Gridding smart park comprehensive management method and system

Through grid layout of sensors and data processing technology, the problem of difficulty in cleaning hidden garbage in park management is solved, efficient monitoring and cleaning of green belts is achieved, and the ecological environment and management efficiency of the park is improved.

CN119026877AInactive Publication Date: 2024-11-26JINAN GAOPIN WEIYE INFORMATION TECH CO LTD
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN202411505185.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-28
Publication Date
2024-11-26
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The existing park management methods are difficult to clean up hidden garbage in the green belt in a timely manner, resulting in ecological environment damage and inefficient management.

Method used

By pre-acquisitioning the greening distribution and area of ​​the park, grid layout sensors monitor garbage entering the green belt in real time, collect and process garbage monitoring data and hidden garbage survey data, integrate the first sampling data and the second sampling data, locate the garbage location, and get close to it Allocation of sanitation workers for cleaning and sorting recycling.

Benefits of technology

It has achieved timely cleaning of hidden garbage in the park's green belt, improved the ecological environment, improved the efficiency and service quality of park management, and built a more comfortable, efficient and green smart park.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119026877A_ABST
    Figure CN119026877A_ABST
Patent Text Reader

Abstract

The invention discloses a gridding smart park comprehensive management method and system, and relates to the technical field of park management, the smart park comprehensive management method comprises the following steps: obtaining park greening distribution and area in advance, and arranging sensors in a gridding manner to monitor garbage entering a green belt in real time; garbage monitoring data and hidden garbage survey data are collected and processed to obtain garbage azimuth data, and then the first sampling data and the second sampling data are fused to obtain a garbage sampling set; according to the garbage sampling set, sanitation workers are distributed nearby for timely cleaning, and the garbage is classified and recycled, so that hidden garbage is subjected to sound wave positioning, the sanitation workers can conveniently and timely clean the hidden garbage in the greening area in the park, the ecological environment of the greening area is improved and maintained, and the environmental protection effect is improved. And moreover, comprehensive management in the park can be well matched, and a more comfortable, efficient and green smart park is constructed.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of park management technology, and in particular to a grid-based smart park comprehensive management method and system. Background Art

[0002] In modern park management, grid-based integrated management of smart parks is very common. This management method divides the park into multiple grid areas, and uses technologies such as the Internet of Things (IoT), big data, and cloud computing to conduct refined management and intelligent services for resources and activities within each grid, aiming to improve operational efficiency and service quality and promote the sustainable development of the park. However, park management often overlooks the management of green areas, especially the garbage hidden under the dense branches and leaves of green belts. These garbage are usually difficult to find and are not cleaned up for a long time, which not only damages the ecological environment of the green area, but also poses a threat to park management.

[0003] In response to traditional park management methods, the present invention obtains the green distribution and area of ​​the park in advance, and deploys sensors in a grid to monitor the entry of garbage into the green belt in real time; then collects and processes garbage monitoring data and hidden garbage survey data to obtain garbage location data for analysis to determine the direction of the next experimental operation.

[0004] The first sampling data collected is merged with the second sampling data to obtain a garbage sampling set; and according to the garbage sampling set, sanitation workers are assigned nearby to clean up in time, and the garbage is classified and recycled to further output the detection results.

[0005] To sum up, how to promptly clean up the hard-to-find hidden garbage in the green belts within the park is an issue that needs to be urgently solved and optimized in the grid-based smart park comprehensive management system. Summary of the invention

[0006] The present invention provides a grid-based integrated management method and system for a smart park, which solves the technical problem of timely cleaning up hidden garbage that is difficult to find in green belts within a park in the related art.

[0007] In order to solve the above technical problems, the present invention provides a grid-based smart park comprehensive management method and system, and the specific technical solutions are as follows:

[0008] In a first aspect, a grid-based integrated management method for a smart park includes the following steps:

[0009] Obtain the distribution and area of ​​green areas in the park in advance, and deploy sensors in a grid-like manner to monitor the entry of garbage into the green belt in real time;

[0010] Collecting garbage monitoring data and hidden garbage survey data and processing them to obtain garbage location data, specifically including: collecting monitoring data in real time based on pre-deployed sensors to obtain first sampling data; further surveying hidden or hard-to-find garbage based on the first sampling data to obtain second sampling data; performing location analysis on the first sampling data and the second sampling data respectively to obtain garbage location data;

[0011] fusing the first sampling data with the second sampling data to obtain a garbage sampling set;

[0012] According to the garbage sampling set, sanitation workers are assigned nearby to clean up the garbage in a timely manner, and the garbage is sorted and recycled.

[0013] As a further optimization scheme of the present invention, the greening distribution and area of ​​the park are obtained in advance, and sensors are deployed to monitor the garbage entering the green belt in real time, including:

[0014] Obtain a detailed map of the park in advance to obtain the park's topography and green areas;

[0015] Organize professionals to conduct on-site surveys based on the acquired park terrain and greening areas to obtain greening information data;

[0016] Multiple sensors are deployed in a grid-like manner according to greening information data and sensor detection range to obtain real-time monitoring data;

[0017] Capture data on garbage entering green belts based on real-time monitoring data and transmit it to the management system;

[0018] The management system establishes a database storage to store the monitoring data transmitted in real time.

[0019] As a further optimization scheme of the present invention, real-time monitoring data is collected based on pre-deployed sensors to obtain first sampling data, including:

[0020] Grid the park's green areas and deploy sensors based on the shape and area of ​​the green belts;

[0021] The acquisition sensor monitors the green belt data in real time to obtain obvious garbage image data;

[0022] Preprocessing the obvious junk image data to obtain preprocessed data;

[0023] The pre-processed data is identified to obtain first sampling data.

[0024] As a further optimization scheme of the present invention, further surveying hidden or hard-to-find garbage based on the first sampling data to obtain second sampling data includes:

[0025] Based on the first sampling data, a sprinkler truck is used to spray water droplets to simulate the sound wave information data of raindrops hitting an object;

[0026] The sound wave data generated by the water droplets sprayed by the sprinkler hitting the object is marked as quiet environment sound wave data;

[0027] Then the sound wave information data of natural raindrops hitting objects is marked as noise environment sound wave data;

[0028] Acquire the sound wave data at multiple points in two different environments, quiet and noisy, to obtain the initial sampling data set;

[0029] The initial sampling data is subjected to anti-interference processing to obtain second sampling data.

[0030] As a further optimization scheme of the present invention, the initial sampling data is subjected to anti-interference processing to obtain second sampling data, including:

[0031] Process the quiet environment sound wave data to obtain clear sound wave data;

[0032] Perform modal decomposition on the sound wave data of the noise environment. According to the noise environment and the modal correlation coefficient R, γ is preset as the interference threshold. The mode with R<γ is considered as noise interference, that is, the modal component with R≧γ is selected to reconstruct the useful signal to obtain the useful sound wave signal;

[0033] Reconstruct the useful sound wave signal to achieve signal noise reduction, and then use the relative delay between the reconstructed signals to obtain the flight time estimation of the sound wave during transmission in a noisy environment;

[0034] Suppress unnecessary acoustic signal interference in a noisy environment based on flight time estimation to obtain clear acoustic data;

[0035] The acoustic wave clear data is combined with the acoustic wave clear data to obtain second sampling data.

[0036] As a further optimization scheme of the present invention, the first sampling data and the second sampling data are respectively subjected to azimuth analysis to obtain garbage azimuth data, including:

[0037] Separating the obvious garbage image data in the preprocessed data to obtain the obvious garbage shape area;

[0038] Extract edge frame information of obvious garbage shape area to obtain obvious garbage contour data;

[0039] Target positioning is performed based on the obvious garbage contour area to obtain the location data of obvious garbage;

[0040] Identifying the second sampling data and extracting the acoustic wave characteristics of the hidden garbage to obtain abnormal acoustic wave characteristic data;

[0041] Based on the characteristic data of abnormal sound waves, multi-point acoustic positioning is carried out to obtain the location data of hidden garbage.

