Intelligent cleaning strategy planning method and system based on fusion feature analysis

Abstract

The application discloses a wisdom cleaning strategy planning method and system based on fusion feature analysis, relates to the field of wisdom environment, solves the problems of low resource utilization efficiency and poor cleaning effect of the existing cleaning strategy planning method, and comprises the following steps: S1, acquiring cleaning planning basic data; S2, analyzing the cleaning planning basic data to obtain cleaning planning analysis data; S3, planning the cleaning frequency of a cleaning sub-region according to the cleaning planning analysis data; and S4, planning the real-time cleaning of the cleaning sub-region according to the cleaning planning basic data.The application can effectively improve the cleaning quality and the utilization efficiency of cleaning resources by dynamically adjusting the cleaning frequency through the acquisition of data with different characteristics, and can make the cleaning strategy timely respond to environmental changes by planning the real-time cleaning according to real-time image data.

Description

Technical Field

[0001] This invention belongs to the field of smart environment and relates to feature fusion technology, specifically a smart cleaning strategy planning method and system based on fusion feature analysis. Background Technology

[0002] Smart cleaning refers to the use of technologies such as the Internet of Things, artificial intelligence, and big data to intelligently manage and optimize the cleaning process. It collects environmental data through devices such as sensors and cameras, and combines intelligent algorithms to analyze and process the data to achieve functions such as intelligent scheduling, intelligent monitoring, and intelligent prediction, thereby improving cleaning efficiency, saving costs, and improving service quality.

[0003] Existing cleaning strategy planning methods have the following shortcomings:

[0004] 1. Existing cleaning strategy planning methods are mostly based on historical data and static rules, lacking dynamic monitoring and adjustment of real-time environmental data, resulting in cleaning strategies failing to respond to environmental changes in a timely manner;

[0005] 2. Existing cleaning strategy planning methods often only consider a single data source or feature, lacking the integration and analysis of multi-source data, and cannot comprehensively assess the cleaning situation and formulate effective strategies.

[0006] To address this, we propose a smart cleaning strategy planning method and system based on fusion feature analysis. Summary of the Invention

[0007] To address the shortcomings of existing technologies, the purpose of this invention is to provide a smart cleaning strategy planning method and system based on fusion feature analysis. This invention is based on acquiring basic cleaning planning data, calculating cleaning frequency planning coefficients corresponding to the first type of sub-region and the second type of sub-region respectively based on the basic cleaning planning data, obtaining cleaning planning analysis data, acquiring the upper limit and lower limit of the benchmark interval of the cleaning frequency planning coefficient, comparing them with the cleaning frequency planning coefficient, performing cleaning frequency planning for the cleaning sub-region based on the numerical comparison results, and performing real-time cleaning planning for the cleaning sub-region based on the basic cleaning planning data.

[0008] To achieve the above objectives, the present invention employs the following technical solution: the intelligent cleaning strategy planning method based on fusion feature analysis includes the following steps:

[0009] Step S1: Obtain the cleaning area distribution map, divide the cleaning area distribution map into multiple cleaning sub-areas, and divide the cleaning sub-areas into first type cleaning sub-areas and second type cleaning sub-areas, and obtain the basic cleaning planning data corresponding to the first type cleaning sub-areas and the second type cleaning sub-areas respectively.

[0010] Step S2: Calculate the cleaning frequency planning coefficients for the first type of sub-area and the second type of sub-area based on the basic cleaning planning data, and obtain the cleaning planning analysis data;

[0011] Step S3: Obtain the upper limit and lower limit of the baseline interval for the cleaning frequency planning coefficient, compare them with the cleaning frequency planning coefficient, and plan the cleaning frequency for the cleaning sub-area based on the comparison results.

[0012] Step S4: Perform real-time cleaning planning for the cleaning sub-areas based on the basic cleaning plan data.

[0013] Furthermore, it also includes a database, which stores data such as a map of the cleaning areas and a schedule for cleaning the areas.

[0014] In step S1, the specific steps for obtaining the basic data for the cleaning plan are as follows:

[0015] Step S11: Extract the cleaning area distribution map from the database, and divide the cleaning area into multiple cleaning sub-areas according to the cleaning area distribution map;

[0016] Step S12: Obtain the number of pedestrians in the corresponding cleaning sub-area;

[0017] The specific steps are as follows:

[0018] Step S121: Within the time frame of operation of the cleaning sub-area, randomly divide it into n cleaning environment monitoring cycles of the same length and mark them as the first to the nth environmental monitoring cycles respectively.

[0019] Step S122: Obtain the number of pedestrians appearing in the cleaning sub-area during the first to nth environmental monitoring cycles through the first acquisition device, obtain the number of pedestrians during the first to nth monitoring cycles, sort the number of pedestrians during the first to nth monitoring cycles in descending order, obtain the descending order queue of pedestrians, obtain the number of pedestrians during the monitoring cycle corresponding to the median of the descending order queue, and mark it as the number of pedestrians in the area.

[0020] Step S13: Extract the area cleaning schedule from the database, obtain the time interval between two consecutive cleanings of the current cleaning sub-area through the area cleaning schedule, and obtain the cleaning time interval value.

[0021] Step S14: Acquire real-time image data corresponding to the cleaning sub-area through the second acquisition device;

[0022] Step S15: Divide the cleaning sub-area into a first type of cleaning sub-area and a second type of cleaning sub-area, and obtain the basic data for cleaning planning corresponding to different types of cleaning sub-areas;

[0023] The specific steps are as follows:

[0024] Step S151: If the cleaning sub-area is located indoors, then the cleaning sub-area is classified as a first type of cleaning sub-area;

[0025] Step S152: If the cleaning sub-area is located outdoors, then the cleaning sub-area is classified as a second type of cleaning sub-area;

[0026] Step S153: Obtain the basic data for cleaning planning corresponding to the first type of cleaning sub-area;

[0027] Step S154: Obtain the basic data of the cleaning plan corresponding to the second type of cleaning sub-area.

[0028] Furthermore, step S153 also includes the following specific steps:

[0029] Step S1531: Obtain the opening area value of each window in the cleaning sub-area, and count the opening time of the corresponding window on the same day. Calculate the sum of the products of the opening area value of each window and the opening time on the same day to obtain the window opening area duration.

[0030] Step S1532: Obtain the daily wind force level range value corresponding to the environment of the cleaning sub-area through the third acquisition device, calculate the midpoint value of the daily wind force level range value, and obtain the daily wind force level value of the area.

[0031] Step S1533: Calculate the impact value of natural ventilation in the area by combining the window opening area and duration with the wind force level of the area on that day;

[0032] Step S1534: Repeat steps S12-S14 to obtain the real-time image data, natural ventilation impact value, cleaning time interval value and number of pedestrians in each first type of cleaning sub-area to obtain the basic cleaning planning data for the first type of cleaning sub-area.

