on a cloud computing platform, as provided in one embodiment of the present invention;

[0042] Figure 2 This is a flowchart illustrating a method for obtaining suspected areas of similar waste, provided in one embodiment of the present invention.

[0043] Figure 3 This is a flowchart illustrating a method for obtaining garbage matching priority according to an embodiment of the present invention. Detailed Implementation

[0044] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a waste management data analysis and optimization method and system based on a cloud computing platform proposed by the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0045] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0046] The following description, in conjunction with the accompanying drawings, details the specific scheme of a waste treatment data analysis and optimization method and system based on a cloud computing platform provided by this invention.

[0047] Please see Figure 1 The document illustrates a flowchart of a waste management data analysis and optimization method based on a cloud computing platform, according to an embodiment of the present invention. The method specifically includes:

[0048] Step S1: Acquire video frame images of the low-velocity sea area containing garbage at each moment, as well as water flow velocity data.

[0049] In embodiments of the present invention, to avoid pollution and damage to the marine environment and improve waste recycling efficiency, accurate identification of waste is necessary. First, an underwater robot equipped with a camera captures video footage of a low-current area, and the video frames are then analyzed. Water flow velocity data of the captured area is collected by a water flow velocity sensor on the underwater robot and uploaded to a cloud-based data acquisition system for further processing. This process acquires video frame images and water flow velocity data of the low-current area containing waste at each moment.

[0050] It should be noted that, in the embodiments of the present invention, the acquired video frame images are placed in the same mapped image according to coordinates, that is, the left vertical edge of the video frame image is used as the vertical axis of the coordinate system, and the lower edge perpendicular to the vertical axis is used as the horizontal axis of the coordinate system, so as to coordinate the position of each pixel point on each image, which facilitates the subsequent analysis of the image.

[0051] It should be noted that, in one embodiment of the present invention, video images collected within a historical 10 seconds at a real-time moment are acquired, with a time interval of 0.2 seconds, i.e., 0.2 seconds as one frame, and the video images are converted into static frame images to acquire multiple video frame images; in other embodiments of the present invention, the time range and time interval can be specifically set according to the specific situation, and are not limited or described in detail here.

[0052] Step S2: Based on the grayscale distribution of pixels between different video frame images, select stable frame images and obtain multiple suspected garbage regions in each stable frame image; based on the color difference of suspected garbage regions between different stable frame images, obtain the color similarity coefficient of suspected garbage regions between different stable frame images.

[0053] Considering that fish near the camera will cover a large area of ​​trash, and that even slight movements can cause significant differences in image quality, resulting in changes in the grayscale values ​​of pixels in the image, analyzing the grayscale distribution of pixels in different video frames can better reflect changes in image quality and the coverage of trash areas. Selecting stable frame images can help analyze the specific morphology of the trash. Based on the grayscale distribution of pixels between different video frames, stable frame images are selected, and multiple suspected trash areas are obtained in each stable frame image.

[0054] Preferably, in one embodiment of the present invention, the method for obtaining a stable frame image includes:

[0055] The image stability of each video frame is obtained based on the grayscale distribution of pixels between different video frames.

[0056] In one embodiment of the present invention, the method for obtaining image stability includes:

[0057] For any video frame image, count the number of pixels corresponding to each gray level, and construct a gray level representation curve with gray level size as the horizontal axis and the number of corresponding pixels as the vertical axis.

[0058] The average correlation coefficient between the grayscale performance curves of each video frame and other different video frames is obtained as the grayscale correlation of each video frame relative to other video frames.

[0059] Obtain the maximum value of the corresponding gray-level correlation in all video frame images as the gray-level reference correlation; obtain the ratio of the corresponding gray-level correlation to the gray-level reference correlation for each video frame image as the image stability of each video frame image.

[0060] It should be noted that, in one embodiment of the present invention, the correlation coefficient is the Pearson correlation coefficient; the specific means are well known to those skilled in the art and will not be described in detail here.