[0042] As a further optimization scheme of the present invention, multi-base acoustic positioning is performed according to the characteristic data of the abnormal sound waves to obtain the position data of the hidden garbage, including:

[0043] Acquire multiple acoustic wave points horizontally, and arrange the multiple acoustic wave points in a ring shape to obtain acoustic wave information data at multiple bases;

[0044] Constructing a surround sound wave information model based on multi-point sound wave information;

[0045] Identify the preliminary location of the abnormal sound wave data points existing in the multi-point sound wave data by using the ambient sound wave data model;

[0046] Extending the initial positions of multiple abnormal sound wave information data to obtain the convergence point of the abnormal sound wave information data;

[0047] The location of the abnormal sound wave information data is obtained according to the abnormal sound wave information data gathering point to obtain the specific location data of the hidden garbage.

[0048] As a further optimization scheme of the present invention, sanitation workers are assigned to clean up the garbage in time according to the garbage sampling set, and the garbage is classified and recycled, including:

[0049] Get the location of each garbage based on the garbage sampling set, and get the distance between each garbage and each worker based on the current location of the sanitation worker: ; In the formula, (x i ,y i ) is the location coordinate of the garbage, (x k ,y k ) are the worker coordinates;

[0050] Match based on distance combined with worker capabilities, and choose the closest distance: ; In the formula, x ki is a binary variable. When worker W k Assigned to garbage C i When x ki =1, otherwise 0; p is the total number of workers, q is the total number of garbage, which dynamically regulates the garbage allocation tasks for workers;

[0051] Workers are assigned to clean up garbage based on proximity, and the collected garbage is sorted and recycled.

[0052] In a second aspect, the system provides an electronic device including a memory, a processor, and a grid-based smart park comprehensive management method program stored in the memory and executable on the processor. When the grid-based smart park comprehensive management method program is executed by the processor, the steps of the grid-based smart park comprehensive management method are implemented.

[0053] Thirdly, a grid-based integrated management system for smart parks includes:

[0054] Monitoring module: It is used to obtain the distribution and area of ​​green areas in the park in advance, and to deploy sensors in a grid-like manner to monitor the entry of garbage into the green belt in real time;

[0055] Data processing module: It is used to collect garbage monitoring data and hidden garbage survey data and process them to obtain garbage location data;

[0056] Data fusion module: used for fusing the first sampling data with the second sampling data to obtain a garbage sampling set;

[0057] Execution management module: It is used to assign sanitation workers to clean up the garbage in a timely manner according to the garbage sampling set, and to classify and recycle the garbage.

[0058] The present invention has at least the following beneficial effects: the present invention obtains the greening distribution and area of ​​the park in advance, and deploys sensors in a grid to monitor the entry of garbage into the green belt in real time; collects garbage monitoring data and hidden garbage survey data and processes them to obtain garbage location data, and then integrates the first sampling data and the second sampling data to obtain a garbage sampling set; according to the garbage sampling set, sanitation workers are assigned nearby to clean up in time, and the garbage is classified and recycled, so that the hidden garbage is located by using sound waves, which can facilitate the timely cleaning of hidden garbage in the greening area of ​​the park, so that the ecological environment of the greening area is improved and maintained, and it can cooperate well with the comprehensive management within the park to build a more comfortable, efficient and green smart park. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 It is a flowchart of a grid-based smart park comprehensive management method provided by an embodiment of the present invention;

[0060] Figure 2 It is a schematic diagram of a grid-based smart park comprehensive management system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0061] The present application is further described in detail below in conjunction with the accompanying drawings. It is necessary to point out here that the following specific implementation methods are only used to further illustrate the present application and cannot be understood as limiting the scope of protection of the present application. Technical personnel in this field can make some non-essential improvements and adjustments to the present application based on the above application content.

[0062] Example 1

[0063] like Figure 1 , Figure 2 As shown, a grid-based smart park comprehensive management method and system, the specific implementation method is as follows:

[0064] like Figure 1 As shown, the present invention provides a grid-based smart park comprehensive management method, comprising the following steps:

[0065] Step 1: Obtain the distribution and area of ​​green areas in the park in advance, and deploy sensors in a grid-like manner to monitor garbage entering the green belt in real time;

[0066] Step 2, collecting garbage monitoring data and hidden garbage survey data and processing them to obtain garbage location data, specifically also includes: step 21, collecting monitoring data in real time based on pre-deployed sensors to obtain first sampling data; step 22, further surveying hidden or hard-to-find garbage based on the first sampling data to obtain second sampling data; step 23, performing location analysis on the first sampling data and the second sampling data respectively to obtain garbage location data;

[0067] Step 3, fusing the first sampling data with the second sampling data to obtain a garbage sampling set;

[0068] Step 4: Assign sanitation workers to clean up the garbage in a timely manner based on the garbage sampling set, and classify and recycle the garbage.

[0069] It should be noted that you should check and adjust each acoustic monitoring camera sensor in advance, install the camera according to the instructions, and fix it in an appropriate position, and ensure that the camera covers the required monitoring area and that the camera is aimed at the area to be monitored. If the camera supports manual focus, turn the focus ring until the image is clear. If the camera supports manual aperture adjustment, adjust the aperture size according to the lighting conditions. A smaller aperture (larger f-number) is usually used during the day, and a larger aperture (smaller f-number) is used at night to get more light;

[0070] Adjust the HD resolution to 1080p (1920x1080) and set it to 30fps to obtain a smooth video effect. Usually use the automatic exposure mode (AE) as the initial setting. If the image is too bright or too dark, you can adjust the exposure compensation value appropriately, usually between 2EV. Select the automatic white balance (AWB) as the default setting. Manually adjust the white balance to around 5500K under specific lighting conditions, and set it to around 6500K on a cloudy day. If there is a strong backlight source (such as the sun) in the monitoring scene, turn on the BLC function to maintain the clarity of the foreground object. Adjust the BLC level according to actual needs, usually gradually adjust from low to high until the foreground is clear. If there is a significant lighting contrast in the monitoring scene, turn on the WDR function and adjust the WDR level according to actual needs, usually gradually adjust from low to high until both the bright and dark areas in the image are clearly visible;

[0071] In low light conditions, increase the gain to improve the image brightness. However, be aware that too high a gain will increase image noise. Night settings are usually adjusted between 0dB and 18dB, with contrast set to 50% and saturation set to 50% to obtain natural color performance. Adjusting the sharpness can make the image edges clearer, but too high a sharpness can cause image distortion. It is usually set to 50% of the same level. If the camera supports infrared night vision, make sure the infrared light is automatically turned on at night. Adjust the infrared light intensity according to actual needs. Generally, it is set to a medium level. Observe the image quality through the monitor to check if there is any need for further adjustment. Make minor adjustments to the above parameters based on the test results until a satisfactory effect is achieved. If the camera has a protective cover, ensure that it is installed correctly, fix the camera on the bracket, and ensure that it is stable and reliable. Configure the camera's IP address, subnet mask, gateway and other network parameters to set the video storage method, whether it is uploaded to the cloud or saved on the local hard disk. Use wide dynamic monitoring to adapt to the strong light changes outdoors, and use the strong light suppression function to reduce the halo when the sun is directly shining. Finally, adjust the contrast and saturation for better visual effects.

[0072] In a preferred embodiment of the present invention, step 1 uses GIS (geographic information system) to obtain accurate distribution and area data of each green belt in the park, so as to determine the range and number of sensor deployment, divide multiple green belt areas of the entire park into grids in a straight line, set a straight line length threshold, and divide each green area into at least two networks according to the threshold. An acoustic wave monitoring camera sensor is deployed in each grid, and considering that the shape of the green area in the park is irregular, usually involving green areas at corners, it is necessary to additionally deploy an acoustic wave monitoring camera sensor at the corner, and determine the monitoring range of the acoustic wave monitoring camera sensor according to the area of ​​each network: , where R iis the coverage radius of the ith acoustic wave surveillance camera sensor, A grid is the area of ​​each grid. A monitoring network covering the entire green belt is formed. The acoustic wave monitoring camera sensor is used to detect the entry of garbage into the green belt in real time. The grid layout ensures that there is no blind spot in the entire park green belt, making the monitoring of garbage more accurate. The acoustic wave monitoring camera sensor captures the entry of garbage in real time, ensuring that the management personnel can discover and deal with garbage problems in time.

