A garbage disposal point evaluation method, device, readable storage medium and server

Abstract

The present invention belongs to the technical field of artificial intelligence, and particularly relates to a method, device, computer-readable storage medium and server for evaluating garbage disposal points. The method includes: obtaining a first remote sensing image of a specified area at a first moment and a second remote sensing image of the specified area at a second moment from a remote sensing image database; calculating a measurement value of the water body change situation of the specified area according to the first remote sensing image and the second remote sensing image; calculating a measurement value of the vegetation change situation of the specified area according to the first remote sensing image and the second remote sensing image; obtaining evaluation information of the garbage disposal point to be evaluated in each evaluation dimension from a garbage disposal point database; constructing a feature vector of the garbage disposal point to be evaluated according to the measurement value of the water body change situation, the measurement value of the vegetation change situation and the evaluation information; and processing the feature vector by using a garbage disposal point evaluation model to obtain an evaluation result of the garbage disposal point to be evaluated.

Description

Technical Field

[0001] The present invention belongs to the technical field of artificial intelligence, and particularly relates to a method, device, computer-readable storage medium and server for evaluating garbage disposal points. Background Art

[0002] Garbage disposal points are important public facilities integrating functions such as garbage collection, storage, and transfer, and are closely related to the daily lives of the general public. Whether garbage disposal points can be effectively managed is a major livelihood matter concerning every household and has received increasing attention and concern.

[0003] In the prior art, in order to promote the enthusiasm of garbage disposal point administrators for garbage disposal point management, effectively improve the environment of garbage disposal points in various places, and enhance the usage experience of local residents, experts are generally invited to evaluate garbage disposal points regularly, and rewards and punishments are given according to the evaluation results. However, this method relying on expert evaluation is very time-consuming and laborious, with low efficiency and difficult to meet the current actual needs. Summary of the Invention

[0004] In view of this, embodiments of the present invention provide a method, device, computer-readable storage medium and server for evaluating garbage disposal points to solve the problem of low efficiency of existing garbage disposal point evaluation methods.

[0005] The first aspect of the embodiments of the present invention provides a method for evaluating a garbage disposal point, which may include:

[0006] Obtaining a first remote sensing image of a specified area at a first moment and a second remote sensing image of the specified area at a second moment from a preset remote sensing image database; the specified area is the area where the garbage disposal point to be evaluated is located;

[0007] Calculating a water body change situation metric value of the specified area according to the first remote sensing image and the second remote sensing image;

[0008] Calculating a vegetation change situation metric value of the specified area according to the first remote sensing image and the second remote sensing image;

[0009] Obtaining evaluation information of the garbage disposal point to be evaluated in preset evaluation dimensions from a preset garbage disposal point database;

[0010] Constructing a feature vector of the garbage disposal point to be evaluated according to the water body change situation metric value, the vegetation change situation metric value and the evaluation information;

[0011] Processing the feature vector using a preset garbage disposal point evaluation model to obtain an evaluation result of the garbage disposal point to be evaluated.

[0012] In a specific implementation manner of the first aspect, calculating the measurement value of the water body change situation in the specified area according to the first remote sensing image and the second remote sensing image includes:

[0013] Perform atmospheric correction processing on the first remote sensing image to obtain a first corrected image;

[0014] Calculate the normalized difference water index (NDWI) of each pixel in the first corrected image respectively to generate a first normalized difference water index image corresponding to the first corrected image; wherein, the calculation formula of the normalized difference water index of any one pixel is:

[0015]

[0016] Green is the remote sensing reflectance value of the pixel in the green band, NIR is the remote sensing reflectance value of the pixel in the near-infrared band, and NDWI is the normalized difference water index of the pixel;

[0017] Use a preset water body segmentation model to perform water body segmentation on the first normalized difference water index image to obtain a first water body segmentation image;

[0018] Perform atmospheric correction processing on the second remote sensing image to obtain a second corrected image;

[0019] Calculate the normalized difference water index of each pixel in the second corrected image respectively to generate a second normalized difference water index image corresponding to the second corrected image;

[0020] Use the water body segmentation model to perform water body segmentation on the second normalized difference water index image to obtain a second water body segmentation image;

[0021] Calculate the measurement value of the water body change situation in the specified area according to the first water body segmentation image and the second water body segmentation image.

[0022] In a specific implementation manner of the first aspect, the water body segmentation model includes a downsampling layer, an intermediate processing layer, and an upsampling layer. Using the preset water body segmentation model to perform water body segmentation on the first normalized difference water index image to obtain a first water body segmentation image includes:

[0023] Use the downsampling layer to perform convolution and downsampling processing on the first normalized difference water index image to obtain a first processing result;

[0024] Use the preset residual module in the intermediate processing layer to perform feature extraction on the first processing result, and superimpose the feature extraction result on the first processing result to obtain a second processing result;

[0025] Use the upsampling layer to perform convolution and upsampling on the second processing result to obtain a first water body segmentation image corresponding to the first normalized water body index image.

