Rural domestic sewage treatment effect image recognition evaluation method, system and platform based on deep learning

Abstract

The invention discloses a rural domestic sewage treatment effect image recognition and evaluation method, system and platform based on deep learning. The method comprises the following steps: constructing a problem picture data set corresponding to a treatment effect evaluation associated object; generating and acquiring first data corresponding to each associated object, labeling and processing a type label of a target object in the first data, and creating a corresponding training set and verification set; building an image recognition model based on deep learning, optimizing model parameters through the training set and the verification set, and generating second data corresponding to each associated object according to the image recognition model; performing scoring processing on the second data according to a preset score assignment rule, and generating third data corresponding to the second data; and the third data is evaluation treatment effect data integrating the scores of the associated objects, and a system and a platform corresponding to the method can solve the problems of high professional requirements of evaluation personnel, poor result consistency, low efficiency and high cost of an existing agricultural pollution effect evaluation mode.

Description

Technical Field

[0001] This invention belongs to the interdisciplinary field of intelligent environmental governance technology, specifically involving a method, system, and platform for image recognition and evaluation of the effectiveness of rural domestic sewage treatment based on deep learning. Background Technology

[0002] Rural domestic sewage treatment is a crucial component of rural living environment improvement and a vital measure in the fight against pollution. It holds significant importance for promoting rural ecological revitalization, rural ecological civilization, and the construction of beautiful villages. In recent years, my country has continuously advanced its efforts in rural domestic sewage treatment, resulting in a steady improvement in treatment levels. However, current assessments of the effectiveness of rural domestic sewage treatment primarily rely on manual on-site visual inspections. This method requires experienced experts and is inefficient. Furthermore, differences in professional backgrounds and subjective judgment among assessors directly affect the accuracy of the assessments, leading to deviations in guidance for standardized facility construction, enhanced operation and maintenance, and improved overall effectiveness. Especially when assessing the effectiveness of a large number of rural domestic sewage treatment facilities, manual on-site visual inspections are not only highly specialized but also time-consuming, labor-intensive, and costly. They can only be achieved through sampling assessments and periodic inspections. The low efficiency, high cost, and inconsistent results of on-site visual inspections fail to meet the needs for routine, comprehensive, and rapid assessment and supervision of rural domestic sewage treatment effectiveness.

[0003] With the continuous development of artificial intelligence algorithms, deep learning-based image recognition technology has replaced manual visual inspection and is widely used for the detection of target objects in images. This method constructs a training set by labeling the types of target objects in the image. After optimizing the parameters of the image recognition model using the training set, it can identify the target type in the image, using the image as the analysis object. For example, the AI-based fireworks analysis method and system based on video images (CN113989735B), the image recognition-based method, equipment, and medium for detecting defects in car seats (CN118864448B), and a visual detection method and system for aquaculture water pollution (CN116863323B) have all achieved the detection of target object categories in images using the above methods. The effectiveness assessment of rural domestic sewage treatment mainly involves on-site visual inspection of whether there is sewage accumulation and the operation of treatment facilities within the village. Regarding the situation within the village, relevant objects include whether there is sewage flowing freely in public spaces (such as alleys, main roads, and open spaces between households); whether drainage ditches have accumulated large amounts of domestic sewage, kitchen waste, and silt; and whether there are stinking ponds. The operational status of the facility mainly focuses on whether wastewater enters the facility, whether each treatment unit is functioning properly, and whether the effluent is normal. In summary, the assessment of agricultural wastewater treatment effectiveness involves numerous related objects and complex problem types. Therefore, how to quickly and accurately identify the specific categories of related objects in the assessment process using deep learning image recognition models and image data, and how to achieve an efficient and low-cost accurate assessment of agricultural wastewater treatment effectiveness, is a technical problem that urgently needs to be solved by those skilled in the art.

[0004] In other words, the current assessment of the effectiveness of rural domestic sewage treatment relies on manual on-site visual inspection. This method is inefficient, costly, and the consistency of assessment results needs further improvement. It cannot meet the needs of normalized, comprehensive, and rapid assessment and supervision of the effectiveness of rural domestic sewage treatment, and seriously restricts the high-quality development of rural domestic sewage treatment.

[0005] Therefore, in order to address the above-mentioned technical problems and shortcomings, there is an urgent need to design and develop an image recognition evaluation method, system, and platform for the effectiveness of rural domestic sewage treatment based on deep learning. Summary of the Invention

[0006] To overcome the shortcomings and difficulties of the existing technologies, the present invention aims to provide a method, system and platform for evaluating the effectiveness of rural domestic sewage treatment based on deep learning, which aims to solve the problems of high professional requirements for evaluators, poor consistency of results, low efficiency and high cost faced by existing methods of evaluating the effectiveness of rural sewage treatment.

[0007] The first objective of this invention is to provide a deep learning-based image recognition and evaluation method for the effectiveness of rural domestic sewage treatment; the second objective of this invention is to provide a deep learning-based image recognition and evaluation system for the effectiveness of rural domestic sewage treatment; and the third objective of this invention is to provide a deep learning-based image recognition and evaluation platform for the effectiveness of rural domestic sewage treatment.

[0008] The first objective of this invention is achieved as follows: the method comprises the following steps:

[0009] Construct a problem dataset corresponding to the objects associated with the governance effectiveness assessment; wherein, the associated objects include the sewage accumulation status in village public spaces, the type of accumulated materials in drainage ditches, the water status of ponds, the water inlet status of facilities, the functional status of facility treatment units, and the water outlet status of facilities;

[0010] Generate and acquire first data corresponding to each associated object, label the type of the target object in the first data, and create corresponding training and validation sets; wherein, the first data includes problematic positive sample and non-problematic negative sample image data;

[0011] A deep learning-based image recognition model is constructed, and the model parameters are optimized using the training set and validation set. Based on the image recognition model, second data corresponding to each associated object is generated; wherein, the second data is specific type recognition result data.