[0061] If the image stability of each video frame is greater than the preset stability threshold, the corresponding video frame image is taken as a stable frame image.

[0062] It should be noted that, in one embodiment of the present invention, the preset stability threshold is 0.78; in other embodiments of the present invention, the preset stability threshold can be set according to specific circumstances, and will not be limited or elaborated here.

[0063] In an image, different objects or regions will exhibit different grayscale values ​​due to factors such as their material, color, and lighting conditions. K-means clustering can divide the pixels in the image into different cluster regions, which may correspond to different objects or regions, and can be analyzed as suspected garbage areas.

[0064] Preferably, in one embodiment of the present invention, the method for obtaining suspected garbage areas includes:

[0065] K-means clustering is performed on all pixels based on the grayscale values ​​of pixels in each stable frame image to obtain multiple pixel cluster regions, which are used as suspected garbage regions.

[0066] The K-means clustering algorithm groups multiple clusters of objects into K specified clusters based on their similarity. Each object belongs to one and only one cluster whose distance to the center of the suspected waste area is minimized by both its distance to the center and its grayscale distance. It should be noted that, in one embodiment of this invention, when using the K-means clustering algorithm to cluster all pixels, the K value is obtained by using the elbow rule to determine the K value, thus obtaining the corresponding number of pixel cluster regions as suspected waste areas. The specific K-means clustering algorithm and elbow rule are well-known techniques to those skilled in the art and will not be elaborated upon here.

[0067] The same seabed object should have similar color representation in different frame images. By analyzing the color differences of suspected garbage areas between different stable frame images, we can assess its color uniformity and variability, quantify the color similarity of suspected garbage areas in different stable frame images, and thus help to identify garbage more accurately. Based on the color differences of suspected garbage areas between different stable frame images, we can obtain the color similarity coefficient of suspected garbage areas between different stable frame images.

[0068] Preferably, in one embodiment of the present invention, the method for obtaining the chromatic similarity coefficient includes:

[0069] The LAB color space is the color space with the widest color gamut, which can more accurately describe and compare colors; obtain the parameter values ​​of each pixel in the suspected garbage region in each stable frame image in the LAB color space; take any suspected garbage region in any stable frame image as the target region, count the number of pixels with the same parameter values ​​between different suspected garbage regions and the target region in each other stable frame image, and use this as the number of overlapping pixels;

[0070] Obtain the overlapping pixel data between each suspected garbage region and the target region in each other stable frame image, and the ratio of the maximum number of overlapping pixels, as the chromatic similarity coefficient between each suspected garbage region and the target region in each other stable frame image; change the target region to obtain the chromatic similarity coefficient of the suspected garbage region between different stable frame images.

[0071] Based on this, analyzing the chromatic similarity coefficients of any two suspected debris regions between different stable frame images is more helpful in assessing the possibility that the two suspected debris regions are the same seabed object.

[0072] Step S3: Based on the location characteristics of suspected garbage areas between different stable frame images and the color similarity coefficient, obtain multiple groups of similar suspected garbage areas between different stable frame images; based on the difference and change characteristics of each group of similar suspected garbage areas between different adjacent stable frame images and the water flow velocity data distribution of stable frame images, obtain the floating passivity of each group of similar suspected garbage areas between different adjacent stable frame images, and filter out the garbage identification area group.

[0073] The positional differences of the same object in adjacent frames are relatively small. The chromaticity similarity coefficient can quantify the color similarity between different suspected garbage regions. When the chromaticity similarity coefficient of two regions is high, it indicates that the regions have similar color features and may belong to the same object. Therefore, by combining positional features and chromaticity similarity coefficients, similar suspected garbage regions in different stable frame images can be identified more accurately. Based on the positional features of suspected garbage regions in different stable frame images and the chromaticity similarity coefficients, multiple sets of similar suspected garbage regions in different stable frame images are obtained.