[0026] In a specific implementation manner of the first aspect, the training process of the water body segmentation model includes:

[0027] Obtain a training sample set from a preset training sample database; the training sample set includes each training sample, and each training sample includes a normalized water body index sample image and an expected water body segmentation image, where the expected water body segmentation image in each training sample has a one-to-one correspondence with the normalized water body index sample image;

[0028] Input the normalized water body index sample images in each training sample into the water body segmentation model for processing to obtain an actual water body segmentation image;

[0029] Calculate the difference degree between the expected water body segmentation image and the actual water body segmentation image in each training sample;

[0030] If the difference degree is greater than a preset first threshold, adjust the model parameters of the water body segmentation model, and return to execute the step of inputting the normalized water body index sample images in each training sample into the water body segmentation model for processing until the difference degree is less than or equal to the first threshold.

[0031] In a specific implementation manner of the first aspect, calculating the water body change situation measurement value of the specified area according to the first water body segmentation image and the second water body segmentation image includes:

[0032] Perform binarization on the first water body segmentation image to obtain a first binary image; perform binarization on the second water body segmentation image to obtain a second binary image;

[0033] Count the number of water body pixels in the first binary image and the number of water body pixels in the second binary image respectively; where the value of the water body pixel is a preset first value, and the value of the non-water body pixel is a preset second value;

[0034] Calculate the water body change situation measurement value of the specified area according to the following formula:

[0035]

[0036] Where WaterPixN1 is the number of water body pixels in the first binary image, WaterPixN2 is the number of water body pixels in the second binary image, and WaterIndex is the water body change situation measurement value.

[0037] In a specific implementation of the first aspect, calculating the vegetation change metric value of the specified area based on the first remote sensing image and the second remote sensing image includes:

[0038] Performing atmospheric correction processing on the first remote sensing image to obtain a third corrected image;

[0039] Calculating the normalized difference vegetation index (NDVI) of each pixel in the third corrected image respectively to generate a first NDVI image corresponding to the third corrected image; wherein, the calculation formula of the NDVI of any pixel is:

[0040]

[0041] Red is the remote sensing reflectance value of the pixel in the red band, NIR is the remote sensing reflectance value of the pixel in the near-infrared band, and NDVI is the normalized difference vegetation index of the pixel;

[0042] Using a preset vegetation segmentation model to perform vegetation segmentation on the first NDVI image to obtain a first vegetation segmentation image;

[0043] Performing atmospheric correction processing on the second remote sensing image to obtain a fourth corrected image;

[0044] Calculating the NDVI of each pixel in the fourth corrected image respectively to generate a second NDVI image corresponding to the fourth corrected image;

[0045] Using the vegetation segmentation model to perform vegetation segmentation on the second NDVI image to obtain a second vegetation segmentation image;

[0046] Calculating the vegetation change metric value of the specified area based on the first vegetation segmentation image and the second vegetation segmentation image.

[0047] In a specific implementation of the first aspect, calculating the vegetation change metric value of the specified area based on the first vegetation segmentation image and the second vegetation segmentation image includes:

[0048] Performing binarization processing on the first vegetation segmentation image to obtain a third binarized image; performing binarization processing on the second vegetation segmentation image to obtain a fourth binarized image;

[0049] Counting the number of vegetation pixels in the third binarized image and the number of vegetation pixels in the fourth binarized image respectively; wherein, the value of the vegetation pixel is a preset third value, and the value of the non-vegetation pixel is a preset fourth value;

[0050] Calculate the measurement value of the vegetation change in the specified area according to the following formula:

[0051]

[0052] Wherein, VegePixN1 is the number of vegetation pixels in the third binary image, VegePixN2 is the number of vegetation pixels in the fourth binary image, and VegeIndex is the measurement value of the vegetation change situation.

[0053] The second aspect of the embodiments of the present invention provides a garbage disposal point evaluation device, which may include functional modules for implementing the steps of any of the above garbage disposal point evaluation methods.

[0054] The third aspect of the embodiments of the present invention provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of any of the above garbage disposal point evaluation methods are implemented.

[0055] The fourth aspect of the embodiments of the present invention provides a server, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of any of the above garbage disposal point evaluation methods are implemented.

[0056] The fifth aspect of the embodiments of the present invention provides a computer program product. When the computer program product runs on a server, the server is caused to execute the steps of any of the above garbage disposal point evaluation methods.