[0073] Then, in step 2, garbage monitoring data is collected in real time based on the acoustic wave monitoring camera sensor, and the initial detected garbage location is recorded. These data constitute the first sampling data set. Based on the first sampling, the second sampling data set is obtained by surveying hidden or difficult-to-find garbage. This usually involves using more sophisticated sensors or higher-resolution monitoring equipment to perform azimuth analysis on the collected first sampling data and second sampling data to determine the precise location of the garbage. The two samplings allow hidden garbage to be effectively detected, which improves the comprehensiveness of garbage monitoring. The azimuth analysis ensures the accurate identification of the garbage location and provides reliable data support for subsequent cleaning work.

[0074] According to the first sample data and the second sample data obtained in step 2, the first sample data and the second sample data are standardized in step 3 so that the two sets of data have the same scale, and .... ;x f,k is the kth sample feature after fusion, x' 1,k and x' 2,k are the kth standardized features of the first sample data and the second sample data, respectively, and w 1 and w 2 is the corresponding data set weight, satisfying w 1 +w 2 =1, the two sets of sampling data are fused to generate a comprehensive garbage sampling set. This sampling set contains all detected garbage location information and related features; data fusion combines multiple sampling results to form a more comprehensive and accurate garbage sampling set, which improves the utilization value of data, reduces data redundancy through fusion, and makes subsequent processing more efficient.

[0075] According to the garbage sampling set in step 3 and the garbage location data recorded in step 2, the distance between the garbage point and the sanitation worker is obtained, and the workers are assigned to clean up using the principle of proximity. During the cleaning process, the sanitation workers classify the garbage by type and place them in different garbage collection containers. The classified garbage is recycled and sent to the corresponding recycling center or processing plant. The allocation of sanitation workers nearby minimizes the time and distance of workers' movement and improves the efficiency of garbage cleaning. Garbage classification and recycling contribute to the rational use of resources and reduce garbage disposal costs and the impact on the environment.

[0076] In a preferred embodiment of the present invention, the above step 1 further includes:

[0077] Step 11, obtaining a detailed map of the park in advance to obtain the park terrain and green areas;

[0078] Step 12: Organize professional personnel to conduct on-site surveys based on the acquired park greening areas to obtain greening information data;

[0079] Step 13, multiple sensors are gridded according to the greening information data and the sensor detection range to obtain real-time monitoring data;

[0080] Step 14, capturing data of garbage entering the green belt according to real-time monitoring data, and transmitting it to the management system;

[0081] Step 15: The management system establishes a database to store the real-time transmitted monitoring data.

[0082] In the embodiment of the present invention, detailed map data of the park is obtained by using GIS (geographic information system) or technologies such as drones and satellite images in step 11: , where M represents the park map, T represents the terrain data, V represents the green area data, and f is the terrain parameter. The map includes the terrain of the park (such as terrain height, building distribution, etc.) and the green area (such as the specific location of lawns, trees, and flower beds). The map data is analyzed to identify and mark the green areas in the park, providing a reference for the subsequent deployment of acoustic wave monitoring camera sensors. By obtaining a detailed park map, the accuracy of the basic data for subsequent work is ensured, and the blind spots caused by the blind deployment of sensors are avoided. The accuracy and comprehensiveness of the map data help to achieve full coverage monitoring of the green area. Based on the terrain and green area in the park map obtained in step 11, professional personnel are organized to conduct on-site surveys of the park through step 12. The survey content includes the shape of the green belt and the specific types of lawns, flower beds, shrubs, vegetation density, vegetation health status, soil type and terrain changes. The greening information data obtained from the survey is: , where G represents the greening information dataset, L nrepresents the location of the nth green area, S n The characteristic data representing the nth green area (such as plant species, coverage area, terrain slope, etc. of each area) is recorded in detail and compared and supplemented with the map data obtained previously. The on-site survey data can effectively verify and supplement the map data, ensure the authenticity and integrity of the data, obtain more detailed greening information, and provide a reliable reference for the deployment of acoustic wave monitoring camera sensors.

[0083] According to the greening information data obtained in step 12, select the acoustic wave monitoring camera sensor suitable for monitoring garbage entering the green belt through step 13, divide the entire green area into several grids, and deploy one or more sensors in each grid according to the detection range of the sensor to ensure full coverage monitoring of the entire green belt. The sensor is deployed at a specified position according to the plan, and the height, angle, sensitivity and other parameters of the sensor are set. The grid-based deployment of sensors can achieve all-round monitoring of the green belt, reduce blind spots, and ensure that each green area can be monitored in real time. And through reasonable planning and deployment of sensors, the situation of garbage entering the green belt can be accurately captured, and the monitoring effect is improved; according to step 13, after the acoustic wave monitoring camera sensor is deployed, the system passes: ; Where D represents the real-time monitoring data set, P i is the garbage location coordinate, T i is the time when garbage enters, S i is the corresponding sensor number, which monitors the situation of the green belt in real time. When the sensor detects that garbage has entered the green belt, the system records the location, time and other relevant information of the garbage, and transmits the captured garbage data to the management system in real time through wireless communication or wired network. The transmitted data includes the specific location coordinates of the garbage entering, the detection time and the sensor number. Real-time monitoring and data transmission enable management personnel to grasp the situation of garbage entering the green belt at the first time, so as to quickly arrange the cleaning work. The real-time transmitted data can be stored in the system in time to avoid data loss or delay, and ensure the continuity of the monitoring work; finally, based on step 14, the management system in step 15 designs a database according to the structure of the monitoring data. The sound wave information of water droplets hitting the branches and leaves of the green belt is entered into the system in advance for storage. The database table includes fields such as garbage location information, timestamp, sensor number, data type, etc. The monitoring data transmitted to the management system in real time is automatically stored in the database, which is expressed as ; In the formula, DB represents database, P i is the location coordinate of the garbage, T i is the time when garbage enters, S i is the corresponding sensor number, F iIt is the information related to sanitation workers. The database is backed up regularly to ensure the security and integrity of the data. In order to improve the efficiency of data access, the database indexes key fields (such as location and time) to facilitate subsequent data query and analysis. The database systematically manages and stores monitoring data, making data access and analysis more convenient. The automatic storage and backup mechanism of the data ensures the security of the monitoring data and prevents data loss or damage.

[0084] In a preferred embodiment of the present invention, the above step 21 may include:

[0085] Step 211, gridding the park green area, and deploying sensors according to the shape and area of ​​the green belt;

[0086] Step 212, collecting sensors to monitor green belt data in real time to obtain obvious garbage image data;

[0087] Step 213: pre-process the obvious junk image data to obtain pre-processed data.

[0088] In the embodiment of the invention, according to the park map and green area information obtained in step 11, the green belt is divided into several grids in step 2: Where A grid,l is the area of ​​each grid, A total is the total area of ​​the green belt, n is the number of grids, f(shape, density) is the correction factor considering the shape of the green belt and the density of vegetation. The size and shape of the grid are adjusted according to the specific form of the green belt (such as width and curvature) to ensure complete coverage. The layout planning of the acoustic monitoring camera sensor is personalized according to the area and shape of each grid. For example, in a narrow area or a corner area, more sensors may need to be deployed for coverage, while the number of sensors can be reduced in a wide area. The location and density of the sensor layout should ensure that there are no dead angles, and take into account the actual vegetation density and monitoring needs. Through personalized layout of sensors, ensure that the sensors can adapt to the characteristics of different green belt shapes so that the monitoring range is not missed. Adjust the number and location of sensors according to the specific situation of the grid, effectively utilize sensor resources, and avoid excessive or insufficient layout; after the sensors deployed in step 21 are started in step 22, they capture the environmental changes in the green belt in real time, especially the garbage with obvious volume, through: ; In the formula, Capture (i, j) is a binary variable, which is 1 when garbage is detected and 0 otherwise. i is the coverage radius of the sensor, D i,jIndicates the distance of garbage detection, thereby being able to automatically detect and capture the existence of garbage and record image data. When the sensor detects that garbage has entered the green belt, it automatically captures or records image data of the corresponding area. These data usually include information such as the shape, size and location of garbage. Real-time monitoring can timely capture the situation of garbage entering the green belt to ensure the immediacy of data. Through image data, the type and location of garbage can be intuitively identified, providing clear information for subsequent processing. Then, step 213 is used to denoise the collected garbage image data, remove background noise and redundant information in the image, improve the clarity and effectiveness of the image, and then enhance the image to make the garbage area more prominent, which is convenient for subsequent identification and classification, and then separate the garbage area from the background. Through segmentation, a clear garbage outline and position can be obtained, and then key features such as the shape, color, size, texture, etc. of garbage are extracted from the preprocessed image. These features will be used as input data for subsequent classification or processing. Through preprocessing, unnecessary noise and interference are removed, so that the quality of the garbage image is improved. The preprocessed image data is clearer and more structured, which helps to improve the accuracy of garbage identification and classification.