[0033] Furthermore, step S154 also includes the following specific steps:

[0034] Step S1541: Obtain the daily suspended particulate matter content range value of the cleaning sub-area through the third acquisition device, calculate the midpoint of the daily suspended particulate matter content range value, and obtain the daily particulate matter content value of the area.

[0035] Step S1542: Repeat step S1532 to obtain the daily wind force level value of the area corresponding to the second type of cleaning sub-area;

[0036] Step S1543: Calculate the regional particulate matter impact value by combining the regional particulate matter content value and the regional wind force level value for the day.

[0037] Step S1544: Repeat steps S12-S14 to obtain the real-time image data, daily particulate matter content, cleaning time interval, and number of pedestrians for each second-type cleaning sub-area, and obtain the basic data for cleaning planning corresponding to the second-type cleaning sub-area.

[0038] Furthermore, in step S2, the specific steps for acquiring the cleaning planning analysis data are as follows:

[0039] Step S21: Obtain basic data for cleaning planning, and obtain the cleaning frequency planning coefficient based on the basic data for cleaning planning;

[0040] Step S22: When the cleaning sub-area is a first-type cleaning sub-area, obtain the cleaning frequency planning coefficient;

[0041] The specific steps are as follows:

[0042] Step S221: Obtain the area's natural ventilation impact value, cleaning time interval value, and number of pedestrians in the area based on the basic data of the cleaning plan;

[0043] Step S222: Calculate the cleaning frequency planning coefficient corresponding to the first type of cleaning sub-area by taking the area's natural ventilation impact value, cleaning time interval value, and number of pedestrians in the area;

[0044] Step S23: When the cleaning sub-area is a second type of cleaning sub-area, obtain the cleaning frequency planning coefficient;

[0045] The specific steps are as follows:

[0046] Step S231: Obtain the regional particulate matter impact value, cleaning time interval value, and number of pedestrians in the area based on the basic data of the cleaning plan;

[0047] Step S232: Calculate the cleaning frequency planning coefficient corresponding to the second type of cleaning sub-area by taking the regional particulate matter impact value, cleaning time interval value, and number of pedestrians in the area;

[0048] Step S24: Define the cleaning frequency planning coefficients corresponding to the first type of cleaning sub-area and the second type of cleaning sub-area as cleaning planning analysis data.

[0049] Furthermore, in step S3, the specific steps for planning the cleaning frequency based on the cleaning planning analysis data are as follows:

[0050] Step S31: Establish a first reference frequency cleaning sub-region and a second reference frequency cleaning sub-region, obtain the cleaning frequency planning coefficients corresponding to the first reference frequency cleaning sub-region and the second reference frequency cleaning sub-region respectively, and mark them as the upper limit of the reference interval of the cleaning frequency planning coefficient and the lower limit of the reference interval of the cleaning frequency planning coefficient respectively.

[0051] Step S32: Obtain cleaning planning analysis data, and obtain the cleaning frequency planning coefficient based on the cleaning planning analysis data;

[0052] Step S33: Compare the cleaning frequency planning coefficient with the upper limit and lower limit of the benchmark interval of the cleaning frequency planning coefficient, and plan the cleaning frequency based on the comparison results.

[0053] The specific steps are as follows:

[0054] Step S331: If the cleaning frequency planning coefficient is greater than the upper limit of the baseline range of the cleaning frequency planning coefficient, the cleaning frequency of the corresponding cleaning sub-area needs to be increased until the cleaning frequency planning coefficient of the corresponding cleaning sub-area is between the upper limit of the baseline range of the cleaning frequency planning coefficient and the lower limit of the baseline range of the cleaning frequency planning coefficient.

[0055] Step S332: If the cleaning frequency planning coefficient is between the upper limit of the cleaning frequency planning coefficient benchmark range and the lower limit of the cleaning frequency planning coefficient benchmark range, then there is no need to adjust the cleaning frequency of the cleaning sub-area.

[0056] Step S333: If the cleaning frequency planning coefficient is less than the lower limit of the benchmark range of the cleaning frequency planning coefficient, the cleaning frequency of the cleaning sub-area needs to be reduced and adjusted until the cleaning frequency planning coefficient of the corresponding cleaning sub-area is between the upper limit and the lower limit of the benchmark range of the cleaning frequency planning coefficient.

[0057] Furthermore, the specific steps of step S4: real-time cleaning planning based on the basic cleaning planning data are as follows:

[0058] Step S41: Obtain real-time image data corresponding to multiple cleaning sub-areas based on the basic cleaning plan data;

[0059] Step S42: Mark the passage area in the real-time image data to obtain the marked passage area image, and identify the visible litter in the marked passage area;

[0060] Step S43: Obtain the image corresponding to the marked passage area after cleaning is completed, and obtain the cleaning image of the passage area. Compare the image of the marked passage area with the cleaning image of the passage area. If the image comparison results are consistent, it is determined that there is no visible garbage in the passage area. If the image comparison results are inconsistent, it is determined that there is visible garbage in the passage area.

[0061] Step S44: If there is visible litter in the marked passage area, immediately plan for cleaning staff to clean the designated sub-area;

[0062] Step S45: If there is no visible litter in the marked passage area, further analysis of the real-time image data is performed.

[0063] Furthermore, step S45 also includes the following specific steps:

[0064] Step S451: Obtain images of the cleaning sub-areas corresponding to multiple different time points in the same cleaning cycle through real-time image data, and mark them as images from the first monitoring time to the m-th monitoring time.

[0065] Step S452: Obtain the garbage distribution density value of the cleaning area corresponding to the image at the first monitoring time, as follows:

[0066] Step S453: Binarize the image at the first monitoring time, mark the pixel value corresponding to the area where the visible garbage is located as 1, and mark the pixel value of the non-visible garbage area as 0, to obtain the binarized image of the cleaning area;

[0067] Step S454: Count the number of pixels with a pixel value of 1 in the binarized image of the cleaning area to obtain the number of pixels in the visible garbage area, and count the number of pixels with a pixel value of 0 in the binarized image of the cleaning area to obtain the number of pixels in the non-visible garbage area.

[0068] Step S455: Summate the number of pixels in the visible waste area with the number of pixels in the non-visible waste area to obtain the total number of pixels in the cleaning area. Calculate the ratio of the number of pixels in the visible waste area to the total number of pixels in the cleaning area to obtain the waste distribution density value at the first monitoring time.