[0057] The beneficial effects of the embodiments of the present invention compared with the prior art are as follows: The embodiments of the present invention obtain a first remote sensing image of a specified area at a first moment and a second remote sensing image of the specified area at a second moment from a preset remote sensing image database; calculate a measurement value of the water body change situation of the specified area according to the first remote sensing image and the second remote sensing image; calculate a measurement value of the vegetation change situation of the specified area according to the first remote sensing image and the second remote sensing image; obtain evaluation information of the garbage disposal point to be evaluated in each preset evaluation dimension from a preset garbage disposal point database; construct a feature vector of the garbage disposal point to be evaluated according to the measurement value of the water body change situation, the measurement value of the vegetation change situation, and the evaluation information; and use a preset garbage disposal point evaluation model to process the feature vector to obtain an evaluation result of the garbage disposal point to be evaluated. Through the embodiments of the present invention, it is possible to evaluate the garbage disposal point through the intelligent processing of the model without relying on experts, greatly improving the efficiency. Description of the Drawings

[0058] To more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the accompanying drawings required for use in the embodiments or the description of the prior art. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can also be obtained based on these drawings.

[0059] Figure 1 It is a flowchart of an embodiment of a garbage disposal point evaluation method in an embodiment of the present invention;

[0060] Figure 2 It is a schematic flowchart for calculating the measurement value of the water body change situation in a specified area based on the first remote sensing image and the second remote sensing image;

[0061] Figure 3 It is a schematic flowchart for calculating the measurement value of the vegetation change situation in a specified area based on the first remote sensing image and the second remote sensing image;

[0062] Figure 4 It is a structural diagram of an embodiment of a garbage disposal point evaluation device in an embodiment of the present invention;

[0063] Figure 5 It is a schematic block diagram of a server in an embodiment of the present invention. Detailed implementation manners

[0064] To make the invention objectives, features, and advantages of the present invention more obvious and understandable, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the embodiments described below are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0065] The embodiments of the present invention can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, sense the environment, acquire knowledge, and use knowledge to obtain the best results in theory, methods, technologies, and application systems.

[0066] Artificial intelligence basic technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technologies, operation / interaction systems, and mechatronics. Artificial intelligence software technologies mainly include several major directions such as computer vision technology, robotics, biometric technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0067] The execution subject of the embodiments of the present invention can be an artificial intelligence-based server, which is used to execute the garbage disposal point evaluation method in the embodiments of the present invention. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, Content Delivery Network (CDN), and big data and artificial intelligence platforms.

[0068] Please refer to Figure 1 , an embodiment of a garbage disposal point evaluation method in the embodiments of the present invention may include:

[0069] Step 101: Obtain a first remote sensing image of a specified area at a first moment and a second remote sensing image of the specified area at a second moment from a preset remote sensing image database.

[0070] The specified area is the area where the garbage disposal point to be evaluated is located, and its specific shape and size can be set according to actual situations. For example, a circular area can be constructed with the garbage disposal point to be evaluated as the center and a preset distance threshold as the radius, and this circular area can be used as the specified area. The distance threshold can be set according to actual situations. For example, it can be set to 5 kilometers. Of course, it can also be set to other values, and the embodiments of the present invention do not make specific limitations on this.

[0071] The first moment is a recent moment, and the second moment is an earlier moment. Both the first moment and the second moment can be set according to actual situations. For example, when it is necessary to evaluate the garbage disposal point to be evaluated, a moment that is before the current moment and the interval between it and the current moment is less than a preset first interval threshold can be used as the first moment. The first interval threshold can be set according to actual situations. For example, it can be set to 1 day, 1 week, etc. A moment that is before the current moment and the interval between it and the current moment is greater than a preset second interval threshold can be used as the second moment. The second interval threshold can be set according to actual situations. For example, it can be set to 1 month, 1 quarter, etc.

[0072] When it is necessary to evaluate the garbage disposal point to be evaluated, the server can obtain the remote sensing image of the specified area at the first moment, that is, the recent remote sensing image, from the remote sensing image database, record it as the first remote sensing image, and obtain the remote sensing image of the specified area at the second moment, that is, the earlier remote sensing image, from the remote sensing image database, and record it as the second remote sensing image.

[0073] Step 102: Calculate the measurement value of the water body change in the specified area according to the first remote sensing image and the second remote sensing image.

[0074] As Figure 2 shown, step 102 may include the following process:

[0075] Step 1021: Perform atmospheric correction processing on the first remote sensing image to obtain a first corrected image.