[0012] The second data is scored according to the preset scoring rules, and a third data corresponding to the second data is generated; wherein, the third data is a comprehensive evaluation of the governance effectiveness based on the scores of various related objects.

[0013] Furthermore, the specific types of the associated objects include:

[0014] The village's public spaces are divided into alleys / main roads / front and back of houses with no water accumulation, with clear water accumulation, and with sewage accumulation;

[0015] Drainage ditches are classified into dry ditches, clear water ditches, and ditches that accumulate domestic sewage / kitchen waste / silt.

[0016] The pond is divided into two states: normal water quality and black water quality; the facility water intake is divided into three states: no water, clear water, and sewage intake.

[0017] The facility treatment units are divided into normal constructed wetlands, severely backwatered constructed wetlands, constructed wetlands with severe vegetation loss, normal reaction tanks, clear reaction tanks, normal main structures, and leaking main structures, etc.

[0018] The facility's water output is categorized into three states: no water output, normal water output, and black water output.

[0019] Furthermore, the step of building a deep learning-based image recognition model, optimizing model parameters using the training and validation sets, and generating second data corresponding to each associated object based on the image recognition model, further includes:

[0020] A deep learning-based image recognition model for evaluating the effectiveness of rural sewage treatment was constructed, and convolutional neural networks were used to extract and fuse image features.

[0021] Generate specific type data corresponding to the objects reflected in the image; wherein, the number of output layer nodes matches the total number of specific types of the associated objects.

[0022] Furthermore, the step of scoring the second data according to a preset scoring rule and generating third data corresponding to the second data also includes:

[0023] A scoring rule corresponding to governance effectiveness is created, and a fourth data point corresponding to governance effectiveness is generated based on the scoring rule; wherein, the fourth data point is governance effectiveness evaluation data;

[0024] Based on the fourth data and combined with the governance effectiveness review process, a fifth data corresponding to the governance effectiveness is generated; wherein, the fifth data is the data indicating poor governance effectiveness assessment results;

[0025] Based on the training set and in conjunction with the corresponding training model, the fifth data is retrained and processed.

[0026] Furthermore, the scoring rules are as follows:

[0027] For the identification result of each associated object, 1 point is assigned if there is no problem and 0 points are assigned if there is a problem.

[0028] Multiply the scores of the 6 related objects together. A total score of 1 is considered good, and 0 is considered poor.

[0029] When the evaluation result is unsatisfactory, the specific type of all associated objects with a score of 0 will be output synchronously.

[0030] The second objective of this invention is achieved as follows: the system is used to implement the deep learning-based image recognition and evaluation method for the effectiveness of rural domestic sewage treatment; the system includes:

[0031] The first data processing unit is used to construct a problem dataset corresponding to the objects associated with the governance effectiveness assessment; wherein, the associated objects include the sewage accumulation status in the village's public spaces, the type of accumulated materials in drainage ditches, the water status of ponds, the water inlet status of facilities, the functional status of facility treatment units, and the water outlet status of facilities;

[0032] The second data processing unit is used to generate and acquire first data corresponding to each associated object, and to label the type of the target object in the first data, and to create corresponding training and validation sets; wherein, the first data includes problematic positive sample and non-problematic negative sample image data;

[0033] The first data generation unit is used to build a deep learning-based image recognition model, optimize the model parameters through the training set and validation set, and generate second data corresponding to each associated object according to the image recognition model; wherein, the second data is specific type recognition result data;

[0034] The second data generation unit is used to score the second data according to a preset scoring rule and generate third data corresponding to the second data; wherein the third data is data that comprehensively evaluates the governance effectiveness based on the scores of various related objects.

[0035] Furthermore, the specific types of the associated objects include:

[0036] The village's public spaces are divided into alleys / main roads / front and back of houses with no water accumulation, with clear water accumulation, and with sewage accumulation;

[0037] Drainage ditches are classified into dry ditches, clear water ditches, and ditches that accumulate domestic sewage / kitchen waste / silt.

[0038] The pond is divided into two states: normal water quality and black water quality; the facility water intake is divided into three states: no water, clear water, and sewage intake.

[0039] The facility treatment units are divided into normal constructed wetlands, severely backwatered constructed wetlands, constructed wetlands with severe vegetation loss, normal reaction tanks, clear reaction tanks, normal main structures, and leaking main structures, etc.

[0040] The facility's water output is classified into three states: no water output, normal water output, and black water output.

[0041] And / or, the first data generation unit further includes:

[0042] The first processing module is used to construct a deep learning-based image recognition model for evaluating the effectiveness of rural sewage treatment, and to extract and fuse image features using a convolutional neural network.

[0043] The first generation module is used to generate specific type data corresponding to the objects reflected in the image; wherein, the number of output layer nodes matches the total number of specific types of associated objects;

[0044] And / or, the second data generation unit further includes:

[0045] The second generation module is used to create scoring rules corresponding to governance effectiveness, and generate fourth data corresponding to governance effectiveness based on the scoring rules; wherein, the fourth data is governance effectiveness evaluation data;

[0046] The third generation module is used to generate fifth data corresponding to the governance effectiveness based on the fourth data and the governance effectiveness review process; wherein, the fifth data is the data of poor governance effectiveness evaluation results;

[0047] The second processing module is used to retrain and process the fifth data based on the training set and in combination with the corresponding training model.

[0048] Furthermore, the scoring rules are as follows:

[0049] For the identification result of each associated object, 1 point is assigned if there is no problem and 0 points are assigned if there is a problem.