[0089] In a preferred embodiment of the present invention, the above step 22 specifically includes:

[0090] Step 221, based on the first sampling data, use a sprinkler truck to spray water droplets to simulate the sound wave information data of raindrops hitting an object;

[0091] Step 222, marking the sound wave data generated by the water droplets sprayed by the sprinkler hitting the object as quiet environment sound wave data;

[0092] Step 223, the sound wave information data of natural rainfall water droplets hitting the object is marked as noise environment sound wave data;

[0093] Step 224, performing multi-point acquisition of sound wave data in two different environments, quiet and noisy, to obtain an initial sampling data set;

[0094] Step 225: perform anti-interference processing on the initial sampling data to obtain second sampling data.

[0095] In the embodiment of the invention, based on the first sampling data (basic environmental data) obtained in step 21, a sprinkler truck is used to spray water droplets in the green area through step 221 to simulate the sound waves of natural raindrops hitting the surface of an object. The sprinkler truck simulates rainfall conditions of different intensities and densities by controlling the amount of water and the spraying method, thereby generating sound wave data similar to natural rainfall, and the sound waves of water droplets hitting an object appear in a green grid area, and there are no obvious sound waves of water droplets hitting an object in the surrounding area. At this time, when the sprinkler truck sprays water droplets, the sound wave monitoring camera sensor is used to monitor multiple The acoustic wave information of water droplets hitting objects is collected at the position, and the data characteristics such as time, frequency, amplitude, etc. of these acoustic waves are recorded. The water droplets sprayed by the sprinkler provide a controllable experimental environment, which can simulate different types of rainfall conditions and facilitate the analysis and research of acoustic wave characteristics. The sprinkler can quickly generate a large amount of simulated acoustic wave data, which is particularly suitable for acoustic wave collection and testing under conditions of long-term drought and no rain. According to step 221, step 222 is used for the acoustic wave data generated by the water droplets sprayed by the sprinkler hitting the object. Since the background environment is relatively quiet and controllable, therefore, by: Where D quiet is a quiet environment sound wave dataset, S sim (t) is the simulated water droplet hitting sound wave signal of the sprinkler truck within time t, and N is the number of sound wave data samples, which is marked as "quiet environment sound wave data". This label is used to distinguish the sound wave data generated in different environments. By classifying and archiving the collected quiet environment sound wave data, a clear data structure is provided for subsequent analysis and processing. Environmental labeling of sound wave data helps to distinguish data in different environments in subsequent analysis and reduce the complexity of analysis. Sound wave data marked as quiet environment are usually less subject to external interference, which helps to accurately study the sound wave characteristics generated when water droplets hit objects.

[0096] Based on the different simulation environments in step 222, the sound wave data generated by natural rainfall is collected in an uncontrollable real environment, which may contain various natural noises, such as wind sound, sound waves of raindrops hitting objects in all areas, and footsteps of people walking in water, etc. These sound wave data collected under natural rainfall conditions are: Where D noise is a noise environment sound wave dataset, S real(t) is the natural rainfall water droplet impact sound wave signal within time t, m is the number of noise data samples, marked as "noise environment sound wave data" to distinguish it from the previous quiet environment data. Because the collection of sound waves from natural rainfall saves water resources and does not require artificial watering of green areas, but natural education is in a noisy environment, the generated sound wave data can better reflect complex situations, which helps to improve the adaptability of the system in real scenarios. By comparing the data of quiet environment and noisy environment, the impact of noise on sound wave signals can be studied, and useful sound wave signals can be obtained through effective noise suppression technology; according to step 222 and step 223, in two different environments, the sound wave monitoring camera sensor is used to synchronously collect sound wave data to ensure that the multi-dimensional changes of sound waves in space can be captured. The data collected by each sound wave monitoring camera sensor reflects the characteristics of the sound wave at a specific position and time point in the green area. By integrating the sound wave data at different points and times, an initial sampling data set is formed, which provides a comprehensive data basis for subsequent processing and analysis. Multi-point acquisition provides rich spatial and temporal dimension data, which is helpful for in-depth analysis of the propagation characteristics of sound waves in different environments. Multi-point acquisition can avoid the data uniformity and insufficiency that may exist in single-point acquisition, ensuring the integrity and representativeness of the data. The initial sampling data of step 224 is subjected to anti-interference processing through step 225 to reduce or eliminate the influence of noise. The data after anti-interference processing is the second sampling data, which represents more reliable sound wave information after processing. Through anti-interference processing, the noise and interference signals in the initial data are effectively suppressed, which improves the clarity and availability of the data. The second sampling data is more accurate and can better reflect the true characteristics of the sound wave, providing a solid foundation for subsequent analysis and application.

[0097] In another preferred embodiment of step 22 of the present invention, based on the first sampling data obtained by the preliminary capture of obvious garbage, these data can preliminarily identify the interference sources or special environmental factors that may exist in the monitoring area, and according to the features identified by the first sampling data, multiple sensors are arranged at different positions to further collect the sound wave information generated by water droplets hitting objects such as water spraying or rain from a sprinkler truck, and the sensors are arranged at different points to obtain the various sound wave features of water spraying or raindrops hitting objects at these points, by: Where S multi (t) is the multi-point acoustic wave data set collected by all sensors within time t, S i,j(t) represents the effective sound wave data scanned by the sensor within time t when i=1 and j=1, n is the number of sensors, and m is the number of different types of sound waves (such as the sound waves of water droplets hitting branches and leaves and the sound waves of garbage of various materials). Multi-point collection of sound wave information generated by sprinklers or rainfall can further identify and distinguish the differences between natural sounds in the environment and other possible interference factors (such as the sound of garbage falling) to prevent false alarms. Multi-point collection ensures that sound wave information is obtained from multiple angles, which is conducive to the comprehensive analysis of environmental factors and enriches the data source;

[0098] According to the acquired sound wave information data, a large amount of environmental noise contained in the collected sound wave information data is preprocessed. In the preprocessing stage, the sound wave data is first denoised to remove noise of irrelevant frequencies, retain the sound wave frequency bands related to the sprinkler or raindrops hitting objects, and enhance the denoised signal: , where S” i,j (t) is the enhanced signal within time t, G() is the enhancement function, S' i,j It is the signal within the unenhanced time t, which can enhance the characteristic components in the signal and highlight the frequency and amplitude of characteristic sound waves (such as the sound of water droplets hitting), which is convenient for subsequent analysis, through: Where, F i,j is the feature vector extracted from the jth acoustic wave signal at the i-th sensor position. ExtractFeatures() is a feature extraction function, which extracts key features from the processed acoustic wave data, such as spectrum features, energy distribution, and time domain features. These features will serve as the basis for the initial sampling data. De-noising greatly reduces the impact of environmental noise on the analysis results and improves the accuracy of the data. By extracting representative acoustic wave features, the data structure is simplified, making subsequent analysis more efficient and accurate.