[0076] First, atmospheric correction processing can be performed on the first remote sensing image to eliminate the influence of factors such as the atmosphere and illumination on the reflection of ground objects. For example, the internal average relative reflectance method can be used for the atmospheric correction of the first remote sensing image, and each pixel in each band of the first remote sensing image is divided by the average value of the gray values of the entire image in that band. Of course, other atmospheric correction processing methods can also be used, and the embodiments of the present invention do not make specific limitations in this regard.

[0077] Step 1022: Calculate the normalized difference water index (NDWI) of each pixel in the first corrected image respectively, and generate a first normalized difference water index image corresponding to the first corrected image.

[0078] After obtaining the first corrected image, taking any one pixel as an example, the normalized difference water index of this pixel can be calculated respectively according to the following formula:

[0079]

[0080] where Green is the remote sensing reflectance value of this pixel in the green band, NIR is the remote sensing reflectance value of this pixel in the near-infrared band, and NDWI is the normalized difference water index of this pixel.

[0081] Traverse each pixel in the first corrected image, and the corresponding normalized difference water indices can be obtained, thereby generating a normalized difference water index image corresponding to the first corrected image, which is denoted as the first normalized difference water index image. The value of any one pixel in the first normalized difference water index image is the normalized difference water index of the pixel at the corresponding position in the first corrected image.

[0082] Step 1023: Use a preset water body segmentation model to perform water body segmentation on the first normalized difference water index image to obtain a first water body segmentation image.

[0083] The water body segmentation model includes three processing layers: a downsampling layer, an intermediate processing layer, and an upsampling layer. The specific processing process may include the following steps:

[0084] Step 10231: Use the downsampling layer to perform convolution and downsampling processing on the first normalized difference water index image to obtain a first processing result.

[0085] Here, the number of layers of the downsampling layer is denoted as LayerNum1. The specific value of LayerNum1 can be set according to the actual situation. In the embodiment of the present invention, it is preferably set that LayerNum1 = 3, and these three layers are sequentially denoted as DL1, DL2, and DL3. Among them, DL1 is the first layer of the downsampling layer, whose input is the first normalized water body index image, and the output is the result obtained by performing convolution and downsampling processing on the first normalized water body index image in DL1. Here, the output of DL1 is denoted as DL1_Res; DL2 is the second layer of the downsampling layer, whose input is DL1_Res, and the output is the result obtained by performing convolution and downsampling processing on DL1_Res in DL2. Here, the output of DL2 is denoted as DL2_Res; DL3 is the third layer of the downsampling layer, whose input is DL2_Res, and the output is the result obtained by performing convolution and downsampling processing on DL2_Res in DL3.

[0086] Taking the specific processing process in DL1 as an example, it is described in detail as follows: The number of channels of the first normalized water body index image is 1, that is, the NDWI value channel. The convolution kernel in DL1 performs convolution processing on the first normalized water body index image to obtain a feature map of the first normalized water body index image. The number of convolution kernels in DL1 can be set according to the actual situation. In this embodiment, it is preferably set to 8. Then, after the first normalized water body index image undergoes convolution processing in DL1, an 8-channel feature map can be obtained. Then, an activation function (for example, the Rectified Linear Unit (ReLU) can be used as the activation function) is used to process the feature map to limit the values in the feature map within the range of [0, 1]. Then, downsampling processing is performed on the feature map to reduce the scale of the feature map. For example, through downsampling processing, the length and width of the feature map can be reduced to half of the original. The feature map after the downsampling processing in DL1 will be used as the input of DL2.

[0087] The processing processes of DL2 and DL3 are similar to that of DL1, which will not be elaborated here. However, it should be noted that the number of convolution kernels in DL2 is twice that of the convolution kernels in DL1, and the number of convolution kernels in DL3 is twice that of the convolution kernels in DL2. In this way, the number of channels of the feature maps output by DL1, DL2, and DL3 are 8, 16, and 32 in sequence. Finally, the feature map output by DL3 is the first processing result.

[0088] Step 10232: Use the preset residual module in the intermediate processing layer to extract features from the first processing result, and superimpose the feature extraction result on the first processing result to obtain a second processing result.

[0089] The intermediate processing layer includes a residual module. The structure of the residual module includes two branches. Among them, the first branch is used to extract deeper features from the first processing result, and the second branch is used to maintain the first processing result. In the first branch, the first processing result can successively go through processes such as convolution processing, ReLU function processing, and convolution processing. The number of convolution kernels for convolution processing in the first branch is the same as the number of convolution kernels in DL3. Therefore, during the entire processing process, the number of channels of the feature map remains unchanged; while in the second branch, through the way of skip connection, the first processing result can skip the processing process in the first branch, and the data on the two branches are weighted and superimposed to obtain the second processing result. Through such a processing method, the high-frequency characteristics of the data can be effectively maintained, and the problems of gradient disappearance and gradient explosion that may be caused by the deepening of the network depth are solved, so that while training a deeper neural network, good performance can be guaranteed.