[0050] Multiply the scores of the 6 related objects together. A total score of 1 is considered good, and 0 is considered poor.

[0051] When the evaluation result is unsatisfactory, the specific type of all associated objects with a score of 0 will be output synchronously.

[0052] The third objective of this invention is achieved as follows: it includes a processor, a memory, and a control program for a deep learning-based image recognition and evaluation platform for the effectiveness of rural domestic sewage treatment; wherein the deep learning-based image recognition and evaluation platform control program is executed on the processor, the deep learning-based image recognition and evaluation platform control program is stored in the memory, and the deep learning-based image recognition and evaluation platform control program implements the deep learning-based image recognition and evaluation method for the effectiveness of rural domestic sewage treatment.

[0053] This invention constructs a problem dataset corresponding to objects associated with the evaluation of governance effectiveness through a method. The associated objects include the sewage accumulation status in village public spaces, the type of accumulated material in drainage ditches, the water status of ponds, the water inlet status of facilities, the functional status of facility treatment units, and the water outlet status of facilities. First data corresponding to each associated object is generated and acquired, and type labels for target objects in the first data are labeled. Corresponding training and validation sets are created. The first data includes image data of positive samples with problems and negative samples without problems. A deep learning-based image recognition model is built, and the model parameters are optimized using the training and validation sets. Based on the image recognition model, second data corresponding to each associated object is generated. The second data is specific type identification result data. The second data is scored according to preset scoring rules, and third data corresponding to the second data is generated. The third data is a comprehensive evaluation of governance effectiveness based on the scores of each associated object. The invention also includes a system and platform corresponding to the method, which can solve the problems of high professional requirements for assessors, poor consistency of results, low efficiency, and high cost faced by existing agricultural pollution control effectiveness evaluation methods. In other words, it achieves efficient, low-cost, and highly consistent evaluation of agricultural pollution control effectiveness. Attached Figure Description

[0054] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying 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.

[0055] Figure 1 This is a schematic diagram of the process of an image recognition and evaluation method for the effectiveness of rural domestic sewage treatment based on deep learning, according to the present invention.

[0056] Figure 2 This is a schematic flowchart of an embodiment of the image recognition and evaluation method for the effectiveness of rural domestic sewage treatment based on deep learning according to the present invention.

[0057] Figure 3 This is a schematic diagram of the architecture of a deep learning-based image recognition and evaluation system for the effectiveness of rural domestic sewage treatment according to the present invention.

[0058] Figure 4 This is a schematic diagram of the architecture of an embodiment of the image recognition and evaluation system for the effectiveness of rural domestic sewage treatment based on deep learning according to the present invention;

[0059] Figure 5 This is a schematic diagram of the architecture of an image recognition and evaluation platform for the effectiveness of rural domestic sewage treatment based on deep learning, according to the present invention. Detailed Implementation

[0060] To facilitate a clearer understanding of the objectives, technical solutions, and advantages of this invention, the invention will be further described below in conjunction with the accompanying drawings and specific embodiments. Those skilled in the art can easily understand other advantages and effects of this invention from the content disclosed in this specification.

[0061] This invention can also be implemented or applied through other different specific examples, and various details in this specification can also be modified and changed based on different viewpoints and applications without departing from the spirit of this invention.

[0062] It should be noted that if the embodiments of the present invention involve directional indicators (such as up, down, left, right, front, back, etc.), the directional indicators are only used to explain the relative positional relationship and movement of the components in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicators will also change accordingly.

[0063] Furthermore, if the embodiments of this invention involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of those features. Secondly, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by this invention.

[0064] Preferably, the deep learning-based image recognition evaluation method for the effectiveness of rural domestic sewage treatment is applied in one or more terminals or servers. The terminal is a device capable of automatically performing numerical calculations and / or information processing according to pre-set or stored instructions. Its hardware includes, but is not limited to, microprocessors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0065] The terminal can be a desktop computer, laptop, handheld computer, or cloud server, etc. The terminal can interact with the customer via a keyboard, mouse, remote control, touchpad, or voice control device.

[0066] This invention provides a method, system, and platform for evaluating the effectiveness of rural domestic sewage treatment based on deep learning through image recognition.

[0067] like Figure 1 The diagram shown is a flowchart of an image recognition and evaluation method for the effectiveness of rural domestic sewage treatment based on deep learning, provided in an embodiment of the present invention.

[0068] In this embodiment, the deep learning-based image recognition evaluation method for the effectiveness of rural domestic sewage treatment can be applied to terminals or fixed terminals with display functions. The terminals are not limited to personal computers, smartphones, tablets, desktop computers or all-in-one computers with cameras, etc.

[0069] The deep learning-based image recognition and evaluation method for the effectiveness of rural domestic sewage treatment can also be applied to a hardware environment consisting of a terminal and a server connected to the terminal via a network. The network includes, but is not limited to, a wide area network (WAN), a metropolitan area network (MAN), or a local area network (LAN). The deep learning-based image recognition and evaluation method for the effectiveness of rural domestic sewage treatment in this embodiment can be executed by the server, by the terminal, or by both the server and the terminal.

[0070] For example, for terminals requiring image recognition and evaluation of rural domestic sewage treatment effectiveness based on deep learning, the deep learning-based image recognition and evaluation function of rural domestic sewage treatment effectiveness provided by the method of this invention can be directly integrated into the terminal, or a client for implementing the method of this invention can be installed. Alternatively, the method provided by this invention can also run on servers or other devices in the form of a Software Development Kit (SDK), providing an interface for the deep learning-based image recognition and evaluation function of rural domestic sewage treatment effectiveness. Terminals or other devices can then implement the deep learning-based image recognition and evaluation function of rural domestic sewage treatment effectiveness through the provided interface. The invention will be further described below with reference to the accompanying drawings.