[0074] Preferably, in one embodiment of the present invention, the method for obtaining suspected areas of similar waste is described in [reference needed]. Figure 2 It illustrates a flowchart of a method for obtaining suspected garbage areas, including:

[0075] Step S201: Obtain the geometric center of each suspected garbage region on each stable frame image; obtain the relative distance between the corresponding geometric centers of suspected garbage regions between different stable frame images, as the region distance.

[0076] To better analyze the matching relationship of suspected trash regions in stable frame images, the distance between the geometric centers of suspected trash regions is analyzed. This helps to quantify the positional changes between regions. The greater the relative distance, the greater the distance between regions, and the less likely it is that they are the same object.

[0077] It should be noted that, in the embodiments of the present invention, the geometric center of each suspected waste area can be obtained by the shape boundary centroid algorithm; or the geometric center can be obtained by calculating the average coordinates of all pixels in the suspected waste area; the specific means are well known to those skilled in the art and will not be described in detail here.

[0078] Step S202: Perform negative correlation mapping on the regional distance, calculate the product of the negative correlation mapping result and the chromatic similarity coefficient of the suspected garbage region between the corresponding stable frame images, and normalize it as the regional similarity of the suspected garbage region between different stable frame images.

[0079] The greater the distance between regions, the greater the variation in location, the less likely the two regions are the same object, and the smaller the regional similarity, showing a negative correlation.

[0080] In one embodiment of the present invention, the formula for regional similarity is expressed as:

[0081]

[0082] Among them, S x,y,u,r V represents the region similarity between the y-th suspected garbage region in the x-th stable frame image and the r-th suspected garbage region in the u-th stable frame image; x,y,u,r D represents the chromatic similarity coefficient between the y-th suspected garbage region in the x-th stable frame image and the r-th suspected garbage region in the u-th stable frame image; x,y,u,r This represents the relative distance between the corresponding geometric centers of the y-th suspected garbage region in the x-th stable frame image and the r-th suspected garbage region in the u-th stable frame image, i.e., the region distance; norm() represents the normalization function.

[0083] In the formula for regional similarity, D x,y,u,r Adding 0.01 to +0.01 is meaningless if the denominator is 0; the greater the distance between suspected garbage areas, the greater the positional difference, and the lower the credibility of the area similarity; the smaller the distance between suspected garbage areas, the smaller the positional difference, and the higher the credibility of the area similarity, and the greater the color similarity coefficient, the greater the possibility that they are the same type of object, and the greater the area similarity.

[0084] It should be noted that in some embodiments of the present invention, existing distance algorithms such as Euclidean distance and Manhattan distance can be used to obtain the relative distance between geometric centers. The specific means are well known to those skilled in the art and will not be described in detail here.

[0085] Step S203: If the regional similarity of suspected garbage regions between different stable frame images is greater than a preset similarity threshold, the corresponding suspected garbage regions between different stable frame images are taken as similar suspected garbage regions, and the similar suspected garbage regions of each suspected garbage region are obtained to form multiple sets of similar suspected garbage regions.

[0086] It should be noted that, in one embodiment of the present invention, the preset similarity threshold is 0.88; in other embodiments of the present invention, the preset similarity threshold can be set according to specific circumstances, and will not be limited or elaborated here.

[0087] Submarine plastic debris floats locally with the flow of seawater. The floating pattern of the debris should be positively correlated with the seawater flow velocity. However, fish have autonomous life in relation to the debris, which does not conform to the relevant laws of seawater flow. Therefore, by analyzing the differences in the variation characteristics of suspected areas of the same type of debris in each group, the positional changes of objects within the area can be reflected. Combined with the distribution of water flow velocity data, the degree of passive floating of suspected debris areas can be quantitatively assessed, thereby screening out groups of confirmed debris areas, which helps in the analysis of debris characteristics. Based on the differences in the variation characteristics of suspected areas of the same type of debris between different adjacent stable frame images, and the distribution of water flow velocity data in the stable frame images, the degree of passive floating of suspected areas of the same type of debris between different adjacent stable frame images can be obtained, and groups of confirmed debris areas can be screened out.