[0090] Step 10233: Use the upsampling layer to perform convolution and upsampling processing on the second processing result to obtain a first water body segmentation image corresponding to the first normalized water body index image.

[0091] The number of layers of the upsampling layer is the same as the number of layers of the downsampling layer. Here, still taking the case where the number of layers is 3 as an example, these three layers are sequentially denoted as UL1, UL2, and UL3. Among them, UL1 is the first layer of the upsampling layer, its input is the second processing result, and its output is the result obtained by performing convolution and upsampling processing on the second processing result in UL1. Here, the output of UL1 is denoted as UL1_Res; UL2 is the second layer of the upsampling layer, its input is UL1_Res, and its output is the result obtained by performing convolution and upsampling processing on UL1_Res in UL2. Here, the output of UL2 is denoted as UL2_Res; UL3 is the third layer of the upsampling layer, its input is UL2_Res, and its output is the result obtained by performing convolution and upsampling processing on UL2_Res in UL3.

[0092] Taking the specific processing process in UL1 as an example, the details are as follows: The number of convolution kernels in UL1 is the same as the number of convolution kernels in DL2. Then, after the second processing result passes through the convolution processing of UL1, a feature map with 16 channels can be obtained. Then, an activation function (for example, ReLU can be used as the activation function) is used to process the feature map to limit the values in the feature map within the range of [0,1]. Then, the feature map is subjected to upsampling processing to expand the scale of the feature map. For example, the length and width of the feature map can be expanded to twice the original through upsampling processing. The feature map after the upsampling processing of UL1 will be used as the input of UL2.

[0093] The processing procedures of UL2 and UL3 are similar to that of UL1, which will not be elaborated here. However, it should be noted that the number of convolutional kernels in UL2 is half of that in UL1, and the number of convolutional kernels in UL3 is 1. In this way, the number of channels of the feature maps output by UL1, UL2, and UL3 are 16, 8, and 1 in sequence. Finally, the feature map output by UL3 can be further processed by 1 time of convolution (the number of convolutional kernels is 1) and 1 time of activation function (for example, Sigmoid can be used as the activation function) to obtain the first water body segmentation image corresponding to the first normalized water body index image. It should be noted that a skip connection is also introduced between the downsampling layer and the upsampling layer. Before each convolution processing in the upsampling layer, the data to be convolved is superimposed with the output result of the same number of channels in the downsampling layer, and the superimposed result is used as the input for the next convolution processing.

[0094] The water body segmentation model is obtained through training based on samples. Before being put into use, it can be pre-trained through the following process:

[0095] Step 201: Obtain a training sample set from a preset training sample database.

[0096] The training sample set includes each training sample, and each training sample includes a normalized water body index sample image and an expected water body segmentation image. Among them, the expected water body segmentation image in each training sample has a one-to-one correspondence with the normalized water body index sample image. The expected water body segmentation image is a binary image. In this image, the pixel value corresponding to the water body is a preset first value, and the pixel value corresponding to the non-water body is a preset second value. The specific values of the first value and the second value can be set according to the actual situation. For example, the first value can be set to 1 and the second value can be set to 0. Of course, the first value and the second value can also be set to other values. The embodiments of the present invention do not make specific limitations on this.

[0097] Step 202: Input the normalized water body index sample images in each training sample into the water body segmentation model for processing to obtain actual water body segmentation images.

[0098] The specific process of Step 202 can refer to the detailed description in Step 1023, which will not be elaborated here.

[0099] Step 203: Calculate the difference degree between the expected water body segmentation image and the actual water body segmentation image in each training sample.

[0100] Specifically, the difference degree can be calculated according to the following formula:

[0101]

[0102] Where n is the serial number of the training sample, 1≤n≤N, N is the total number of training samples, pix is ​​the serial number of the pixel, 1≤pix≤PixN, PixN is the total number of pixels in the image, s n,pix is the value of the pixth pixel of the expected water body segmentation image of the nth training sample, y n,pix is the value of the pixth pixel of the actual water body segmentation image of the nth training sample, and Loss is the difference.

[0103] Step 204: Determine whether the difference is greater than a preset first threshold.

[0104] If the difference is greater than the first threshold, step 205 is executed; if the difference is less than or equal to the first threshold, step 206 is executed.

[0105] Step 205: Adjust the model parameters of the water body segmentation model.

[0106] After completing the parameter adjustment, the process returns to step 202 , that is, continues to use the training sample set to train the water body segmentation model until the difference is less than or equal to the first threshold.