[0069] Step S456: Repeat the process of obtaining the garbage distribution density value at the first monitoring time, and obtain the garbage distribution density value corresponding to the image at each monitoring time to obtain the garbage distribution density values ​​at the first to the mth monitoring times.

[0070] Step S457: Mark the image acquisition time values ​​corresponding to the first monitoring time image to the m-th detection time image as the first to the m-th time values, and establish a line graph of waste density change over time with the time values ​​as the horizontal axis and the waste distribution density values ​​as the vertical axis.

[0071] Step S458: Using image analysis tools, obtain the slope value of the line connecting every two consecutive coordinate points in the line graph of waste density change over time, obtain the waste density growth rate corresponding to the image at different detection times, obtain the waste density growth rate threshold, compare each waste density growth rate with the waste density growth rate threshold, and make real-time cleaning plans based on the comparison results.

[0072] Furthermore, step S458 also includes the following specific steps:

[0073] Step S4581: When the rate of increase in waste density is greater than or equal to the threshold for the rate of increase in waste density, it is determined to be an abnormal rate of increase in waste density;

[0074] Step S4582: When the rate of increase in waste density is less than the threshold for the rate of increase in waste density, it is determined to be a normal rate of increase in waste density.

[0075] Step S4583: If there is an abnormal growth rate of garbage density in the line graph of garbage density change over time, immediately arrange for cleaning personnel to clean the sub-area.

[0076] Step S4584: If there is no abnormal growth rate of garbage density in the line graph of garbage density change over time, then clean the sub-area according to the regular cleaning frequency.

[0077] The intelligent cleaning strategy planning system based on fusion feature analysis includes a data acquisition module, a data analysis module, a frequency planning module, and a real-time planning module.

[0078] Data acquisition module: Acquires basic data for cleaning planning;

[0079] Data Analysis Module: Analyzes the basic data of the cleaning plan to obtain cleaning plan analysis data;

[0080] Frequency planning module: Plans cleaning frequency for cleaning sub-areas based on cleaning planning analysis data;

[0081] Real-time planning module: Performs real-time cleaning planning for cleaning sub-areas based on basic cleaning planning data.

[0082] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0083] 1. This invention dynamically adjusts the cleaning frequency by acquiring data with various characteristics, which can effectively improve the cleaning quality and the utilization efficiency of cleaning resources;

[0084] 2. This invention dynamically acquires real-time image data and performs real-time cleaning planning based on the real-time image data, enabling the cleaning strategy to respond promptly to environmental changes and effectively improving the targeting of the cleaning strategy. Attached Figure Description

[0085] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0086] Figure 1 This is a diagram illustrating the implementation steps of the present invention;

[0087] Figure 2 This is a system framework diagram of the present invention;

[0088] Figure 3 This is a line graph showing the time-varying change in waste density in this invention;

[0089] In the figure: 1. Waste distribution density value at the first monitoring time; 2. Waste distribution density value at the second monitoring time; m. Waste distribution density value at the m-th monitoring time. Detailed Implementation

[0090] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0091] Example 1

[0092] Please see Figure 1 The present invention provides a technical solution: a smart cleaning strategy planning system based on fusion feature analysis, including a data acquisition module, a data analysis module, a frequency planning module, a real-time planning module and a server, wherein the data acquisition module, the data analysis module, the frequency planning module and the real-time planning module are respectively connected to the server;

[0093] The data acquisition module acquires basic data for the cleaning plan;

[0094] It also includes a database, which stores data such as a map of the cleaning area and a schedule for cleaning the area.

[0095] Extract the cleaning area distribution map from the database, and divide the cleaning area into multiple cleaning sub-areas based on the cleaning area distribution map;

[0096] Within the time frame of operation of the cleaning sub-area, n cleaning environment monitoring cycles of the same length are randomly divided and marked as the first to the nth environmental monitoring cycle.

[0097] The first acquisition device acquires the number of pedestrians appearing in the cleaning sub-area during the first to nth environmental monitoring cycles, obtains the number of pedestrians during the first to nth monitoring cycles, sorts the number of pedestrians during the first to nth monitoring cycles in descending order, obtains the descending order queue of pedestrians, obtains the number of pedestrians during the monitoring cycle corresponding to the median of the descending order queue, and marks it as the number of pedestrians in the area.

[0098] It should be noted that if the number of values ​​in the descending order queue is odd, the median is the middle value; if the number of values ​​in the descending order queue is even, the median is the average of the two middle values. The median is a positive integer; if the average is not an integer, it is rounded up.

[0099] Extract the area cleaning schedule from the database, obtain the time interval between two consecutive cleanings of the current cleaning sub-area through the area cleaning schedule, and get the cleaning time interval value;

[0100] The second acquisition device acquires real-time image data corresponding to the cleaning sub-area;

[0101] The cleaning sub-areas are divided into Type I cleaning sub-areas and Type II cleaning sub-areas. The specific division process is as follows:

[0102] If the cleaning sub-area is located indoors, then the cleaning sub-area will be classified as a first-type cleaning sub-area;

[0103] If the cleaning sub-area is located outdoors, then the cleaning sub-area will be classified as a second type of cleaning sub-area.

[0104] It should be noted here that:

[0105] In this application, "indoor" refers to the interior space of a building enclosed by structures such as walls and roofs, including but not limited to indoor corridors and indoor halls;

[0106] Outdoor space refers to the space outside a building that is not enclosed by walls, roofs, or other structures, including but not limited to plazas and roads.

[0107] Obtain the basic cleaning plan data corresponding to the first type of cleaning sub-area, as follows:

[0108] Obtain the opening area value of each window in the cleaning sub-area, and count the opening time of the corresponding window on the same day. Calculate the sum of the products of the opening area value of each window and the opening time on the same day to obtain the window opening area duration.

[0109] The opening area value of each window is obtained as follows:

[0110] Obtain the building area value of each window in the cleaning sub-area;

[0111] Within the cleaning sub-area, i windows are randomly selected as sample windows and labeled as the first sample window to the i-th sample window, respectively.

[0112] Obtain the opening angle value corresponding to each sample window and mark them as the first to the i-th opening angle values ​​respectively. Calculate the average value of the first to the i-th opening angle values ​​to obtain the average window opening angle corresponding to the cleaning sub-area.

[0113] Obtain the opening angle reference value for each sample window, and get the first to the i-th opening angle reference values. Calculate the average value of the first to the i-th opening angle reference values ​​to obtain the average window opening angle reference value for the cleaning sub-area.

[0114] Calculate the ratio of the average window opening angle to the baseline value of the average window opening angle to obtain the window opening area ratio;

[0115] Calculate the product of the building area of ​​each window and the ratio of the window's opening area to obtain the opening area of ​​each window.