[0071] like Figures 1-2 As shown, this invention provides an image recognition and evaluation method for the effectiveness of rural domestic sewage treatment based on deep learning. The method includes the following steps:

[0072] S1. Construct a problem dataset corresponding to the objects associated with the governance effectiveness assessment; wherein, the associated objects include the sewage accumulation status in village public spaces, the type of accumulated material in drainage ditches, the water status of ponds, the water inlet status of facilities, the functional status of facility treatment units, and the water outlet status of facilities;

[0073] S2. Generate and acquire first data corresponding to each associated object, and label the type of the target object in the first data, and create corresponding training and validation sets; wherein, the first data includes problematic positive sample and non-problematic negative sample image data;

[0074] S3. Build a deep learning-based image recognition model, optimize the model parameters using the training set and validation set, and generate second data corresponding to each associated object based on the image recognition model; wherein, the second data is specific type recognition result data;

[0075] S4. The second data is scored according to the preset scoring rules, and a third data corresponding to the second data is generated; wherein, the third data is the data for evaluating the governance effectiveness by comprehensively considering the scores of each related object.

[0076] The specific types of the associated objects include:

[0077] The village's public spaces are divided into alleys / main roads / front and back of houses with no water accumulation, with clear water accumulation, and with sewage accumulation;

[0078] Drainage ditches are classified into dry ditches, clear water ditches, and ditches that accumulate domestic sewage / kitchen waste / silt.

[0079] The pond is divided into two states: normal water quality and black water quality; the facility water intake is divided into three states: no water, clear water, and sewage intake.

[0080] The facility treatment units are divided into normal constructed wetlands, severely backwatered constructed wetlands, constructed wetlands with severe vegetation loss, normal reaction tanks, clear reaction tanks, normal main structures, and leaking main structures, etc.

[0081] The facility's water output is categorized into three states: no water output, normal water output, and black water output.

[0082] The step of building a deep learning-based image recognition model, optimizing model parameters using the training and validation sets, and generating second data corresponding to each associated object based on the image recognition model, further includes:

[0083] S31. Construct a deep learning-based image recognition model for evaluating the effectiveness of rural sewage treatment, and combine it with a convolutional neural network to extract and fuse image features respectively;

[0084] S32. Generate specific type data corresponding to the objects reflected in the image; wherein the number of output layer nodes matches the total number of specific types of the associated objects.

[0085] The step of scoring the second data according to a preset scoring rule and generating third data corresponding to the second data further includes:

[0086] S41. Create scoring rules corresponding to governance effectiveness, and generate fourth data corresponding to governance effectiveness based on the scoring rules; wherein, the fourth data is governance effectiveness evaluation data;

[0087] S42. Based on the fourth data and combined with the governance effectiveness review process, generate the fifth data corresponding to the governance effectiveness; wherein, the fifth data is the data indicating poor governance effectiveness evaluation results;

[0088] S43. Based on the training set and in conjunction with the corresponding training model, retrain and process the fifth data.

[0089] The scoring rules are as follows: for the identification result of each associated object, 1 point is assigned if there is no problem and 0 points are assigned if there is a problem; the scores of the 6 associated objects are multiplied together, and the evaluation result is good when the total score is 1 point and poor when it is 0 points; when the evaluation result is poor, the specific type of all associated objects with a score of 0 is output simultaneously.

[0090] Specifically, in this embodiment of the invention, a deep learning-based image recognition and evaluation method for the effectiveness of rural domestic sewage treatment is provided, comprising the following steps: constructing a problem dataset of associated objects for evaluating treatment effectiveness, wherein the associated objects include: sewage accumulation status in village public spaces, accumulation type in drainage ditches, pond water status, facility influent status, facility treatment unit functional status, and facility effluent status; collecting images of problematic positive samples and non-problematic negative samples for each associated object, labeling the type of the target object in the image, and constructing a training set and a validation set; building a deep learning-based image recognition model, optimizing the model parameters through the training set and validation set, so that the model can identify the specific type of the target object in the input image; inputting the rural sewage treatment site image to be evaluated into the trained model, and outputting the specific type recognition result of each associated object; scoring the recognition result according to a preset scoring rule, and comprehensively evaluating the treatment effectiveness based on the scores of each associated object.

[0091] The specific types of objects associated with the steps include: village public spaces are divided into alleys / main roads / front and back of houses with no water accumulation, with clear water accumulation, and with sewage accumulation; drainage ditches are divided into ditches with no water, with clear water, and with accumulated domestic sewage / kitchen waste / silt; ponds are divided into two states: normal water and black water; facility inlet water is divided into three states: no water, clear water, and sewage inlet; facility treatment units are divided into at least 6 process states, such as normal artificial wetlands and severely backwater artificial wetlands; facility effluent water is divided into three states: no effluent, normal effluent, and black effluent.

[0092] The image recognition model in the above steps uses a convolutional neural network to extract features and includes a feature fusion module. The number of nodes in its output layer matches the total number of specific types of associated objects.

[0093] The scoring rules in the steps are as follows: for each associated object's identification result, assign 1 point for a problem-free state and 0 points for a problem state; multiply the scores of the 6 associated objects, and the total score is 1 point for an evaluation result of "good" and 0 points for "poor"; when the evaluation result is "poor", simultaneously output the specific type of all associated objects with a score of 0. The model optimization also includes the steps of: manually reviewing the cases with an evaluation result of "poor", and adding the incorrectly identified images to the training set to retrain the model.