[0107] Step 206: The training of the water body segmentation model is completed.

[0108] When the difference is less than or equal to the first threshold, it means that the training has achieved the predetermined effect and the training can be terminated. At this time, the determined water body segmentation model has been trained with a large number of samples, and its difference is kept in a small range. Using the water body segmentation model to perform water body segmentation on the first normalized water body index image can obtain a better processing effect.

[0109] Step 1024: Perform atmospheric correction processing on the second remote sensing image to obtain a second corrected image.

[0110] Step 1025: Calculate the normalized water index of each pixel in the second corrected image respectively, and generate a second normalized water index image corresponding to the second corrected image.

[0111] Step 1026: Use the water body segmentation model to perform water body segmentation on the second normalized water body index image to obtain a second water body segmentation image.

[0112] The specific process of step 1024 to step 1026 can refer to the detailed description of step 1021 to step 1023, which will not be repeated here.

[0113] Step 1027. Calculate the water body change measurement value of the specified area according to the first water body segmentation image and the second water body segmentation image.

[0114] First, perform binarization processing on the first water body segmentation image to obtain a first binary image; perform binarization processing on the second water body segmentation image to obtain a second binary image.

[0115] After binarization processing, pixels are divided into two categories, namely water body pixels and non-water body pixels. Among them, the value of water body pixels is the first numerical value, and the value of non-water body pixels is the second numerical value.

[0116] Then, respectively count the number of water body pixels in the first binary image and the number of water body pixels in the second binary image, and calculate the water body change measurement value of the specified area according to the following formula:

[0117]

[0118] where WaterPixN1 is the number of water body pixels in the first binary image, WaterPixN2 is the number of water body pixels in the second binary image, and WaterIndex is the water body change measurement value.

[0119] Step 103. Calculate the vegetation change measurement value of the specified area according to the first remote sensing image and the second remote sensing image.

[0120] As Figure 3 shown, step 103 may include the following process:

[0121] Step 1031. Perform atmospheric correction processing on the first remote sensing image to obtain a third corrected image.

[0122] The specific process of step 1031 may refer to the detailed description in step 1021 and will not be elaborated here.

[0123] Step 1032. Calculate the normalized difference vegetation index of each pixel in the third corrected image respectively, and generate a first normalized difference vegetation index image corresponding to the third corrected image.

[0124] After obtaining the third corrected image, taking any one pixel as an example, the normalized difference vegetation index of this pixel can be calculated respectively according to the following formula:

[0125]

[0126] Wherein, Red is the remote sensing reflectance value of the pixel in the red band, NIR is the remote sensing reflectance value of the pixel in the near-infrared band, and NDVI is the normalized difference vegetation index of the pixel.

[0127] By traversing each pixel in the third corrected image, the corresponding normalized difference vegetation indices can be obtained, thereby generating a normalized difference vegetation index image corresponding to the third corrected image, which is denoted as the first normalized difference vegetation index image. The value of any pixel in the first normalized difference vegetation index image is the normalized difference vegetation index of the pixel at the corresponding position in the third corrected image.

[0128] Step 1033: Use a preset vegetation segmentation model to perform vegetation segmentation on the first normalized difference vegetation index image to obtain a first vegetation segmentation image.

[0129] The structure, processing process, and training process of the vegetation segmentation model are similar to those of the water body segmentation model. For specific details, reference can be made to the detailed description of the water body segmentation model, which will not be elaborated here.

[0130] It should be noted that during its training process, each training sample includes a normalized difference vegetation index sample image and an expected vegetation segmentation image. Among them, the expected vegetation segmentation image in each training sample has a one-to-one correspondence with the normalized difference vegetation index sample image. The expected vegetation segmentation image is a binary image. In this image, the value of the pixel corresponding to the vegetation is a preset third value, and the value of the pixel corresponding to non-vegetation is a preset fourth value. The specific values of the third value and the fourth value can be set according to actual situations. For example, the third value can be set to 1 and the fourth value can be set to 0. Of course, the third value and the fourth value can also be set to other values. The embodiments of the present invention do not make specific limitations on this.

[0131] Step 1034: Perform atmospheric correction processing on the second remote sensing image to obtain a second corrected image.

[0132] Step 1035: Calculate the normalized difference vegetation index of each pixel in the second corrected image respectively to generate a second normalized difference vegetation index image corresponding to the second corrected image.

[0133] Step 1036: Use the vegetation segmentation model to perform vegetation segmentation on the second normalized difference vegetation index image to obtain a second vegetation segmentation image.

[0134] The specific processes of Step 1034 to Step 1036 can refer to the detailed description in Step 1031 to Step 1033, which will not be elaborated here.