[0116] The wind force level range value corresponding to the environment of the cleaning sub-area on that day is obtained by the third acquisition device, and the midpoint value of the wind force level range value on that day is calculated to obtain the wind force level value of the area on that day.

[0117] The impact value of natural ventilation in the area is obtained by calculating the area and duration of window opening and the wind force level of the area on that day.

[0118] The calculation of the impact of natural ventilation in the area is performed using the following formula:

[0119] Qtf = Kkq + Fl*a1;

[0120] Where Qtf is the regional natural ventilation impact value, Kkq is the window opening area and duration, Fl is the regional wind force level value for the day, and a1 is the set proportional coefficient with a1 greater than 0.

[0121] Repeat the above process of obtaining the regional natural ventilation impact value, cleaning time interval value and regional pedestrian number, and obtain the real-time image data, regional natural ventilation impact value, cleaning time interval value and regional pedestrian number for each first type of cleaning sub-area respectively, to obtain the basic data of cleaning planning for the first type of cleaning sub-area;

[0122] Obtain the basic cleaning plan data corresponding to the second type of cleaning sub-area, as follows:

[0123] The daily suspended particulate matter content range of the cleaning sub-area is obtained by acquiring the range value of the daily suspended particulate matter content range through the third acquisition device, and the midpoint of the range value of the daily suspended particulate matter content range is calculated to obtain the daily particulate matter content value of the area.

[0124] Repeat the above process of obtaining the daily wind force level value for the area corresponding to the first type of cleaning sub-area, and obtain the daily wind force level value for the area corresponding to the second type of cleaning sub-area.

[0125] The regional particulate matter impact value is obtained by calculating the regional particulate matter content value and the regional wind force level value on the same day.

[0126] The formula for calculating the impact of particulate matter on the region is configured as follows:

[0127] Qky = Khl + Fl * a2;

[0128] Where Qky is the regional particulate matter impact value, Khl is the regional particulate matter content value on the day, Fl is the regional wind force level value on the day, and a2 is a set proportional coefficient with a2 greater than 0.

[0129] Repeat the above process of obtaining the impact value of natural ventilation, the cleaning time interval value and the number of pedestrians in the area, and obtain the real-time image data, the daily particulate matter content value, the cleaning time interval value and the number of pedestrians in the area corresponding to each second type of cleaning sub-area, to obtain the basic data of cleaning planning corresponding to the second type of cleaning sub-area.

[0130] It should be noted here that:

[0131] The first acquisition device involved here is a surveillance camera, the second acquisition device is a high-definition camera, and the third acquisition device is a weather forecast.

[0132] The data acquisition module acquires the basic data for cleaning planning and transmits it to the frequency planning module and the real-time planning module;

[0133] The data analysis module analyzes the basic data of the cleaning plan to obtain cleaning plan analysis data;

[0134] Obtain basic data for cleaning planning, and obtain cleaning frequency planning coefficients based on the basic data for cleaning planning;

[0135] When the cleaning sub-area is a Type I cleaning sub-area, the cleaning frequency planning coefficient is obtained as follows:

[0136] Based on the basic data of the cleaning plan, obtain the impact value of natural ventilation in the area, the cleaning time interval value, and the number of pedestrians in the area;

[0137] The cleaning frequency planning coefficient for the first type of cleaning sub-area is obtained by calculating the regional natural ventilation impact value, the cleaning time interval value, and the number of pedestrians in the area.

[0138] The specific formula for the cleaning frequency planning coefficient is configured as follows:

[0139]

[0140] Wherein, Bgx1 is the cleaning frequency planning coefficient corresponding to the first type of cleaning sub-area, Qtf is the area natural ventilation impact value, Qxs is the number of pedestrians in the area, Sjj is the cleaning time interval value, and b1 is the set proportional coefficient and b1 is greater than 0.

[0141] When the cleaning sub-area is a type II cleaning sub-area, the cleaning frequency planning coefficient is obtained as follows:

[0142] Based on the basic data of the cleaning plan, obtain the regional particulate matter impact value, cleaning interval value, and number of pedestrians in the area;

[0143] The cleaning frequency planning coefficient for the second type of cleaning sub-area is obtained by calculating the regional particulate matter impact value, the cleaning time interval value, and the number of pedestrians in the area.

[0144] The specific formula for the cleaning frequency planning coefficient is configured as follows:

[0145]

[0146] Wherein, Bgx1 is the cleaning frequency planning coefficient corresponding to the second type of cleaning sub-area, Qky is the particulate matter impact value of the area, Qxs is the number of pedestrians in the area, Sjj is the cleaning time interval value, and b2 is the set proportional coefficient and b2 is greater than 0.

[0147] The cleaning frequency planning coefficients corresponding to the first type of cleaning sub-area and the second type of cleaning sub-area are defined as cleaning planning analysis data.

[0148] The frequency planning module plans the cleaning frequency based on the cleaning planning analysis data.

[0149] Establish a first reference frequency cleaning sub-region and a second reference frequency cleaning sub-region, obtain the cleaning frequency planning coefficients corresponding to the first reference frequency cleaning sub-region and the second reference frequency cleaning sub-region respectively, and mark them as the upper limit of the reference interval of the cleaning frequency planning coefficient and the lower limit of the reference interval of the cleaning frequency planning coefficient respectively.

[0150] It should be noted here that:

[0151] In this application, the first reference frequency cleaning sub-area and the second reference frequency cleaning sub-area mentioned here are both manually arranged cleaning scenarios used to determine the cleaning frequency planning coefficients.

[0152] The first reference frequency cleaning sub-area involved here refers to the cleaning area at the lower limit of the cleaning qualification and the cleaning area at the upper limit of the cleaning qualification. The lower limit of the cleaning qualification means that if the cleaning frequency planning coefficient corresponding to the cleaning area is greater than the cleaning frequency planning coefficient corresponding to the first reference frequency cleaning sub-area, it means that the cleaning of this area is unqualified. The upper limit of the cleaning qualification means that if the cleaning frequency planning coefficient corresponding to the cleaning area is less than the cleaning frequency planning coefficient corresponding to the second reference frequency cleaning sub-area, it means that the cleaning of this area is over-cleaned.

[0153] Obtain cleaning planning analysis data, and obtain cleaning frequency planning coefficients based on the cleaning planning analysis data;

[0154] If the cleaning frequency planning coefficient is greater than the upper limit of the benchmark range of the cleaning frequency planning coefficient, the cleaning frequency of the corresponding cleaning sub-area needs to be increased until the cleaning frequency planning coefficient of the corresponding cleaning sub-area is between the upper limit and the lower limit of the benchmark range of the cleaning frequency planning coefficient.