[0088] Preferably, in one embodiment of the present invention, obtaining the degree of floating passivity of each group of suspected garbage regions between different adjacent stable frame images, and filtering out garbage-identified region groups, includes:

[0089] The number of difference pixels in each group of suspected garbage regions of the same type between adjacent stable frame images is obtained as the floating intensity parameter of each group of suspected garbage regions of the same type between adjacent stable frame images.

[0090] The DTW distance between the curve composed of the floating intensity parameters of each group of suspected garbage areas at the corresponding time and the curve composed of the water flow velocity data is obtained. The DTW distance is negatively correlated and normalized to be mapped as the floating passivity of each group of suspected garbage areas between different adjacent stable frame images.

[0091] In one embodiment of the present invention, the formula for the degree of floating passivity is expressed as:

[0092]

[0093] Among them, IB h Indicates the degree of passive floating of the h-th group of suspected waste areas; DTW hThe DTW distance is represented between the curve composed of the floating intensity parameter and the curve composed of the water flow velocity data at the corresponding time in the h-th group of suspected garbage areas of the same type; norm() represents the normalization function.

[0094] In the formula for the degree of floating passivity, DTW h The addition of 0.01 to +0.01 is to avoid the denominator being 0, as this would render the formula meaningless. For the suspected areas of the same type of garbage in the h-th group, the larger the DTW distance between the curves formed by the floating intensity parameters and the water flow velocity data at the corresponding time for each suspected area of ​​the same type of garbage, the smaller the similarity between the object's floating and the water flow velocity in the time sequence, the stronger the object's autonomous action awareness, the smaller the degree of passive floating, and the smaller the possibility that it is garbage.

[0095] If the degree of fluctuation of suspected waste areas in each group of similar waste exceeds the preset threshold, the suspected waste areas in the corresponding group of similar waste will be identified as waste areas.

[0096] It should be noted that, in one embodiment of the present invention, the preset degree threshold is 0.8; in other embodiments of the present invention, the preset degree threshold can be set according to specific circumstances, and will not be limited or elaborated here.

[0097] Step S4: Based on the morphological features, texture features, and corresponding floating passivity of different garbage determination areas in each group of garbage determination areas, obtain the garbage matching priority of each garbage determination area in each group of garbage determination areas; based on the garbage matching priority of each garbage determination area in different groups of garbage determination areas, filter out the garbage to be matched areas in each group of garbage determination areas.

[0098] Because the alignment rate between trash and the camera is low at certain times, fewer feature points are exposed, increasing the difficulty of subsequent trash template matching. Therefore, the morphological features of the trash-identified areas are analyzed; the larger the morphological features, the higher the priority of trash matching. Seawater contains many biological metabolites, leading to complex turbidity variations and distortion in some trash areas. Therefore, the texture features of the trash-identified areas are analyzed; the more obvious the texture features, the more obvious the features, and the better for trash identification. The floating passivity of the trash-identified areas indicates the correlation between objects within the area and seawater flow. The greater the floating passivity, the more likely the floating is caused by seawater flow, and the greater the correlation, the more likely it is a trash area. Therefore, based on the morphological features, texture features, and corresponding floating passivity of different trash-identified areas in each group of trash-identified areas, the trash matching priority of each trash-identified area in each group of trash-identified areas is obtained.

[0099] Preferably, in one embodiment of the present invention, the method for obtaining the garbage matching priority is described in [reference needed]. Figure 3It shows a flowchart of a method for obtaining garbage matching priority, including:

[0100] Step S301: For each group of garbage-determined regions, obtain the number of pixels in each garbage-determined region as a morphological feature; normalize the morphological feature of each garbage-determined region, and calculate the product of the normalization result and the corresponding floating passive intensity as the garbage target degree of each garbage-determined region in each group of garbage-determined regions.