[0094] Among them, a method for evaluating the effectiveness of rural domestic sewage treatment based on a deep learning image recognition model is provided, which includes the following:

[0095] (1) Construction of a dataset of related objects and problems for governance effectiveness. Based on the main related objects for evaluating the effectiveness of rural domestic sewage treatment, the effectiveness of rural domestic sewage treatment is assessed from six related objects: whether there is sewage accumulation in public spaces (such as alleys, main roads, and open spaces between households), whether domestic sewage accumulates in drainage ditches, whether pond water is black and smelly, whether sewage flows into facilities, whether the treatment unit of the facility is functioning normally, and whether the effluent from the facility is normal. Common problems affecting the effectiveness of treatment are collected and sorted out for each related object, and the related objects and corresponding problems are summarized to establish a dataset of problems for the six objects related to the effectiveness of treatment.

[0096] (2) Construction of the Image Dataset for the Correlation of Governance Effectiveness. Based on the problem types of the six objects associated with the effectiveness of rural domestic sewage treatment, photos of objects with problems (positive samples) and without problems (subsamples) were collected for the assessment of the effectiveness of rural domestic sewage treatment. The photos were manually labeled with the specific type of object reflected in the image. Images without problems were labeled as "no problem," thus constructing the image dataset for the correlation of governance effectiveness. A portion of the images in the dataset was used as the training dataset, and the remainder was used as the test dataset.

[0097] (3) Construction and Training Optimization of Image Recognition Model for Effectiveness Evaluation. A deep learning-based image recognition model for evaluating the effectiveness of rural sewage treatment was constructed to achieve feature extraction, feature fusion, and feature-based recognition of images. The output result is the specific type of the object reflected in the image. The recognition model was trained using images labeled in the database as input. The model optimizes its internal parameters by learning the features in the images, thereby improving the accuracy of the model in recognizing the input images.

[0098] (4) Image object type identification, governance effectiveness assessment, and specific type output of problematic objects. Input the photos of the six assessed objects into the optimized image recognition model to obtain the specific categories of the target objects in the images. Assign scores based on the image recognition results of the specific categories of the assessed objects; a score of 0 is given for objects with problems, and a score of 1 is given for objects without problems. Multiply the scores of the six assessed object photos to evaluate the effectiveness of agricultural pollution control. A score of 1 is recorded as "good," and the output result is "good"; a score of 0 is recorded as "poor," and the specific categories of all assessed objects with a score of 0 are also output.

[0099] This study identifies key stakeholders in evaluating the effectiveness of rural domestic sewage treatment under the facility-based treatment model. Typical problems affecting treatment effectiveness across these stakeholders are identified, leading to a summary of six key stakeholders and their corresponding issues. First, for village public spaces, the focus is on whether sewage accumulates in alleys, main roads, and around houses. Second, for village drainage ditches, the focus is on whether there is accumulation of black and odorous domestic sewage, kitchen waste, and silt. Third, for village ponds, the focus is on whether the pond water is black and odorous. Fourth, regarding sewage inflow into facilities, the focus is on whether sewage is present in the inspection wells closest to the facilities. Fifth, regarding the normal functioning of facility treatment units, the focus is on issues that significantly affect facility function or operation, such as severe water backlog on the surface of constructed wetlands, clear biological reactors in biological processes, and leakage in the main structure. Sixth, regarding the normality of facility effluent, the focus is on whether the facilities effluent and whether the effluent is black. This leads to the establishment of a dataset of six stakeholders and their corresponding issues for evaluating treatment effectiveness.

[0100] Based on the problem types of six objects associated with the effectiveness of rural domestic sewage treatment, photos of objects with problems (positive samples) and without problems (no problems) were collected for the assessment of the effectiveness of rural domestic sewage treatment. The photos were manually labeled, with the label information indicating the specific category of the object reflected in the image. Objects without problems were labeled as "no problem," thus constructing a dataset based on images of the six objects associated with the effectiveness assessment of treatment. The specific steps are as follows: First, for public spaces within the village, photos of alleyways, main roads, and areas in front of and behind houses with no water accumulation, with clear water accumulation, and with sewage accumulation were collected. Target objects were selected and labeled sequentially as alleyways with no water accumulation, alleyways with clear water accumulation, and alleyways with sewage accumulation; and alleyways with no water accumulation, alleyways with clear water accumulation, and alleyways with sewage accumulation. The first step is to collect photos of the main roads, including areas around houses with water accumulation (no water, clear water, and sewage). The second step is to collect photos of village drainage ditches, including those with no water, those with clear water, and those with accumulated black and foul-smelling domestic sewage, kitchen waste, and silt. The selected areas are then labeled as "ditches with no water," "ditches with clear water," and "ditches with accumulated black and foul-smelling domestic sewage, kitchen waste, and silt." The third step is to collect photos of village ponds, including those with water that is not black and those with black water. The selected areas are then labeled as "ponds with normal water" and "ponds with black water." The fourth step is to collect photos of... Whether sewage is flowing into the facility is mainly determined by identifying whether there is sewage in the inspection well closest to the facility. Photos of inspection wells with no water, with clean water, and with sewage are collected, and the target objects are selected and labeled sequentially as "no water entering the facility," "clean water entering the facility," and "sewage entering the facility." Fifth, whether the facility's treatment units are functioning normally. Regarding constructed wetlands, photos of normal constructed wetlands, severely waterlogged constructed wetlands, and constructed wetlands with severely depleted vegetation are collected, and the target objects are selected and labeled sequentially as "normal constructed wetlands," "severely waterlogged constructed wetlands," and "severely depleted vegetation." For industrial wetlands; for biological process bioreactors, collect photos of normal and clear reactors, and label the target objects as "normal reactor" and "clear reactor" respectively; for the main structures of the corresponding facilities, collect photos of normal and leaking main structures, and label the target objects as "normal main structure" and "leaking main structure" respectively; sixth, regarding whether the effluent from the facilities is normal, collect photos of facilities with no effluent, normal effluent, and black effluent, and label the target objects as "facilities with no effluent, normal effluent, and black effluent" respectively. Furthermore, there are sufficient photos for each of the above objects to ensure they can be used to train a deep learning-based image recognition model;