[0155] If the cleaning frequency planning coefficient is between the upper limit and the lower limit of the benchmark range of the cleaning frequency planning coefficient, then there is no need to adjust the cleaning frequency of the cleaning sub-area.

[0156] If the cleaning frequency planning coefficient is less than the lower limit of the benchmark range of the cleaning frequency planning coefficient, the cleaning frequency of the cleaning sub-area needs to be reduced and adjusted until the cleaning frequency planning coefficient of the corresponding cleaning sub-area is between the upper limit and the lower limit of the benchmark range of the cleaning frequency planning coefficient.

[0157] The real-time planning module performs real-time cleaning planning based on the basic cleaning planning data.

[0158] Real-time image data corresponding to multiple cleaning sub-areas are obtained based on the basic data of the cleaning plan.

[0159] The passage area is marked in the real-time image data to obtain the marked passage area image. The visible litter in the marked passage area is identified. The identification results include whether there is visible litter in the passage area and whether there is no visible litter in the passage area.

[0160] The identification of visible litter is as follows:

[0161] Obtain the image corresponding to the marked passage area after cleaning is completed, and obtain the cleaning image of the passage area. Compare the image of the marked passage area with the cleaning image of the passage area. If the image comparison results are consistent, it is determined that there is no visible trash in the passage area. If the image comparison results are inconsistent, it is determined that there is visible trash in the passage area.

[0162] It should be noted here that:

[0163] The images of the access areas involved here are the images of the access areas that were cleaned in the previous cleaning at the current moment.

[0164] The visible litter referred to here specifically refers to litter or debris that is visible in images or videos, and is distinct from litter such as dust that is difficult to see in images;

[0165] If there is visible litter in the marked access area, cleaning staff will be immediately scheduled to clean the designated area.

[0166] If there is no visible litter within the marked passage area, the real-time image data will be further analyzed, as follows:

[0167] Images of the cleaning sub-areas corresponding to multiple different time points in the same cleaning cycle are acquired by real-time image data and marked as images from the first monitoring time to the mth monitoring time.

[0168] The garbage distribution density value of the cleaning area corresponding to the image at the first monitoring time is obtained as follows:

[0169] The image at the first monitoring time was binarized, and the pixel values ​​corresponding to the areas where visible garbage was located were marked as 1, while the pixel values ​​of the areas where non-visible garbage was located were marked as 0, thus obtaining a binarized image of the cleaning area.

[0170] The number of pixels with a value of 1 in the binarized image of the cleaning area is counted to obtain the number of pixels in the visible garbage area, and the number of pixels with a value of 0 in the binarized image of the cleaning area is counted to obtain the number of pixels in the non-visible garbage area.

[0171] The total number of pixels in the cleaning area is obtained by summing the pixel count values ​​of the visible garbage area and the non-visible garbage area.

[0172] Calculate the ratio of the number of pixels in the visible waste area to the total number of pixels in the cleaning area to obtain the waste distribution density value at the first monitoring time.

[0173] Repeat the process of acquiring the garbage distribution density value at the first monitoring time, and acquire the garbage distribution density value corresponding to the image at each monitoring time to obtain the garbage distribution density values ​​at the first to the mth monitoring times.

[0174] The image acquisition time values ​​corresponding to the images from the first monitoring time to the m-th monitoring time are marked as the first to the m-th time values;

[0175] Create a line graph showing the change in waste density over time, with time values ​​on the x-axis and waste distribution density values ​​on the y-axis. (See [link to relevant documentation]). Figure 3 ;

[0176] Based on the image analysis tool, the slope value of the line connecting every two consecutive coordinate points in the line graph of waste density change over time is obtained, and the waste density growth rate corresponding to the image at different detection times is obtained.

[0177] Obtain the threshold for the rate of increase in waste density, and compare each rate of increase in waste density with the threshold for the rate of increase in waste density.

[0178] It should be noted here that:

[0179] The image analysis tool mentioned here can be MATLAB;

[0180] In this application, the threshold for the rate of increase in garbage density is the maximum rate of increase in garbage density under the normal cleaning frequency. The threshold for the rate of increase in garbage density set here is only applicable to this application. In actual implementation, the threshold for the rate of increase in garbage density needs to be set according to the actual situation.

[0181] The specific numerical comparison process is as follows:

[0182] When the rate of increase in waste density is greater than or equal to the threshold for the rate of increase in waste density, it is judged as an abnormal rate of increase in waste density.

[0183] When the rate of increase in waste density is less than the threshold for the rate of increase in waste density, it is judged to be the normal growth rate of waste density.

[0184] If an abnormal growth rate of garbage density is found in the line graph of garbage density over time, cleaning staff should be immediately arranged to clean the designated sub-area.

[0185] If there is no abnormal growth rate of garbage density in the line graph of garbage density over time, then the cleaning sub-area will be cleaned according to the regular cleaning frequency.

[0186] In this application, if a corresponding calculation formula appears, the above calculation formula is a dimensionless calculation. The weighting coefficient, proportional coefficient and other coefficients in the formula are set to quantify each parameter to obtain a result value. The size of the weighting coefficient and proportional coefficient is only required to not affect the proportional relationship between the parameter and the result value.

[0187] Example 2

[0188] Please see Figure 1 Based on another concept of the same invention, a smart cleaning strategy planning method based on fusion feature analysis is proposed, including the following steps:

[0189] Step S1: Obtain basic data for cleaning planning;

[0190] It also includes a database, which stores data such as a map of the cleaning area and a schedule for cleaning the area.

[0191] Step S11: Extract the cleaning area distribution map from the database, and divide the cleaning area into multiple cleaning sub-areas according to the cleaning area distribution map;

[0192] Step S12: Obtain the number of pedestrians in the corresponding cleaning sub-area;

[0193] Step S121: Within the time frame of operation of the cleaning sub-area, randomly divide it into n cleaning environment monitoring cycles of the same length and mark them as the first to the nth environmental monitoring cycles respectively.

[0194] Step S122: Obtain the number of pedestrians appearing in the cleaning sub-area during the first to nth environmental monitoring cycles through the first acquisition device, obtain the number of pedestrians during the first to nth monitoring cycles, sort the number of pedestrians during the first to nth monitoring cycles in descending order, obtain the descending order queue of pedestrians, obtain the number of pedestrians during the monitoring cycle corresponding to the median of the descending order queue, and mark it as the number of pedestrians in the area.

[0195] Step S13: Extract the area cleaning schedule from the database, obtain the time interval between two consecutive cleanings of the current cleaning sub-area through the area cleaning schedule, and obtain the cleaning time interval value.