[0101] The number of pixels, as a morphological feature, can intuitively reflect the size and shape of the identified area of ​​garbage. The larger the morphological feature, the more feature points it exhibits. Normalization can map the feature values ​​of different areas to the same scale, which is convenient for subsequent comparison and analysis. The floating passive strength serves as the confidence level for judging whether something is garbage. The greater the floating passive strength, the more likely it is to conform to the relevant laws of seawater flow, and the more comprehensively and accurately the degree of garbage target assessment.

[0102] In one embodiment of the present invention, the formula for the degree of garbage target is expressed as:

[0103]

[0104] Among them, IE m,n Indicates the degree of waste targeting in the nth waste-identified area within the mth waste-identified area; IB m Q represents the degree of passive floating of the m-th group of garbage in the defined region; m,n Q represents the number of pixels in the nth garbage-identified region within the mth garbage-identified region, i.e., the morphological features of the nth garbage-identified region; m,max This represents the maximum number of pixels within all garbage-defined regions in the m-th garbage-defined region.

[0105] In the formula for the degree of garbage target, This represents the ratio of the number of pixels in the nth garbage region within the m-th garbage region to the maximum number of pixels in all garbage regions. In other words, it normalizes the morphological features of each garbage region. The larger the ratio, the larger the normalization result. The larger the number of pixels in the nth garbage region within the m-th garbage region, the larger the morphological features, the more likely it is to show more feature information, the greater the degree of passive floating, the less autonomous action awareness, and the greater the degree of garbage target.

[0106] Step S302: Obtain the corner points of each garbage-determined region based on the corner detection algorithm; obtain the average gradient value of all pixels in the neighborhood of each corner point as the texture saliency value of each corner point; obtain the average texture saliency value of all corner points in each garbage-determined region as the texture performance level of each garbage-determined region.

[0107] In a defined waste region, corners may correspond to the edges, inflection points, or key structural points of the waste. Corner detection algorithms can accurately identify corners in an image, and the gradient mean can reflect the degree of texture change in a local area of ​​the image. By calculating the gradient mean within the neighborhood of a corner, texture information around the corner can be obtained. Furthermore, the texture performance level of each defined waste region can be evaluated through the mean, which can reflect the overall texture characteristics of the waste region and help distinguish different types of waste.

[0108] It should be noted that, in one embodiment of the present invention, the neighborhood range of a corner point is the range formed by the corner pixel as the center and all adjacent pixels; in other embodiments of the present invention, the neighborhood range of a corner point can be specifically set according to the specific situation, and will not be limited or described in detail here.

[0109] Step S303: Obtain the product of the garbage target degree and texture representation level of each garbage determination region in each group of garbage determination regions, and perform normalization mapping as the garbage matching priority of each region in each group of garbage determination regions.

[0110] In one embodiment of the present invention, the formula for garbage matching priority is expressed as:

[0111] IF m,n =norm(IE m,n ×U m,n );

[0112] Among them, IF m,n This indicates the garbage matching priority of the nth garbage region within the m-th garbage region; IE m,n U represents the degree of waste targeting in the nth waste-identified area within the mth waste-identified area; m,n This represents the texture representation level of the nth garbage region within the m-th garbage region; norm() represents the normalization function.

[0113] In the formula for garbage matching priority, the greater the degree of garbage target, the more garbage features the region displays, the greater the level of texture representation, the lower the degree of distortion, and the higher the garbage matching priority.

[0114] Waste matching priority is used to assess the likelihood of waste and the degree of texture distortion. The higher the waste matching priority, the higher the classification accuracy and the richer the texture information. The more information that represents the characteristics of waste, the better it is for waste identification. Based on the waste matching priority of each waste determination area in different groups of waste determination areas, waste to be matched areas in each group of waste determination areas are selected.

[0115] Preferably, in one embodiment of the present invention, the method for obtaining the garbage region to be matched includes:

[0116] Select the waste region with the highest waste matching priority value among all waste regions in each group of waste regions, and use the corresponding waste region as the waste region to be matched.

[0117] Step S5: Identify the waste based on the waste matching area.

[0118] By accurately identifying the areas to be matched, the system can more accurately determine the type of waste, thereby reducing the possibility of misclassification.