[0135] Step 1037: Calculate the vegetation change measurement value of the specified area according to the first vegetation segmentation image and the second vegetation segmentation image.

[0136] First, perform binarization processing on the first vegetation segmentation image to obtain a third binarized image; perform binarization processing on the second vegetation segmentation image to obtain a fourth binarized image.

[0137] After binarization processing, pixels are divided into two categories, namely vegetation pixels and non-vegetation pixels. Among them, the value of vegetation pixels is the third numerical value, and the value of non-vegetation pixels is the fourth numerical value.

[0138] Then, respectively count the number of vegetation pixels in the third binarized image and the number of vegetation pixels in the fourth binarized image, and calculate the vegetation change measurement value of the specified area according to the following formula:

[0139]

[0140] Among them, VegePixN1 is the number of vegetation pixels in the third binarized image, VegePixN2 is the number of vegetation pixels in the fourth binarized image, and VegeIndex is the vegetation change measurement value.

[0141] Step 104: Obtain the evaluation information of the to-be-evaluated garbage disposal point in each preset evaluation dimension from the preset garbage disposal point database.

[0142] Specific evaluation dimensions can be set according to actual situations. For example, they can include but are not limited to tidiness, attitude of management staff, garbage classification publicity, supervision, activity organization, signpost setting, hardware equipment, citizen participation rate, classification accuracy rate of waste disposal, and garbage bin setting, etc. The evaluation information for each evaluation dimension is represented in the form of scores.

[0143] For the evaluation dimension of tidiness, if the ground and wall environment around the garbage disposal point is very tidy, the score is 5 points; if it is tidy, the score is 3 points; if it is average, the score is 0 points; if it is untidy, the score is -5 points.

[0144] For the evaluation dimension of the attitude of management staff, if the attitude is good, the score is 5 points; if it is average, the score is 0 points; if it is relatively bad, the score is -5 points.

[0145] For the evaluation dimension of garbage classification publicity, check whether there are public service advertisements and classification common sense publicity on garbage classification in the public area of the garbage disposal point to create a strong atmosphere for promoting the whole people to participate in garbage classification. If so, the score is 5 points; if not, the score is 0 points.

[0146] For the evaluation dimension of supervision, the score for setting up a publicity platform for domestic waste classification at waste disposal points and giving play to the supervision role of the public and the media is 5 points if it is carried out, and 0 points if it is not carried out.

[0147] For the evaluation dimension of activity organization, the score for organizing publicity activities, such as prize-winning activities and knowledge contests, which are rich and colorful and popular among the masses, to attract the enthusiasm of the masses to participate in waste classification is 5 points if it is carried out, and 0 points if it is not carried out.

[0148] For the evaluation dimension of signpost setting, the main content of the signpost includes: disposal time, domestic waste classification guide, name of the collection and transportation unit and the person in charge, complaint phone number, name and phone number of the supervisor, etc. The score is 0 points if it is not set. If it is set in an irregular manner, 1 point will be deducted for each missing item until the full deduction. The score is 5 points if it is set perfectly.

[0149] For the evaluation dimension of hardware equipment, check whether the waste disposal points are equipped with hardware equipment such as washbasins, hand sanitizers, paper towels or dryers. The score for the disposal points with perfect hardware equipment is 5 points, and 1 point will be deducted for each missing item until the full deduction.

[0150] For the evaluation dimension of the public participation rate, in the areas where domestic waste classification is implemented, the public participation rate in the community should not be less than 80%. For each percentage point less, 1 point will be deducted until the full deduction. The score is 5 points if it is higher than 80%. The specific calculation formula for the public participation rate in the community is:

[0151] Public participation rate in the community = Number of people with records of participating in waste classification disposal during the assessment period / Total number of people × 100%

[0152] This score is randomly checked. One sub-district is randomly selected, and 10% of the communities in the randomly selected sub-district are evaluated, and the score is calculated based on the average.

[0153] For the evaluation dimension of the accurate classification disposal rate, in the areas where domestic waste classification is implemented, the accurate classification disposal rate of community residents should not be less than 90%. For each percentage point less, 1 point will be deducted until the full deduction. The score is 5 points if it is higher than 90%. The specific calculation formula for the accurate classification disposal rate is:

[0154] Accurate classification disposal rate = Amount of waste that should be disposed of in the randomly inspected trash cans / Total amount of waste collected in the randomly inspected trash cans × 100%. This score is randomly checked.

[0155] For the evaluation dimension of trash can setting, if the ground is not hardened, the classified trash cans are placed in a messy manner, the appearance of the can body is not clean, incomplete, damaged, or not airtight, 1 point will be deducted for each discovered case until the full deduction. The score is 5 points if all items are standardized and complete.