[0101] Based on six objects associated with the rural domestic sewage treatment effectiveness assessment, a deep learning-based image recognition model was constructed. A partial dataset of images representing these six objects was used as input to train the deep learning-based image recognition model. This model extracts, fuses, and identifies features from the images, outputting the specific categories of objects contained within them, such as alleyways with or without water accumulation, facilities with no or blackened effluent, etc. The model learns from the features in the images to optimize its internal parameters, thereby improving the accuracy of identifying specific categories of objects in the input images.

[0102] Photos reflecting village public spaces, drainage ditches, ponds, water intake, treatment unit functionality, and effluent conditions are collected. These photos are then input into an optimized image recognition model for effectiveness assessment, yielding image recognition results indicating the specific categories of target objects within the images. Based on the specific categories of objects reflected in each image, a score is assigned to each image according to a scoring table for relevant objects and specific categories in rural domestic sewage treatment effectiveness assessment. The scores of the six assessment object photos are multiplied; a score of 1 is recorded as "good," and a score of 0 is recorded as "poor." The specific categories of all assessment objects with a score of 0 are also output. This process simultaneously achieves the assessment of treatment effectiveness and the output of specific categories of relevant objects affecting treatment effectiveness.

[0103] Table 1. Scoring of Relevant Objects and Specific Categories in the Evaluation of Rural Domestic Sewage Treatment Effectiveness

[0104]

[0105]

[0106] In addition, to further improve the accuracy of the deep learning-based image recognition model in evaluating the effectiveness of agricultural pollution control, for cases where the output results are unsatisfactory, manual sampling is used to verify whether the specific type of the object reflected in the image matches the specific type output by the image recognition model. Inconsistent images are used as training datasets for model optimization.

[0107] The key point and protected point of this invention is to propose a method for rapidly assessing the effectiveness of agricultural pollution control by detecting and scoring specific categories of objects associated with the assessment of agricultural pollution control effectiveness based on a deep learning image recognition model; and simultaneously outputting the assessment of control effectiveness and the specific categories of objects associated with the control effectiveness based on the deep learning image recognition model.

[0108] To achieve the above objectives, the present invention also provides an image recognition and evaluation system for the effectiveness of rural domestic sewage treatment based on deep learning, such as... Figures 2-3As shown, the system is used to implement the deep learning-based image recognition and evaluation method for the effectiveness of rural domestic sewage treatment; the system specifically includes:

[0109] The first data processing unit is used to construct a problem dataset corresponding to the objects associated with the governance effectiveness assessment; wherein, the associated objects include the sewage accumulation status in the village's public spaces, the type of accumulated materials in drainage ditches, the water status of ponds, the water inlet status of facilities, the functional status of facility treatment units, and the water outlet status of facilities;

[0110] The second data processing unit is used to generate and acquire first data corresponding to each associated object, and to label the type of the target object in the first data, and to create corresponding training and validation sets; wherein, the first data includes problematic positive sample and non-problematic negative sample image data;

[0111] The first data generation unit is used to build a deep learning-based image recognition model, optimize the model parameters through the training set and validation set, and generate second data corresponding to each associated object according to the image recognition model; wherein, the second data is specific type recognition result data;

[0112] The second data generation unit is used to score the second data according to a preset scoring rule and generate third data corresponding to the second data; wherein the third data is data that comprehensively evaluates the governance effectiveness based on the scores of various related objects.

[0113] The specific types of the associated objects include:

[0114] The village's public spaces are divided into alleys / main roads / front and back of houses with no water accumulation, with clear water accumulation, and with sewage accumulation;

[0115] Drainage ditches are classified into dry ditches, clear water ditches, and ditches that accumulate domestic sewage / kitchen waste / silt.

[0116] The pond is divided into two states: normal water quality and black water quality; the facility water intake is divided into three states: no water, clear water, and sewage intake.

[0117] The facility treatment units are divided into normal constructed wetlands, severely backwatered constructed wetlands, constructed wetlands with severe vegetation loss, normal reaction tanks, clear reaction tanks, normal main structures, and leaking main structures, etc.

[0118] The facility's water output is classified into three states: no water output, normal water output, and black water output.

[0119] And / or, the first data generation unit further includes:

[0120] The first processing module is used to construct a deep learning-based image recognition model for evaluating the effectiveness of rural sewage treatment, and to extract and fuse image features using a convolutional neural network.

[0121] The first generation module is used to generate specific type data corresponding to the objects reflected in the image; wherein, the number of output layer nodes matches the total number of specific types of associated objects;

[0122] And / or, the second data generation unit further includes:

[0123] The second generation module is used to create scoring rules corresponding to governance effectiveness, and generate fourth data corresponding to governance effectiveness based on the scoring rules; wherein, the fourth data is governance effectiveness evaluation data;

[0124] The third generation module is used to generate fifth data corresponding to the governance effectiveness based on the fourth data and the governance effectiveness review process; wherein, the fifth data is the data of poor governance effectiveness evaluation results;

[0125] The second processing module is used to retrain and process the fifth data based on the training set and in combination with the corresponding training model.