[0119] It should be noted that, in another embodiment of the present invention, after obtaining the garbage to be matched area, the garbage can be identified, including: mapping each garbage to be matched area to an image with a grayscale value of 0 to obtain the corresponding garbage to be matched image; the image analysis module in the cloud computing platform performs a template matching algorithm on the garbage to be matched image to achieve garbage identification and classification; and then the data transmission module sends the seabed garbage identification and classification results to the seabed garbage recycling center, which helps to formulate a recycling strategy based on the identification and classification situation, and to clean and recycle seabed plastic garbage according to the recycling strategy by the garbage recycling submersible.

[0120] In summary, this invention analyzes the grayscale distribution of pixels across different video frames to obtain the chromatic similarity coefficient of suspected waste regions between different stable frame images; combines the positional features of suspected waste regions between different stable frame images to obtain multiple groups of similar suspected waste regions between different stable frame images; obtains the floating passivity of each group of similar suspected waste regions between different adjacent stable frame images, and filters out groups of confirmed waste regions; based on the morphological features, texture features, and corresponding floating passivity of different confirmed waste regions in each group of confirmed waste regions, obtains the waste matching priority of each confirmed waste region in each group of confirmed waste regions; and filters out the waste to be matched regions in each group of confirmed waste regions for waste identification. This invention improves waste identification efficiency by obtaining accurate matching regions for waste.

[0121] This invention also proposes a waste management data analysis and optimization system under a cloud computing platform, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements any of the steps of a waste management data analysis and optimization method under a cloud computing platform.

[0122] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0123] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for analyzing and optimizing waste management data under a cloud computing platform, characterized in that, The method includes: Acquire video frame images of low-velocity sea areas containing debris at every moment, along with water flow velocity data; Based on the grayscale distribution of pixels between different video frame images, stable frame images are selected, and multiple suspected garbage regions in each stable frame image are obtained; based on the color difference of the suspected garbage regions between different stable frame images, the color similarity coefficient of the suspected garbage regions between different stable frame images is obtained. Based on the location characteristics of suspected garbage areas between different stable frame images and the color similarity coefficient, multiple groups of similar suspected garbage areas between different stable frame images are obtained; based on the difference and change characteristics of each group of similar suspected garbage areas between different adjacent stable frame images and the water flow velocity data distribution of stable frame images, the floating passivity of each group of similar suspected garbage areas between different adjacent stable frame images is obtained, and garbage-identified area groups are selected. Based on the morphological and textural features of different waste-identified areas in each group of waste-identified areas, as well as the corresponding floating passivity, the waste matching priority of each waste-identified area in each group of waste-identified areas is obtained; based on the waste matching priority of each waste-identified area in different groups of waste-identified areas, waste-to-match areas in each group of waste-identified areas are filtered out. Waste is identified based on the area to be matched; The process of obtaining the degree of floating passivity of suspected garbage regions in each group between different adjacent stable frame images, and filtering out garbage-identified region groups, includes: The number of difference pixels in each group of suspected garbage regions of the same type between adjacent stable frame images is obtained as the floating intensity parameter of the corresponding suspected garbage regions of the same type between adjacent stable frame images. The DTW distance between the curve composed of the floating intensity parameters and the curve composed of the water flow velocity data at the corresponding time of each group of suspected garbage areas of the same type is obtained. The DTW distance is negatively correlated and normalized to be mapped as the floating passivity of each group of suspected garbage areas of the same type between different adjacent stable frame images. If the degree of fluctuation of suspected waste areas in each group of similar waste exceeds the preset threshold, the suspected waste areas in the corresponding group of similar waste will be identified as waste areas.

2. The waste management data analysis and optimization method under a cloud computing platform according to claim 1, characterized in that, The method for obtaining the stable frame image includes: The image stability of each video frame is obtained based on the grayscale distribution of pixels between different video frames. If the image stability of each video frame is greater than the preset stability threshold, the corresponding video frame image is taken as a stable frame image.