[0099] Finally, pattern recognition is performed on the initial sampling data: , where It is the type of sound wave to be identified (such as the sound waves of water droplets hitting different objects generated by a sprinkler or rainfall). Classifier() is a classification model, which is used to classify features, identify specific sound wave patterns and distinguish them from other background sounds, and classify the initial sampling data by type. For example, the sound waves of water droplets hitting branches and leaves are separated from the sound waves hitting garbage (including hidden garbage), and noise signals irrelevant to these events are filtered out, so as to facilitate the comparison of the two sounds later, so as to obtain more stable and accurate second sampling data. Through in-depth analysis of the initial sampling data, different types of sound wave data can be accurately identified and classified, the accuracy of the data can be ensured, and irrelevant noise can be filtered out, thereby improving the quality of the sampling data and providing a reliable basis for subsequent decision-making and processing.

[0100] In a preferred embodiment of the present invention, the above step 225 specifically includes:

[0101] Step 2251, processing the quiet environment sound wave data to obtain clear sound wave data;

[0102] Step 2252, performing modal decomposition on the noise environment sound wave data, according to the noise environment and the modal correlation coefficient R, γ is preset as the interference threshold, and the mode with R<γ is considered as noise interference, that is, the modal component with R≧γ is selected to reconstruct the useful signal to obtain the useful sound wave signal;

[0103] Step 2253, reconstructing the useful sound wave signal to achieve signal noise reduction, and then using the relative delay between the reconstructed signals to obtain the flight time estimation during the transmission of the sound wave in the noisy environment;

[0104] Step 2254, suppressing the interference of redundant acoustic wave signals in a noisy environment according to the flight time estimation to obtain clear acoustic wave data;

[0105] Step 2255, combining the sound wave clear data with the sound wave clear data to obtain second sampling data.

[0106] In the embodiment of the invention, the acquired initial sampling data is subjected to anti-interference processing through step 225, and specifically, the sound wave data collected in a quiet environment is processed through step 2251. Since the environment is relatively quiet, there is less noise in the data, and the main goal of the processing is to eliminate small noise interference and improve the signal quality. Commonly used methods include bandpass filtering, wavelet denoising, etc., which perform signal enhancement on the cleaned sound wave data to increase the strength of the useful signal, making the signal clearer and more recognizable. By processing the sound wave data in a quiet environment, a clear sound wave signal is obtained to provide high-quality basic data for subsequent analysis and processing. After eliminating small noise, the error in the analysis is reduced, and the accuracy of the signal is ensured; at the same time, the sound wave data in a noisy environment is subjected to modal decomposition processing through step 2252 to separate modal components of different frequencies. Specifically, each modal component is initialized to α=1000 and the decomposition layer number k=2, and the original signal is decomposed. By decomposing the original signal, several modal signals are extracted, and each modal signal corresponds to a different frequency component, through: Where R k is the correlation coefficient of the kth mode; Corr() is the correlation coefficient calculation function; U k(t) is the noise interference parameter of the kth mode within time t, and the correlation coefficient R between each frequency component and the original signal after decomposition is obtained. It is judged whether R<γ is established. If not, k=k+1 is set, and the previous operation is returned to continue to calculate the correlation coefficient. If it is established, the current k is regarded as the optimal layer k, where γ is a preset threshold. Finally, the decomposition layer is fixed at k, and 1 / R is used as the fitness function. An optimized penalty factor α is obtained by using the particle swarm algorithm. It is considered that the mode contains useful signal components and a useful sound wave signal is obtained. Therefore, through mode decomposition and selection, the interference components in the noise environment can be effectively suppressed and useful sound wave signals can be extracted. The process of mode selection makes signal processing more accurate and reduces the influence of unnecessary noise components on the analysis results. The useful sound wave signal selected in step 2252 is passed through step 2253: Where S useful (t) is the useful sound wave signal reconstructed within time t, γ is the interference threshold, kwhere is the index variable, indicating the number of the modal signal, k is the kth modal signal, which indicates the value of the kth component within time t, and the signal is reconstructed accordingly to further remove the residual noise components, making the signal clearer. Through noise reduction processing, redundant background noise (such as the sound of rain hitting objects outside the green area, the sound of people stepping on the water, etc.) can be eliminated to improve the signal quality, and the relative delay between the reconstructed useful signals can be used: ; In the formula, Δt TOF is the estimated flight time, t recv is the time it takes for the sensor to receive the sound wave signal, t emit It is the time of sound wave emission when raindrops hit an object, thus obtaining the estimated flight time of sound waves in a noisy environment. The flight time estimation is used to analyze the time from the emission to the reception of sound waves, which helps to further suppress noise. The noise is further removed through signal reconstruction to obtain a purer useful signal and improve the accuracy of analysis. The flight time estimation provides the transmission characteristic data of sound waves in a noisy environment, which is helpful for subsequent noise suppression and signal processing.

[0107] Using the flight time estimate calculated in step 2253, the propagation path and time characteristics of the sound wave in the noise environment are analyzed in step 2254 to identify and suppress the interference of redundant noise signals. This step usually involves a filtering technology based on the flight time to exclude noise signals that do not meet the propagation characteristics, by: Where S clean is the clear sound wave data after noise suppression, S useful is the useful acoustic wave signal reconstructed before, Δt TOFis the flight time estimation, Suppress() is the noise suppression function, which suppresses the noise signal interference and obtains the final clear sound wave data. Through flight timing analysis and noise suppression, the final obtained sound wave clear data is more accurate, reducing the impact of noise on data quality. Through noise suppression, the effective signal is more prominent, and the efficiency of subsequent signal recognition and processing is improved; finally, through step 2255, the sound wave clear data from the quiet environment (obtained in step 2251) and the sound wave clear data in the noisy environment (obtained in step 2254) are integrated. The integration process may include weighted averaging, feature fusion and other methods to ensure that data in different environments can be effectively combined to form more comprehensive sound wave information. After integration processing, the second sampling data is finally formed. This data set integrates useful sound wave information in different environments, provides reliable data support for further analysis or practical application, and obtains a more comprehensive and representative sound wave data set by integrating data in quiet and noisy environments.

[0108] In a preferred embodiment of the present invention, the above step 23 specifically includes:

[0109] Step 231, separating the obvious garbage image data in the pre-processed data to obtain an obvious garbage shape area;

[0110] Step 232, extracting edge frame information of the obvious garbage shape area to obtain obvious garbage contour data;

[0111] Step 233, target positioning is performed according to the obvious garbage contour area to obtain the location data of the obvious garbage;

[0112] Step 234, identifying the hidden junk sound wave information data and extracting the sound wave characteristics of the hidden junk to obtain sound wave characteristic data;

[0113] Step 235, performing multi-point acoustic positioning based on the acoustic wave characteristic data to obtain the location data of the hidden garbage.

[0114] In the embodiment of the invention, the garbage image pre-processed in step 21 is separated by step 231, and the obvious garbage area is separated from the background. Common image segmentation methods include threshold segmentation, edge detection, region growth, Kmeans clustering, and semantic segmentation based on deep learning (such as UNet model). After segmentation, the shape area belonging to garbage in the image is extracted, irrelevant background information is removed, and only the pixel area that may be garbage is retained. Through image segmentation, the noise and interference information in the background are removed, making the subsequent shape analysis more accurate, and a large amount of irrelevant information in the original image is eliminated, and only the part related to garbage is retained, which greatly reduces the scale of processing data; according to step 231, the edge of the separated garbage shape area is extracted by using edge detection algorithms such as Sobel and Canny. The edge detection algorithm identifies the shape edge according to the rate of change of the gray value in the image, and generates a boundary box (BoundingBox) surrounding the garbage area according to the extracted edge data. The edge frame is usually represented by a rectangle, a minimum circumscribed circle or a polygon. The specific choice depends on the shape and distribution of the garbage. Extracting the edge frame information can accurately describe the outer contour of the garbage and provide basic data for subsequent target positioning and classification. The edge frame data can also be used for shape analysis and feature extraction, such as judging the size of the garbage, the degree of irregularity of the shape, etc.; According to the edge frame information extracted in step 232, the coordinates of the center point of the garbage in the image are obtained: , where (x c ,y c ) is the coordinate of the center point of the garbage in the image, (x min ,y min )、(x max ,y max ) is the vertex coordinate of the edge box. Then, through the mapping relationship between the image and the actual scene, the position of the garbage in the actual space (orientation data) is determined, and the image coordinates are converted into actual world coordinates using the internal and external parameters of the sensor: ; In the formula, (X W , Y W , Z W ) is the world coordinate of the garbage in the real space, R is the rotation matrix of the camera, T is the translation matrix, K is the camera intrinsic matrix, (x c ,y c ,1) is the three-dimensional coordinates of the image, which realizes the accurate positioning of the garbage. ,Through coordinate conversion, the spatial position of the garbage in the park can be ,accurately determined, which provides a basis for the subsequent garbage cleaning and ,treatment. Target positioning can achieve real-time positioning, so that the system can ,respond quickly and arrange sanitation workers for cleaning.