[0126] The scoring rules are as follows: for the identification result of each associated object, 1 point is assigned if there is no problem and 0 points are assigned if there is a problem; the scores of the 6 associated objects are multiplied together, and the evaluation result is good when the total score is 1 point and poor when it is 0 points; when the evaluation result is poor, the specific type of all associated objects with a score of 0 is output simultaneously.

[0127] In the system solution embodiment of the present invention, the specific details of the method steps involved in the image recognition evaluation of the effectiveness of rural domestic sewage treatment based on deep learning have been described above. That is to say, the functional modules in the system are used to implement the steps or sub-steps in the above method embodiment, and will not be repeated here.

[0128] To achieve the above objectives, the present invention also provides an image recognition and evaluation platform for the effectiveness of rural domestic sewage treatment based on deep learning, such as... Figure 5 As shown, the system includes a processor, a memory, and a control program for a deep learning-based image recognition and evaluation platform for the effectiveness of rural domestic sewage treatment. The processor executes the deep learning-based image recognition and evaluation platform control program, which is stored in the memory. This control program implements the steps of the deep learning-based image recognition and evaluation method for the effectiveness of rural domestic sewage treatment. For example:

[0129] S1. Construct a problem dataset corresponding to the objects associated with the governance effectiveness assessment; wherein, the associated objects include the sewage accumulation status in village public spaces, the type of accumulated material in drainage ditches, the water status of ponds, the water inlet status of facilities, the functional status of facility treatment units, and the water outlet status of facilities;

[0130] S2. Generate and acquire first data corresponding to each associated object, and label the type of the target object in the first data, and create corresponding training and validation sets; wherein, the first data includes problematic positive sample and non-problematic negative sample image data;

[0131] S3. Build a deep learning-based image recognition model, optimize the model parameters using the training set and validation set, and generate second data corresponding to each associated object based on the image recognition model; wherein, the second data is specific type recognition result data;

[0132] S4. The second data is scored according to the preset scoring rules, and a third data corresponding to the second data is generated; wherein, the third data is the data for evaluating the governance effectiveness by comprehensively considering the scores of each related object.

[0133] The specific details of the steps have been explained above and will not be repeated here.

[0134] In this embodiment of the invention, the built-in processor of the deep learning-based rural domestic sewage treatment effectiveness image recognition and evaluation platform can be composed of integrated circuits. For example, it can be composed of a single packaged integrated circuit, or multiple integrated circuits with the same or different functions, including combinations of one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and various control chips. The processor connects to various components using various interfaces and lines, and executes programs or units stored in memory, as well as calling data stored in memory, to perform various functions of deep learning-based rural domestic sewage treatment effectiveness image recognition and evaluation and to process data.

[0135] The memory is used to store program code and various data. It is installed in the deep learning-based image recognition and evaluation platform for the effectiveness of rural domestic sewage treatment and enables high-speed and automatic access to programs or data during operation.

[0136] The memory includes read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, disk storage, magnetic tape storage, or any other computer-readable medium capable of carrying or storing data.

[0137] This invention constructs a problem dataset corresponding to objects associated with the evaluation of governance effectiveness through a method. The associated objects include the sewage accumulation status in village public spaces, the type of accumulated material in drainage ditches, the water status of ponds, the water inlet status of facilities, the functional status of facility treatment units, and the water outlet status of facilities. First data corresponding to each associated object is generated and acquired, and type labels for target objects in the first data are labeled. Corresponding training and validation sets are created. The first data includes image data of positive samples with problems and negative samples without problems. A deep learning-based image recognition model is built, and the model parameters are optimized using the training and validation sets. Based on the image recognition model, second data corresponding to each associated object is generated. The second data is specific type identification result data. The second data is scored according to preset scoring rules, and third data corresponding to the second data is generated. The third data is a comprehensive evaluation of governance effectiveness based on the scores of each associated object. The invention also includes a system and platform corresponding to the method, which can solve the problems of high professional requirements for assessors, poor consistency of results, low efficiency, and high cost faced by existing agricultural pollution control effectiveness evaluation methods. In other words, it achieves efficient, low-cost, and highly consistent evaluation of agricultural pollution control effectiveness.

[0138] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A deep learning-based image recognition evaluation method for the effectiveness of rural domestic sewage treatment, characterized in that, The method includes the following steps: Construct a dataset of problem images corresponding to the objects associated with the governance effectiveness assessment; wherein, the associated objects include the sewage accumulation status in village public spaces, the type of accumulated material in drainage ditches, the water status of ponds, the water inlet status of facilities, the functional status of facility treatment units, and the water outlet status of facilities; Generate and acquire first data corresponding to each associated object, label the type of the target object in the first data, and create corresponding training and validation sets; wherein, the first data includes problematic positive sample and non-problematic negative sample image data; A deep learning-based image recognition model is constructed, and the model parameters are optimized using the training set and validation set. Based on the image recognition model, second data corresponding to each associated object is generated; wherein, the second data is specific type recognition result data. The second data is scored according to the preset scoring rules, and a third data corresponding to the second data is generated; wherein, the third data is a comprehensive evaluation of the governance effectiveness based on the scores of various related objects.

2. The image recognition and evaluation method for the effectiveness of rural domestic sewage treatment based on deep learning as described in claim 1, characterized in that, The specific types of the associated objects include: The village's public spaces are divided into alleys / main roads / front and back of houses with no water accumulation, with clear water accumulation, and with sewage accumulation; Drainage ditches are classified into dry ditches, clear water ditches, and ditches that accumulate domestic sewage / kitchen waste / silt. The pond is divided into two states: normal water quality and black water quality; the facility water intake is divided into three states: no water, clear water, and sewage intake. The facility treatment units are divided into normal constructed wetlands, severely backwatered constructed wetlands, constructed wetlands with severe vegetation loss, normal reaction tanks, clear reaction tanks, normal main structures, and leaking main structures. The facility's water output is categorized into three states: no water output, normal water output, and black water output.