[0196] Step S14: Acquire real-time image data corresponding to the cleaning sub-area through the second acquisition device;

[0197] Step S15: Divide the cleaning sub-area into a first type of cleaning sub-area and a second type of cleaning sub-area, and obtain the basic data for cleaning planning corresponding to different types of cleaning sub-areas;

[0198] Step S151: If the cleaning sub-area is located indoors, then the cleaning sub-area is classified as a first type of cleaning sub-area;

[0199] Step S152: If the cleaning sub-area is located outdoors, then the cleaning sub-area is classified as a second type of cleaning sub-area;

[0200] Step S153: Obtain the basic data for cleaning planning corresponding to the first type of cleaning sub-area;

[0201] Step S153 further includes the following specific steps:

[0202] Step S1531: Obtain the opening area value of each window in the cleaning sub-area, and count the opening time of the corresponding window on the same day. Calculate the sum of the products of the opening area value of each window and the opening time on the same day to obtain the window opening area duration.

[0203] Step S1532: Obtain the daily wind force level range value corresponding to the environment of the cleaning sub-area through the third acquisition device, calculate the midpoint value of the daily wind force level range value, and obtain the daily wind force level value of the area.

[0204] Step S1533: Calculate the impact value of natural ventilation in the area by combining the window opening area and duration with the wind force level of the area on that day;

[0205] Step S1534: Repeat the above process of obtaining the influence value of natural ventilation, the cleaning time interval value and the number of pedestrians in the area. Obtain the real-time image data, the influence value of natural ventilation, the cleaning time interval value and the number of pedestrians in the area corresponding to each first type of cleaning sub-area to obtain the basic data of cleaning planning corresponding to the first type of cleaning sub-area.

[0206] Step S154: Obtain the basic data for cleaning planning corresponding to the second type of cleaning sub-area;

[0207] Step S154 further includes the following specific steps:

[0208] Step S1541: Obtain the daily suspended particulate matter content range value of the cleaning sub-area through the third acquisition device, calculate the midpoint of the daily suspended particulate matter content range value, and obtain the daily particulate matter content value of the area.

[0209] Step S1542: Repeat step S1532 to obtain the daily wind force level value of the area corresponding to the second type of cleaning sub-area;

[0210] Step S1543: Calculate the regional particulate matter impact value by combining the regional particulate matter content value and the regional wind force level value for the day.

[0211] Step S1544: Repeat the above process of obtaining the impact value of natural ventilation in the area, the cleaning time interval value and the number of pedestrians in the area. Obtain the real-time image data, the daily particulate matter content value, the cleaning time interval value and the number of pedestrians in the area corresponding to each second type of cleaning sub-area to obtain the basic data of cleaning planning corresponding to the second type of cleaning sub-area.

[0212] Step S2: Analyze the basic data of the cleaning plan to obtain the cleaning plan analysis data;

[0213] Step S21: Obtain basic data for cleaning planning, and obtain the cleaning frequency planning coefficient based on the basic data for cleaning planning;

[0214] Step S22: When the cleaning sub-area is a first-type cleaning sub-area, obtain the cleaning frequency planning coefficient;

[0215] Step S22 further includes the following specific steps:

[0216] Step S221: Obtain the area's natural ventilation impact value, cleaning time interval value, and number of pedestrians in the area based on the basic data of the cleaning plan;

[0217] Step S222: Calculate the cleaning frequency planning coefficient corresponding to the first type of cleaning sub-area by taking the area's natural ventilation impact value, cleaning time interval value, and number of pedestrians in the area;

[0218] Step S23: When the cleaning sub-area is a second type of cleaning sub-area, obtain the cleaning frequency planning coefficient;

[0219] Step S23 further includes the following specific steps:

[0220] Step S231: Obtain the regional particulate matter impact value, cleaning time interval value, and number of pedestrians in the area based on the basic data of the cleaning plan;

[0221] Step S232: Calculate the cleaning frequency planning coefficient corresponding to the second type of cleaning sub-area by taking the regional particulate matter impact value, cleaning time interval value, and number of pedestrians in the area;

[0222] Step S24: Define the cleaning frequency planning coefficients corresponding to the first type of cleaning sub-area and the second type of cleaning sub-area as cleaning planning analysis data;

[0223] Step S3: Plan the cleaning frequency based on the cleaning planning analysis data;

[0224] Step S31: Establish a first reference frequency cleaning sub-region and a second reference frequency cleaning sub-region, obtain the cleaning frequency planning coefficients corresponding to the first reference frequency cleaning sub-region and the second reference frequency cleaning sub-region respectively, and mark them as the upper limit of the reference interval of the cleaning frequency planning coefficient and the lower limit of the reference interval of the cleaning frequency planning coefficient respectively.

[0225] Step S32: Obtain cleaning planning analysis data, and obtain the cleaning frequency planning coefficient based on the cleaning planning analysis data;

[0226] Step S33: Compare the cleaning frequency planning coefficient with the upper limit and lower limit of the benchmark interval of the cleaning frequency planning coefficient, and plan the cleaning frequency based on the comparison results.

[0227] Step S33 further includes the following specific steps:

[0228] Step S331: If the cleaning frequency planning coefficient is greater than the upper limit of the baseline range of the cleaning frequency planning coefficient, the cleaning frequency of the corresponding cleaning sub-area needs to be increased until the cleaning frequency planning coefficient of the corresponding cleaning sub-area is between the upper limit of the baseline range of the cleaning frequency planning coefficient and the lower limit of the baseline range of the cleaning frequency planning coefficient.

[0229] Step S332: If the cleaning frequency planning coefficient is between the upper limit of the cleaning frequency planning coefficient benchmark range and the lower limit of the cleaning frequency planning coefficient benchmark range, then there is no need to adjust the cleaning frequency of the cleaning sub-area.

[0230] Step S333: If the cleaning frequency planning coefficient is less than the lower limit of the benchmark range of the cleaning frequency planning coefficient, the cleaning frequency of the cleaning sub-area needs to be reduced and adjusted until the cleaning frequency planning coefficient of the corresponding cleaning sub-area is between the upper limit and the lower limit of the benchmark range of the cleaning frequency planning coefficient.

[0231] Step S4: Perform real-time cleaning planning based on the basic cleaning plan data;

[0232] Step S41: Obtain real-time image data corresponding to multiple cleaning sub-areas based on the basic cleaning plan data;

[0233] Step S42: Mark the passage area in the real-time image data to obtain the marked passage area image, and identify the visible litter in the marked passage area;

[0234] Step S43: Obtain the image corresponding to the marked passage area after cleaning is completed, and obtain the cleaning image of the passage area. Compare the image of the marked passage area with the cleaning image of the passage area. If the image comparison results are consistent, it is determined that there is no visible garbage in the passage area. If the image comparison results are inconsistent, it is determined that there is visible garbage in the passage area.