[0115] According to the acoustic information data pre-processed in step 22, the acoustic signal generated by water droplets hitting hidden garbage (such as garbage hidden deep in the grass or under leaves) is captured by using a sensor in step 234: ; In the formula, s is the captured sound wave signal, A is the amplitude of the sound wave, f is the frequency of the sound wave, t is the current time, φ is the phase, and n(t) is the noise component within time t. Through pattern recognition, these signals are distinguished from the sound waves of water droplets hitting branches and leaves, and key features are extracted from the identified sound waves: , where F is the acoustic wave eigenvector, f n Represents the nth acoustic wave feature (such as frequency, amplitude, duration, spectral energy distribution, etc.). These features can describe the acoustic wave characteristics of hidden garbage. Through acoustic wave feature extraction, hidden garbage that is difficult to detect directly by visual means can be accurately identified. Acoustic wave information can complement visual data to form a more comprehensive garbage monitoring system and reduce missed detections. Then, by selecting multiple acoustic position points, acoustic wave signals at different locations are collected, and the time difference and amplitude difference of these signals are used for positioning. Common positioning algorithms include the arrival time difference method and the triangulation method. Based on the sensor position and the acoustic wave propagation model, the specific location of the hidden garbage is obtained, and its spatial coordinates are finally determined.

[0116] In a preferred embodiment of the present invention, the above step 235 may include:

[0117] Step 2351, horizontally acquiring multiple acoustic wave points, and arranging the multiple acoustic wave points in a ring shape, so as to obtain acoustic wave information data of multiple base points;

[0118] Step 2352, constructing a surround sound wave information model according to the multi-base sound wave information;

[0119] Step 2353, identifying the preliminary position of the abnormal sound wave information point where the multi-point sound wave information data exists according to the ambient sound wave information model;

[0120] Step 2354, extending the preliminary positions of the plurality of abnormal sound wave information data to obtain a convergence point of the abnormal sound wave information data;

[0121] Step 2355, obtain the position of the abnormal sound wave information data according to the abnormal sound wave information data gathering point to obtain the specific location data of the hidden garbage.

[0122] In the embodiment of the invention, based on the multiple sound position points obtained in step 235, two or more sensors are deployed in the target area in step 2351, and each sensor captures the sound wave signal at a different point, and multiple different sound position points are selected: , where P i represents the spatial three-dimensional coordinates of the i-th acoustic wave point, (x i ,y i, z i ) is the three-dimensional coordinate of the sensor's position. Each sound position point is arranged in a horizontal ring: Where, P j is the spatial three-dimensional coordinate of the sound position point, R is the radius of the circular arrangement, θ i is the position angle of the ith sound position base, and z is the height of the sound position base (assuming that all sensors are at the same height), so as to ensure that sound wave information can be captured from different angles. By arranging multiple sound position bases in a horizontal ring, it can be ensured that the captured sound wave information has a 360-degree coverage range. This all-round collection of sound position bases helps to collect different sound wave positions at different sound source positions and obtain valuable sound wave signals. The circularly arranged sensors can achieve all-round coverage of the target area, and the multi-point horizontal layout can obtain sound wave information from different directions, which helps to improve the accuracy of subsequent positioning; based on step 2351, the sound wave data collected from multiple points through step 2352 are integrated, and a surround sound wave information model is constructed through comprehensive analysis of the multi-point sound wave data: ; In the formula, S (P i , t) represents the sound wave information collected at point P, A i is the amplitude of the sound wave, f i is the frequency of the sound wave; φ i is the phase of the ith sound wave point, n i (t) is the noise recording of the sound wave characteristics of the ith point within the time t, such as arrival time, amplitude, and frequency. The model can simulate the direction of the sound waves between different points, especially capture the propagation path of the sound waves. The surround sound wave information model can perceive the direction of the sound waves in three dimensions from multiple directions, which helps to accurately determine the location of the sound source of the sound waves. Therefore, by using the comprehensive analysis of multi-point data, the model can automatically correct the errors caused by environmental interference and improve the accuracy of the model. The surround sound wave information model constructed in step 2353 is used to analyze the sound wave data of multiple points and identify the emission direction of the sound waves. When the sound waves are usually manifested as different amplitudes and frequencies at certain points, the directions of the sound waves are preliminarily determined according to the characteristic differences of the sound waves at different points. This is achieved by analyzing the time difference or amplitude difference of the sound waves propagating in the annular model. The sound waves from different directions can be effectively detected, providing a data basis for subsequent precise positioning. At the same time, the preliminary direction judgment can provide preliminary directional indications for subsequent precise positioning and narrow the search range.

[0123] Step 2354 is used to linearly extend the multiple initial positions of the abnormal sound waves identified in step 2353: , where L i From point P i The extended path, is the direction vector of sound wave propagation, λ i is the scale factor of path extension. By simulating its propagation path in space, the propagation paths of the abnormal sound waves at different points will form an intersection in space. The intersection point of the extended paths is used to determine the convergence point of the abnormal sound wave information data. The convergence point is the intersection of multiple propagation paths, representing the possible source position of the abnormal sound wave. By extending and intersecting the azimuths of multiple points, the actual source of the abnormal sound wave can be more accurately focused on, and the positioning accuracy can be improved. The determination of the intersection point can eliminate the influence caused by the data error of individual points, and obtain a more reliable sound source position; finally, through step 2355, the convergence point of step 2354 is used as the intersection position of the abnormal sound wave, and the abnormal sound wave is further compared with the sound wave characteristics of water droplets hitting branches and leaves, and the specific position of the hidden garbage is further accurately located. Considering various interference factors in the propagation of sound waves (such as reflection, attenuation, etc.), the specific position of the convergence point is calibrated to ensure that the final azimuth data is accurate and reliable, and finally the specific position of the hidden garbage is determined: , where P 垃圾 is the final location of hidden garbage, P 汇集 It is the convergence point of the abnormal sound waves, and δP is the position calibration correction parameter, so that the subsequent cleaning work can be targeted and reduce omissions. By comprehensively considering various possible interference factors, the position is calibrated to ensure the high accuracy of the positioning results.

[0124] In a preferred embodiment of the present invention, the above step 3 may include:

[0125] Step 31, synchronizing the first sampling data with the second sampling data to obtain synchronization data D1 and D2;

[0126] Step 32, standardizing the synchronization data D1 and D2 so that D1 and D2 have the same scale;

[0127] Step 33, extracting the features of the standardized synchronous data D1 and D2 and fusing them to form a garbage sampling set;

[0128] Step 34: verify and adjust the garbage sampling set to obtain a final garbage sampling set Df.