3. The image recognition and evaluation method for the effectiveness of rural domestic sewage treatment based on deep learning as described in claim 1, characterized in that, The step of building a deep learning-based image recognition model, optimizing model parameters using the training and validation sets, and generating second data corresponding to each associated object based on the image recognition model, further includes: A deep learning-based image recognition model for evaluating the effectiveness of rural sewage treatment was constructed, and convolutional neural networks were used to extract and fuse image features. Generate specific type data corresponding to the objects reflected in the image; wherein, the number of output layer nodes matches the total number of specific types of the associated objects.

4. The image recognition and evaluation method for the effectiveness of rural domestic sewage treatment based on deep learning as described in claim 1, characterized in that, The step of scoring the second data according to a preset scoring rule and generating third data corresponding to the second data further includes: A scoring rule corresponding to governance effectiveness is created, and a fourth data point corresponding to governance effectiveness is generated based on the scoring rule; wherein, the fourth data point is governance effectiveness evaluation data; Based on the fourth data and combined with the governance effectiveness review process, a fifth data corresponding to the governance effectiveness is generated; wherein, the fifth data is the data indicating poor governance effectiveness assessment results; Based on the training set and in conjunction with the corresponding training model, the fifth data is retrained and processed.

5. A method for evaluating the effectiveness of rural domestic sewage treatment based on deep learning, as described in claim 1 or 4, characterized in that... The scoring rules are as follows: For the identification result of each associated object, 1 point is assigned if there is no problem and 0 points are assigned if there is a problem. Multiply the scores of the 6 related objects together. A total score of 1 is considered good, and 0 is considered poor. When the evaluation result is unsatisfactory, the specific type of all associated objects with a score of 0 will be output synchronously.

6. A deep learning-based image recognition and evaluation system for the effectiveness of rural domestic sewage treatment, characterized in that, The system is used to implement the image recognition and evaluation method for the effectiveness of rural domestic sewage treatment based on deep learning, as described in any one of claims 1 to 5; the system includes: The first data processing unit is used to construct a problem dataset corresponding to the objects associated with the governance effectiveness assessment; wherein, the associated objects include the sewage accumulation status in the village's public spaces, the type of accumulated materials in drainage ditches, the water status of ponds, the water inlet status of facilities, the functional status of facility treatment units, and the water outlet status of facilities; The second data processing unit is used to generate and acquire first data corresponding to each associated object, and to label the type of the target object in the first data, and to create corresponding training and validation sets; wherein, the first data includes problematic positive sample and non-problematic negative sample image data; The first data generation unit is used to build a deep learning-based image recognition model, optimize the model parameters through the training set and validation set, and generate second data corresponding to each associated object according to the image recognition model; wherein, the second data is specific type recognition result data; The second data generation unit is used to score the second data according to a preset scoring rule and generate third data corresponding to the second data; wherein the third data is data that comprehensively evaluates the governance effectiveness based on the scores of various related objects.

7. The image recognition and evaluation system for the effectiveness of rural domestic sewage treatment based on deep learning, as described in claim 6, is characterized in that... The specific types of the associated objects include: The village's public spaces are divided into alleys / main roads / front and back of houses with no water accumulation, with clear water accumulation, and with sewage accumulation; Drainage ditches are classified into dry ditches, clear water ditches, and ditches that accumulate domestic sewage / kitchen waste / silt. The pond is divided into two states: normal water quality and black water quality; the facility water intake is divided into three states: no water, clear water, and sewage intake. The facility treatment units are divided into normal constructed wetlands, severely backwatered constructed wetlands, constructed wetlands with severe vegetation loss, normal reaction tanks, clear reaction tanks, normal main structures, and leaking main structures. The facility's water output is classified into three states: no water output, normal water output, and black water output. And / or, the first data generation unit further includes: The first processing module is used to construct a deep learning-based image recognition model for evaluating the effectiveness of rural sewage treatment, and to extract and fuse image features using a convolutional neural network. The first generation module is used to generate specific type data corresponding to the objects reflected in the image; wherein, the number of output layer nodes matches the total number of specific types of associated objects; And / or, the second data generation unit further includes: The second generation module is used to create scoring rules corresponding to governance effectiveness, and generate fourth data corresponding to governance effectiveness based on the scoring rules; wherein, the fourth data is governance effectiveness evaluation data; The third generation module is used to generate fifth data corresponding to the governance effectiveness based on the fourth data and the governance effectiveness review process; wherein, the fifth data is the data of poor governance effectiveness evaluation results; The second processing module is used to retrain and process the fifth data based on the training set and in combination with the corresponding training model.

8. A deep learning-based image recognition and evaluation system for the effectiveness of rural domestic sewage treatment, as described in claim 6 or 7, is characterized in that... The scoring rules are as follows: For the identification result of each associated object, 1 point is assigned if there is no problem and 0 points are assigned if there is a problem. Multiply the scores of the 6 related objects together. A total score of 1 is considered good, and 0 is considered poor. When the evaluation result is unsatisfactory, the specific type of all associated objects with a score of 0 will be output synchronously.

9. A deep learning-based image recognition and evaluation platform for the effectiveness of rural domestic sewage treatment, characterized in that, The system includes a processor, a memory, and a control program for a deep learning-based image recognition and evaluation platform for the effectiveness of rural domestic sewage treatment. The processor executes the deep learning-based image recognition and evaluation platform control program, which is stored in the memory. This deep learning-based image recognition and evaluation platform control program implements the deep learning-based image recognition and evaluation method for the effectiveness of rural domestic sewage treatment as described in any one of claims 1 to 5.

Images