3. The waste management data analysis and optimization method under a cloud computing platform according to claim 2, characterized in that, The method for obtaining the image stability includes: For any video frame image, count the number of pixels corresponding to each gray level, and construct a gray level representation curve with gray level size as the horizontal axis and the number of corresponding pixels as the vertical axis. The average correlation coefficient between the grayscale performance curves of each video frame and other different video frames is obtained as the grayscale correlation of each video frame relative to other video frames. Obtain the maximum value of the corresponding gray-level correlation in all video frame images as the gray-level reference correlation; obtain the ratio of the corresponding gray-level correlation to the gray-level reference correlation for each video frame image as the image stability of each video frame image.

4. The waste management data analysis and optimization method under a cloud computing platform according to claim 1, characterized in that, The method for obtaining the suspected garbage area includes: K-means clustering is performed on all pixels based on the grayscale values ​​of pixels in each stable frame image to obtain multiple pixel cluster regions, which are used as suspected garbage regions.

5. The waste management data analysis and optimization method under a cloud computing platform according to claim 1, characterized in that, The method for obtaining the chromatic similarity coefficient includes: Obtain the parameter values ​​of each pixel in the suspected garbage region in each stable frame image in the LAB space; For any suspected garbage region in any stable frame image, take the target region as the target region. Count the number of pixels with the same parameter value between different suspected garbage regions and the target region in each other stable frame image, and use this as the number of overlapping pixels. Obtain the overlapping pixel data between each suspected garbage region and the target region in each other stable frame image, and the ratio of the maximum number of overlapping pixels, as the chromatic similarity coefficient between each suspected garbage region and the target region in each other stable frame image; change the target region to obtain the chromatic similarity coefficient of the suspected garbage region between different stable frame images.

6. The waste management data analysis and optimization method under a cloud computing platform according to claim 1, characterized in that, The methods for obtaining suspected areas of similar waste include: Obtain the geometric center of each suspected garbage region on each stable frame image; obtain the relative distance between the corresponding geometric centers of suspected garbage regions between different stable frame images, as the region distance; Negative correlation mapping is performed on the regional distances. The product of the negative correlation mapping result and the chromatic similarity coefficient of the suspected garbage regions between the corresponding stable frame images is calculated and normalized to serve as the regional similarity of the suspected garbage regions between different stable frame images. If the regional similarity of suspected garbage regions between different stable frame images is greater than a preset similarity threshold, the corresponding suspected garbage regions between different stable frame images are taken as similar suspected garbage regions, and the similar suspected garbage regions of each suspected garbage region are obtained to form multiple sets of similar suspected garbage regions.

7. The waste management data analysis and optimization method under a cloud computing platform according to claim 1, characterized in that, The method for obtaining the garbage matching priority includes: For each group of garbage-identified regions, the number of pixels in each garbage-identified region is obtained as a morphological feature; the morphological feature of each garbage-identified region is normalized, and the product of the normalization result and the corresponding floating passive intensity is calculated as the garbage target degree of each garbage-identified region in each group of garbage-identified regions. The corner points of each garbage-defined region are obtained based on the corner detection algorithm; the average gradient of all pixels in the neighborhood of each corner point is obtained as the texture saliency value of each corner point; the average texture saliency value of all corner points in each garbage-defined region is obtained as the texture performance level of each defined region. Obtain the product of the garbage target degree and texture representation level of each garbage-defined region in each group of garbage-defined regions, and perform normalization mapping as the garbage matching priority of each region in each group of garbage-defined regions.

8. The waste management data analysis and optimization method under a cloud computing platform according to claim 1, characterized in that, The method for obtaining the garbage to be matched region includes: Select the waste region with the highest waste matching priority value among all waste regions in each group of waste regions, and use the corresponding waste region as the waste region to be matched.

9. A waste management data analysis and optimization system based on a cloud computing platform, the system comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the waste disposal data analysis and optimization method under a cloud computing platform as described in any one of claims 1 to 8.

Images