[0129] In the embodiment of the present invention, the preliminary garbage monitoring obtained in step 21, such as preliminary image data, sound wave data, etc., is used as the first sampling data through step 31; and the second sampling data in step 22 is the fine data further collected based on the preliminary monitoring results. The two sets of data may have certain differences in time or space. In order to effectively compare and merge the two sets of data, data synchronization is required. The synchronization process includes time synchronization: ; where t D1,i is the timestamp of the ith sample of the first synchronous sampling data D1, tD2,j is the timestamp of the jth sample of the second synchronous sampling data D2, Δt is the time difference, which is corrected and spatially synchronized by interpolation or time series alignment method: Where, P D1,i is the spatial coordinate of the i-th sample in the first synchronous sampling data D1, P D2,j is the spatial coordinate of the jth sample in the second synchronous sampling data D2, R is the rotation matrix, and T is the translation matrix. To ensure that the first synchronous sampling data D1 and the second synchronous sampling data D2 can be compared and processed under the same time axis or spatial framework, data synchronization can eliminate the time or space differences between different data sets, so that subsequent analysis is more consistent. Through synchronization processing, the information in the two data sets can be more accurately combined, thereby improving the accuracy of data fusion; according to step 31 and then through step 32, the synchronized data have the same scale, and the two data sets D1 and D2 need to be standardized. The purpose of standardization is to eliminate the dimensional differences between the data so that data from different sources can be compared and analyzed at the same scale. Mean standard deviation standardization or minimum and maximum standardization is usually used. Mean standard deviation standardization adjusts the mean and standard deviation of the data so that the data is concentrated near zero and has the same variance. Standardization eliminates the scale differences between different data so that they can be directly combined. Through standardization, the data distribution can be made more stable, which is helpful for subsequent feature extraction and model training.

[0130] Perform feature extraction based on the standardized synchronous data D1 and D2 obtained in step 32: Where, F D1 and F D2are feature vectors extracted from data sets D1 and D2 respectively. ExtractFeatures() is a feature extraction function that extracts key features that are helpful for garbage identification and classification. Features may include the spatial location, shape features, sound features, etc. of garbage. The features extracted from the two data sets are fused. The fusion method can be weighted average, principal component analysis (PCA) or more complex deep learning methods. The fused data will form a comprehensive garbage sampling set for subsequent analysis and processing. Feature fusion enables data from different sources to complement each other and form a more complete description of garbage, which helps to improve the accuracy of detection and classification. Through feature extraction and fusion, data redundancy can be reduced, processing complexity can be reduced, and key information can be retained. Finally, the fused garbage sampling set is verified through step 34 to detect its accuracy and representativeness. Verification includes comparison with real data, accuracy test using classifiers, etc. According to the verification results, the data in the sampling set is adjusted. This includes removing abnormal samples, adjusting feature weights, resynchronizing or standardizing data, etc., to further improve the quality of the data set. After the adjustment is completed, the final garbage sampling set Df will be formed as the basic data for subsequent analysis and decision-making. Through verification and adjustment, problems in the data can be effectively discovered and corrected, and the overall quality of the data set can be improved. The adjusted data set can improve the stability and accuracy of the model, and provide reliable data support for subsequent classification, detection and other tasks.

[0131] In another preferred embodiment of the present invention, the above step 4 may include:

[0132] Step 41, obtain the location of each garbage based on the garbage sampling set, and obtain the distance between each garbage and each worker according to the current location of the sanitation worker: ; In the formula, (x i ,y i ) is the location coordinate of the garbage, (x k ,y k ) are the worker coordinates;

[0133] Step 42, match based on distance and worker capabilities, and select the closest distance: ; In the formula, x ki is a binary variable. When worker W k Assigned to garbage C i When x ki =1, otherwise 0; p is the total number of workers, q is the total number of garbage, which dynamically regulates the garbage allocation tasks for workers;

[0134] Step 43, assigning workers to clean up garbage based on proximity, and classifying and recycling the acquired garbage.

[0135] In the embodiment of the present invention, based on the garbage sampling set in step 3, multiple adjacent or dense garbage points are clustered by a clustering algorithm (such as Kmeans or DBSCAN) in step 41, and each garbage represents a group of garbage points that are relatively close in space. Where, P cluster,j is the location of the jth garbage, P garbage,i is the location of the i-th garbage point, n j is the number of garbage points in the jth garbage. The geometric center of each garbage is calculated, that is, the location of the garbage. The geometric center can be expressed as the average value of the locations of all garbage points in the cluster. The current location of each sanitation worker is obtained, and the distance between each garbage and each sanitation worker is calculated. This can be achieved through formulas such as Euclidean distance or Manhattan distance. By identifying garbage, sanitation workers can concentrate on handling garbage in an area and improve cleaning efficiency. Obtaining the distance between workers and garbage helps to achieve dynamic response and ensure that cleaning tasks are completed in a timely and efficient manner.

[0136] The sanitation workers' capabilities are evaluated in step 42, including the amount of garbage they can handle (such as weight or volume) and their working status (such as whether they are idle, current task load, etc.). This information is combined with the distance data obtained in step 41 to reasonably allocate garbage cleaning tasks, and match each garbage with each worker through an optimization algorithm (such as the Hungarian algorithm or the greedy algorithm): Where, C k,j is an element of the matching matrix, indicating whether worker k is assigned to process garbage j, D k,j is the distance between worker k and garbage j, L k is the processing capacity (such as maximum load) of worker k, W j is the amount of garbage j, and threshold is the distance threshold, which determines whether the task can be accepted. Usually, the worker who is closest or has sufficient processing capacity temporarily (for example, the worker himself does not carry much garbage) is given priority to handle a certain garbage. Task allocation is dynamic. The system will continuously adjust the allocation strategy according to the real-time distance and worker status to ensure that the task can be completed in time. The matching algorithm can ensure that each sanitation worker can handle the most suitable garbage in the shortest time, improve overall work efficiency, and dynamic control ensures the system's ability to respond quickly to emergencies, such as timely adjustment of tasks when workers complete tasks or encounter new garbage;

[0137] Finally, according to the task assignment in step 42, the sanitation workers go to the garbage assigned to them and perform garbage cleaning work. During the cleaning process, the workers classify the garbage through step 43. Garbage is usually divided into recyclable garbage, non-recyclable garbage, hazardous garbage, kitchen waste, etc. The classification work can be done manually by the workers, or it can be combined with portable smart devices or sensors to assist in classification. After the classification is completed, different types of garbage are placed in corresponding collection containers, and recyclable garbage is sent to recycling points or resource recycling centers. Workers can quickly respond to tasks assigned by the system to ensure that garbage cleaning work is carried out in a timely manner. Through classification and recycling, the amount of landfill can be effectively reduced, and the opportunity for resource reuse can be increased, which has significant environmental benefits.

[0138] like Figure 2 As shown, a grid-based smart park integrated management system includes:

[0139] Monitoring module 201: It is used to obtain the distribution and area of ​​green areas in the park in advance, and to deploy sensors in a grid-like manner to monitor the entry of garbage into the green belt in real time;

[0140] Data processing module 202: used to collect garbage monitoring data and hidden garbage survey data and process them to obtain garbage location data;

[0141] Data fusion module 203: used for fusing the first sample data with the second sample data to obtain a garbage sample set;

[0142] Execution management module 204: It is used to assign sanitation workers to clean up the garbage in a timely manner according to the garbage sampling set, and to classify and recycle the garbage.

[0143] An embodiment of the present invention further provides an electronic device including a memory, a processor, and a grid-based smart park comprehensive management method program stored in the memory and executable on the processor based on a grid-based smart park comprehensive management method and system. When the grid-based smart park comprehensive management method program is executed by the processor, the steps of the grid-based smart park comprehensive management method are implemented. All implementation methods in the above method embodiments are applicable to this embodiment and can achieve the same technical effect.

[0144] Those skilled in the art will also appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0145] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0146] In the embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0147] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0148] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0149] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, ROM, RAM, magnetic disks, or optical disks.

[0150] In addition, it should be pointed out that in the apparatus and method of the present invention, it is obvious that each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent schemes of the present invention. Moreover, the steps of performing the above series of processing can naturally be performed in chronological order according to the order of description, but it is not necessary to perform them in chronological order, and some steps can be performed in parallel or independently of each other. For those of ordinary skill in the art, it can be understood that all or any steps or components of the method and apparatus of the present invention can be implemented in hardware, firmware, software or a combination thereof in any computing device (including processors, storage media, etc.) or a network of computing devices, which can be achieved by those of ordinary skill in the art using basic programming skills after reading the description of the present invention.