[0235] Step S44: If there is visible litter in the marked passage area, immediately plan for cleaning staff to clean the designated sub-area;

[0236] Step S45: If there is no visible litter in the marked passage area, further analysis of the real-time image data is performed;

[0237] Step S45 further includes the following specific steps:

[0238] Step S451: Obtain images of the cleaning sub-areas corresponding to multiple different time points in the same cleaning cycle through real-time image data, and mark them as images from the first monitoring time to the m-th monitoring time.

[0239] Step S452: Obtain the garbage distribution density value of the cleaning area corresponding to the image at the first monitoring time, as follows:

[0240] Step S453: Binarize the image at the first monitoring time, mark the pixel value corresponding to the area where the visible garbage is located as 1, and mark the pixel value of the non-visible garbage area as 0, to obtain the binarized image of the cleaning area;

[0241] Step S454: Count the number of pixels with a pixel value of 1 in the binarized image of the cleaning area to obtain the number of pixels in the visible garbage area, and count the number of pixels with a pixel value of 0 in the binarized image of the cleaning area to obtain the number of pixels in the non-visible garbage area.

[0242] Step S455: Summate the number of pixels in the visible waste area with the number of pixels in the non-visible waste area to obtain the total number of pixels in the cleaning area. Calculate the ratio of the number of pixels in the visible waste area to the total number of pixels in the cleaning area to obtain the waste distribution density value at the first monitoring time.

[0243] Step S456: Repeat the process of obtaining the garbage distribution density value at the first monitoring time, and obtain the garbage distribution density value corresponding to the image at each monitoring time to obtain the garbage distribution density values ​​at the first to the mth monitoring times.

[0244] Step S457: Mark the image acquisition time values ​​corresponding to the first monitoring time image to the m-th detection time image as the first to the m-th time values, and establish a line graph of waste density change over time with the time values ​​as the horizontal axis and the waste distribution density values ​​as the vertical axis.

[0245] Step S458: Using image analysis tools, obtain the slope value of the line connecting every two consecutive coordinate points in the line graph of garbage density change over time, obtain the garbage density growth rate corresponding to the image at different detection times, obtain the garbage density growth rate threshold, compare each garbage density growth rate with the garbage density growth rate threshold, and make real-time cleaning plans based on the comparison results.

[0246] Step S458 further includes the following specific steps:

[0247] Step S4581: When the rate of increase in waste density is greater than or equal to the threshold for the rate of increase in waste density, it is determined to be an abnormal rate of increase in waste density;

[0248] Step S4582: When the rate of increase in waste density is less than the threshold for the rate of increase in waste density, it is determined to be a normal rate of increase in waste density.

[0249] Step S4583: If there is an abnormal growth rate of garbage density in the line graph of garbage density change over time, immediately arrange for cleaning personnel to clean the sub-area.

[0250] Step S4584: If there is no abnormal growth rate of garbage density in the line graph of garbage density change over time, then clean the sub-area according to the regular cleaning frequency;

[0251] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A smart cleaning strategy planning method based on fusion feature analysis, characterized in that, Includes the following steps: Step S1: Obtain the cleaning area distribution map, divide the cleaning area distribution map into multiple cleaning sub-areas, and divide the cleaning sub-areas into first type cleaning sub-areas and second type cleaning sub-areas, and obtain the basic cleaning planning data corresponding to the two types of cleaning sub-areas respectively. Step S2: Calculate the cleaning frequency planning coefficients for the first type of sub-area and the second type of sub-area based on the basic cleaning planning data, and obtain the cleaning planning analysis data; Step S3: Obtain the upper limit and lower limit of the baseline interval for the cleaning frequency planning coefficient, compare them with the cleaning frequency planning coefficient, and plan the cleaning frequency for the cleaning sub-area based on the comparison results. Step S4: Based on the basic cleaning plan data, perform real-time cleaning planning for the cleaning sub-areas. The specific steps of step S4 are as follows: Step S41: Obtain real-time image data corresponding to multiple cleaning sub-areas based on the basic cleaning plan data; Step S42: Mark the passage area in the real-time image data to obtain the marked passage area image, and identify the visible litter in the marked passage area; Step S43: Obtain the image corresponding to the marked passage area after cleaning, obtain the cleaning image of the passage area, and compare it with the image of the marked passage area. If the image comparison results are consistent, it is determined that there is no visible garbage in the passage area. If the image comparison results are inconsistent, it is determined that there is visible garbage in the passage area. Step S44: If there is visible litter in the marked passage area, clean the cleaning sub-area immediately; Step S45: If there is no visible litter in the marked passage area, further analysis of the real-time image data is performed.

2. The intelligent cleaning strategy planning method based on fusion feature analysis according to claim 1, characterized in that, In step S1, the specific steps are as follows: Step S11: Obtain the cleaning area distribution map and divide the cleaning area into multiple cleaning sub-areas based on the cleaning area distribution map; Step S12: Obtain the number of pedestrians in the corresponding cleaning sub-area; Step S121: Within the time frame of operation of the cleaning sub-area, randomly divide it into n cleaning environment monitoring cycles of the same length to obtain the first to the nth environmental monitoring cycles; Step S122: Obtain the number of pedestrians appearing in the cleaning sub-area in n environmental monitoring cycles, obtain the number of pedestrians in n monitoring cycles, sort the number of pedestrians in n monitoring cycles in descending order, obtain the descending order queue of pedestrians, obtain the number of pedestrians in the monitoring cycle corresponding to the median of the descending queue, and mark it as the number of pedestrians in the area. Step S13: Obtain the area cleaning schedule, query the time interval between two consecutive cleanings of each sub-area, and obtain the cleaning time interval value; Step S14: Obtain real-time image data corresponding to the cleaning sub-area; Step S15: Divide the cleaning sub-area into a first type of cleaning sub-area and a second type of cleaning sub-area, and obtain the basic data for cleaning planning corresponding to different types of cleaning sub-areas; Step S151: If the cleaning sub-area is located indoors, then the cleaning sub-area is classified as a first type of cleaning sub-area; Step S152: If the cleaning sub-area is located outdoors, then the cleaning sub-area is classified as a second type of cleaning sub-area; Step S153: Obtain the basic data for cleaning planning corresponding to the first type of cleaning sub-area; Step S154: Obtain the basic data of the cleaning plan corresponding to the second type of cleaning sub-area.