[0151] Therefore, the purpose of the present invention can also be achieved by running a program or a group of programs on any computing device. The computing device can be a well-known general device. Therefore, the purpose of the present invention can also be achieved by simply providing a program product containing a program code that implements the method or device. That is to say, such a program product also constitutes the present invention, and the storage medium storing such a program product also constitutes the present invention. Obviously, the storage medium can be any well-known storage medium or any storage medium developed in the future. It should also be pointed out that in the device and method of the present invention, it is obvious that each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent schemes of the present invention. In addition, the steps of performing the above-mentioned series of processing can naturally be performed in chronological order according to the order of description, but it is not necessary to perform them in chronological order. Some steps can be performed in parallel or independently of each other.

[0152] The above describes an embodiment of the present invention, but this embodiment is not limited to the above-mentioned specific implementation mode. The above-mentioned specific implementation mode is merely illustrative and not restrictive. Under the guidance of this embodiment, ordinary technicians in this field can also make more forms of equivalent embodiments, all of which are within the protection of this embodiment.

[0153] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention.

Claims

1. A grid-based integrated management method for a smart park, characterized in that: The following steps are involved: Obtain the distribution and area of ​​green areas in the park in advance, and deploy sensors in a grid-like manner to monitor the entry of garbage into the green belt in real time; Collecting garbage monitoring data and hidden garbage survey data and processing them to obtain garbage location data, specifically including: collecting monitoring data in real time based on pre-deployed sensors to obtain first sampling data; further surveying hidden or hard-to-find garbage based on the first sampling data to obtain second sampling data; performing location analysis on the first sampling data and the second sampling data respectively to obtain garbage location data; fusing the first sampling data with the second sampling data to obtain a garbage sampling set; According to the garbage sampling set, sanitation workers are assigned nearby to clean up the garbage in a timely manner, and the garbage is sorted and recycled.

2. A grid-based smart park integrated management method according to claim 1, characterized in that: Obtain the distribution and area of ​​green areas in the park in advance, and deploy sensors to monitor garbage entering the green belt in real time, including: Obtain a detailed map of the park in advance to obtain the park's topography and green areas; Organize professionals to conduct on-site surveys based on the acquired park terrain and greening areas to obtain greening information data; Multiple sensors are deployed in a grid-like manner according to greening information data and sensor detection range to obtain real-time monitoring data; Capture data on garbage entering green belts based on real-time monitoring data and transmit it to the management system; The management system establishes a database storage to store the monitoring data transmitted in real time.

3. A grid-based smart park integrated management method according to claim 2, characterized in that: The monitoring data is collected in real time based on the pre-deployed sensors to obtain the first sampling data, including: Grid the park's green areas and deploy sensors based on the shape and area of ​​the green belts; The acquisition sensor monitors the green belt data in real time to obtain obvious garbage image data; Preprocessing the obvious junk image data to obtain preprocessed data; The pre-processed data is identified to obtain first sampling data.

4. A grid-based integrated management method for smart parks according to claim 3, characterized in that: Based on the first sampling data, further survey hidden or hard-to-find garbage to obtain second sampling data, including: Based on the first sampling data, a sprinkler truck is used to spray water droplets to simulate the sound wave information data of raindrops hitting an object; The sound wave data generated by the water droplets sprayed by the sprinkler hitting the object is marked as quiet environment sound wave data; Then the sound wave information data of natural raindrops hitting objects is marked as noise environment sound wave data; Acquire the sound wave data at multiple points in two different environments, quiet and noisy, to obtain the initial sampling data set; The initial sampling data is subjected to anti-interference processing to obtain second sampling data.

5. A grid-based integrated management method for smart parks according to claim 4, characterized in that: The initial sampling data is subjected to anti-interference processing to obtain second sampling data, including: Process the quiet environment sound wave data to obtain clear sound wave data; Perform modal decomposition on the sound wave data of the noise environment. According to the noise environment and the modal correlation coefficient R, γ is preset as the interference threshold. The mode with R<γ is considered as noise interference, that is, the modal component with R≧γ is selected to reconstruct the useful signal to obtain the useful sound wave signal; Reconstruct the useful sound wave signal to achieve signal noise reduction, and then use the relative delay between the reconstructed signals to obtain the flight time estimation of the sound wave during transmission in a noisy environment; Suppress unnecessary acoustic signal interference in a noisy environment based on flight time estimation to obtain clear acoustic data; The acoustic wave clear data is combined with the acoustic wave clear data to obtain second sampling data.

6. A grid-based integrated management method for smart parks according to claim 5, characterized in that: Performing position analysis on the first sampling data and the second sampling data respectively to obtain garbage position data, including: Separating the obvious garbage image data in the preprocessed data to obtain the obvious garbage shape area; Extract edge frame information of obvious garbage shape area to obtain obvious garbage contour data; Target positioning is performed based on the obvious garbage contour area to obtain the location data of obvious garbage; Identifying the second sampling data and extracting the acoustic wave characteristics of the hidden garbage to obtain abnormal acoustic wave characteristic data; Based on the characteristic data of abnormal sound waves, multi-point acoustic positioning is carried out to obtain the location data of hidden garbage.

7. A grid-based integrated management method for a smart park according to claim 6, characterized in that: Based on the characteristic data of abnormal sound waves, multi-base acoustic positioning is carried out to obtain the location data of hidden garbage, including: Acquire multiple acoustic wave points horizontally, and arrange the multiple acoustic wave points in a ring shape to obtain acoustic wave information data at multiple bases; Constructing a surround sound wave information model based on multi-point sound wave information; Identify the preliminary location of the abnormal sound wave data points existing in the multi-point sound wave data by using the ambient sound wave data model; Extending the initial positions of multiple abnormal sound wave information data to obtain the convergence point of the abnormal sound wave information data; The location of the abnormal sound wave information data is obtained according to the abnormal sound wave information data gathering point to obtain the specific location data of the hidden garbage.

8. A grid-based integrated management method for smart parks according to claim 7, characterized in that: According to the garbage sampling set, sanitation workers are assigned to clean up the garbage in a timely manner, and the garbage is classified and recycled, including: Get the location of each garbage based on the garbage sampling set, and get the distance between each garbage and each worker based on the current location of the sanitation worker: ; In the formula, (x i ,y i ) is the location coordinate of the garbage, (x k ,y k ) are the worker coordinates; Match based on distance combined with worker capabilities, and choose the closest distance: ; In the formula, x ki is a binary variable, when a worker is assigned to a garbage, x ki =1, otherwise 0; p is the total number of workers, q is the total number of garbage, which dynamically regulates the garbage allocation tasks for workers; Workers are assigned to clean up garbage based on proximity, and the collected garbage is sorted and recycled.

9. A grid-based integrated management system for smart parks, characterized in that: The system sets up an electronic device including a memory, a processor and a grid-based smart park comprehensive management method program stored in the memory and executable on the processor. When the grid-based smart park comprehensive management method program is executed by the processor, the steps of the grid-based smart park comprehensive management method are implemented.

10. A grid-based integrated management system for smart parks according to claim 9, characterized in that: The system includes: Monitoring module: It is used to obtain the distribution and area of ​​green areas in the park in advance, and to deploy sensors in a grid-like manner to monitor the entry of garbage into the green belt in real time; Data processing module: It is used to collect garbage monitoring data and hidden garbage survey data and process them to obtain garbage location data; Data fusion module: used for fusing the first sampling data with the second sampling data to obtain a garbage sampling set; Execution management module: It is used to assign sanitation workers to clean up the garbage in a timely manner according to the garbage sampling set, and to classify and recycle the garbage.

Citation Information

Patent Citations

  • Floating garbage removal ship and method

    CN108557002A

  • Garbage classification method and system based on ultrasonic sensor arrays

    CN109622416A

  • 5G-based smart park ground garbage identification method and system

    CN116416573A

  • Metering calibration method and system for sound source identification positioning system

    CN118731848A

  • Water surface garbage cleaning robot based on ultrasonic wave and visual identification

    CN221091176U