3. The intelligent cleaning strategy planning method based on fusion feature analysis according to claim 2, characterized in that, The specific steps of step S153 are as follows: Step S1531: Obtain the opening area value of each window in the cleaning sub-area, count the opening time of the corresponding window, calculate the sum of the products of the opening area value of each window and the opening time of the corresponding day, and obtain the window opening area duration. Step S1532: Obtain the daily wind force level range value corresponding to the environment of the cleaning sub-area, calculate the midpoint of the daily wind force level range value, and obtain the daily wind force level value of the area. Step S1533: Calculate the impact value of natural ventilation in the area by combining the window opening area and duration with the wind force level of the area on that day; Step S1534: Obtain the real-time image data, natural ventilation impact value, cleaning time interval value and number of pedestrians in each first type of cleaning sub-area to obtain the basic cleaning planning data corresponding to the first type of cleaning sub-area.

4. The intelligent cleaning strategy planning method based on fusion feature analysis according to claim 2, characterized in that, The specific steps of step S154 are as follows: Step S1541: Obtain the daily suspended particulate matter content range of the cleaning sub-area, calculate the midpoint of the daily suspended particulate matter content range, and obtain the daily particulate matter content value of the area. Step S1542: Obtain the daily wind force level value for the area corresponding to the second type of cleaning sub-area; Step S1543: Calculate the regional particulate matter impact value by combining the regional particulate matter content value and the regional wind force level value for the day. Step S1544: Obtain the real-time image data, daily particulate matter content, cleaning time interval, and number of pedestrians in the area corresponding to the second type of cleaning sub-area to obtain the basic data for cleaning planning corresponding to the second type of cleaning sub-area.

5. The intelligent cleaning strategy planning method based on fusion feature analysis according to claim 1, characterized in that, In step S2, the specific steps are as follows: Step S21: Obtain basic data for cleaning planning, and obtain the cleaning frequency planning coefficient based on the basic data for cleaning planning; Step S22: When the cleaning sub-area is a first-type cleaning sub-area, obtain the cleaning frequency planning coefficient; Step S23: When the cleaning sub-area is a second type of cleaning sub-area, obtain the cleaning frequency planning coefficient; Step S24: Define the cleaning frequency planning coefficients corresponding to the first type of cleaning sub-area and the second type of cleaning sub-area as cleaning planning analysis data.

6. The intelligent cleaning strategy planning method based on fusion feature analysis according to claim 1, characterized in that, In step S3, the specific steps are as follows: Step S31: Establish the first reference frequency cleaning sub-region and the second reference frequency cleaning sub-region, obtain the cleaning frequency planning coefficients of the two sub-regions, and mark them as the upper limit of the reference interval of the cleaning frequency planning coefficient and the lower limit of the reference interval of the cleaning frequency planning coefficient, respectively. Step S32: Obtain cleaning planning analysis data, and obtain the cleaning frequency planning coefficient based on the cleaning planning analysis data; Step S33: Compare the cleaning frequency planning coefficient with the upper limit and lower limit of the benchmark interval of the cleaning frequency planning coefficient, and plan the cleaning frequency based on the comparison results. Step S331: If the frequency exceeds the upper limit of the baseline range of the cleaning frequency planning coefficient, increase the cleaning frequency of the corresponding cleaning sub-area. Step S332: If the cleaning frequency is between the upper limit of the baseline range of the cleaning frequency planning coefficient and the lower limit of the baseline range of the cleaning frequency planning coefficient, then there is no need to adjust the cleaning frequency of the cleaning sub-area; Step S333: If it is less than the lower limit of the baseline range of the cleaning frequency planning coefficient, then adjust the cleaning frequency accordingly.

7. The intelligent cleaning strategy planning method based on fusion feature analysis according to claim 1, characterized in that, In step S45, the specific details are as follows: Step S451: Obtain images of the cleaning sub-areas corresponding to multiple different time points in the same cleaning cycle through real-time image data, and mark them as images from the first monitoring time to the m-th monitoring time. Step S452: Obtain the garbage distribution density value of the cleaning area corresponding to the image at the first monitoring time, as follows: Step S453: Binarize the image at the first monitoring time, mark the pixel value corresponding to the area where the visible garbage is located as 1, and mark the pixel value of the non-visible garbage area as 0, to obtain the binarized image of the cleaning area; Step S454: Count the number of pixels with a pixel value of 1 in the binarized image of the cleaning area to obtain the number of pixels in the visible garbage area, and count the number of pixels with a pixel value of 0 in the binarized image of the cleaning area to obtain the number of pixels in the non-visible garbage area. Step S455: Summate the number of pixels in the visible waste area with the number of pixels in the non-visible waste area to obtain the total number of pixels in the cleaning area. Calculate the ratio of the number of pixels in the visible waste area to the total number of pixels in the cleaning area to obtain the waste distribution density value at the first monitoring time. Step S456: Repeat the process of obtaining the garbage distribution density value at the first monitoring time, and obtain the garbage distribution density value corresponding to the image at each monitoring time to obtain the garbage distribution density values ​​at the first to the mth monitoring times. Step S457: Mark the image acquisition time values ​​corresponding to the first monitoring time image to the m-th detection time image as the first to the m-th time values, and establish a line graph of waste density change over time with the time values ​​as the horizontal axis and the waste distribution density values ​​as the vertical axis. Step S458: Using image analysis tools, obtain the slope value of the line connecting every two consecutive coordinate points in the line graph of waste density change over time, obtain the waste density growth rate corresponding to the image at different detection times, obtain the waste density growth rate threshold, compare each waste density growth rate with the waste density growth rate threshold, and make real-time cleaning plans based on the comparison results.

8. The intelligent cleaning strategy planning method based on fusion feature analysis according to claim 7, characterized in that, In step S458, the specific steps are as follows: Step S4581: When the rate of increase in waste density is greater than or equal to the threshold for the rate of increase in waste density, it is determined to be an abnormal rate of increase in waste density; Step S4582: When the rate of increase in waste density is less than the threshold for the rate of increase in waste density, it is determined to be a normal rate of increase in waste density. Step S4583: If there is an abnormal growth rate of garbage density in the line graph of garbage density change over time, immediately arrange for cleaning personnel to clean the sub-area. Step S4584: If there is no abnormal growth rate of garbage density in the line graph of garbage density change over time, then clean the sub-area according to the regular cleaning frequency.

9. A smart cleaning strategy planning system based on fusion feature analysis, applicable to the smart cleaning strategy planning method based on fusion feature analysis as described in any one of claims 1-8, characterized in that, It includes a data acquisition module, a data analysis module, a frequency planning module, and a real-time planning module; Data acquisition module: Acquires basic data for cleaning planning; Data Analysis Module: Analyzes the basic data of the cleaning plan to obtain cleaning plan analysis data; Frequency planning module: Plans cleaning frequency for cleaning sub-areas based on cleaning planning analysis data; Real-time planning module: Performs real-time cleaning planning for cleaning sub-areas based on basic cleaning planning data.

Images