[0156] It should be noted that the above evaluation dimensions are only examples and not limitations. Other evaluation dimensions can also be set according to actual situations.

[0157] Step 105: Construct a feature vector of the garbage disposal point to be evaluated according to the water body change measurement value, the vegetation change measurement value, and the evaluation information.

[0158] Specifically, the feature vector of the garbage disposal point to be evaluated can be constructed according to the following formula:

[0159] InfoVec=(WaterIndex,VegeIndex,InfoEm 1 ,InfoEm 2 ,...,InfoEm dn ,...,InfoEm DN )

[0160] where dn is the serial number of the evaluation dimension, 1≤dn≤DN, DN is the total number of evaluation dimensions, InfoEm dn is the evaluation information on the dn-th evaluation dimension, and InfoVec is the feature vector of the garbage disposal point to be evaluated.

[0161] Step 106: Use a preset garbage disposal point evaluation model to process the feature vector to obtain an evaluation result of the garbage disposal point to be evaluated.

[0162] The garbage disposal point evaluation model in the embodiments of the present invention may include an input layer, a hidden layer, and an output layer. The input layer is used to receive input data from the outside, including more than two input layer nodes. The hidden layer is used to process the data, including more than two hidden layer nodes. The output layer is used to output the processing result, including one output layer node. The processing process of the garbage disposal point evaluation model may include:

[0163] Step S1061: Determine the feature vector as the input layer node data of the garbage disposal point evaluation model.

[0164] The input layer nodes correspond one by one to the elements in the feature vector. For example, if there are 12 elements in the feature vector, namely element 1, element 2, element 3,..., then the number of input layer nodes of the corresponding garbage disposal point evaluation model should also be 12, namely input layer node 1, input layer node 2, input layer node 3,..., where input layer node 1 corresponds to element 1, input layer node 2 corresponds to element 2, input layer node 3 corresponds to element 3, and so on.

[0165] Step S1062: At the hidden layer nodes of the waste disposal point evaluation model, the input layer node data is processed respectively using the fuzzy Gaussian membership function to obtain the hidden layer node data.

[0166] In this embodiment, the hidden layer node data can be obtained through the following calculation formula:

[0167]

[0168] Where:

[0169] i is the label of the input layer node, and its value range is [1, n], where n is the number of input layer nodes;

[0170] j is the label of the hidden layer node, and its value range is [1, h], where h is the number of hidden layer nodes;

[0171] Φ j (x) is the hidden layer node data of the j-th hidden layer node;

[0172] G ij (x i ) is the i-th fuzzy Gaussian membership function of the j-th hidden layer node;

[0173] x is the input layer node data, and xi is the input layer node data of the i-th input layer node among them;

[0174] μ ij is the mathematical expectation of the i-th fuzzy Gaussian membership function of the j-th hidden layer node;

[0175] σ ij is the standard deviation of the i-th fuzzy Gaussian membership function of the j-th hidden layer node.

[0176] Preferably, the hidden layer node data can also be normalized to narrow the difference of the hidden layer node data. Specifically, the maximum value and the minimum value in the hidden layer node data can be obtained, and then the hidden layer node data is normalized according to the maximum value and the minimum value to obtain the normalized hidden layer node data.

[0177] For example, the hidden layer node data can be normalized through the following formula:

[0178]

[0179] Where:

[0180] Ψ j (x) is the normalized hidden layer node data of the j-th hidden layer node;

[0181] Φmax (x) is Φ j the maximum value in (x);

[0182] Φ min (x) is Φ j the minimum value in (x).

[0183] Step S1063: Use preset weights to perform weighted summation on the hidden layer node data respectively to obtain the output value of the garbage disposal point evaluation model.

[0184] For the hidden layer node data that has not been normalized, the calculation formula for the output value can be:[[]]

[0185]

[0186] For the normalized hidden layer node data, the calculation formula for the output value can be:[[]]

[0187]

[0188] Where:[[]]

[0189] ω j is the weight corresponding to the hidden layer node data of the jth hidden layer node;

[0190] R(x) is the output layer node data, that is, the output value of the garbage disposal point evaluation model.

[0191] Step S1064: Determine the evaluation result of the garbage disposal point to be evaluated according to the output value.

[0192] In the embodiment of the present invention, output value intervals corresponding to various evaluation results can be preset in advance. After obtaining the output value of the garbage disposal point evaluation model, it is determined which output value interval corresponding to an evaluation result the output value belongs to, and the evaluation result corresponding to the output value interval to which the output value belongs is determined as the evaluation result of the garbage disposal point to be evaluated. The evaluation results can include but are not limited to excellent, good, general, poor, extremely poor, etc.

[0193] Before the garbage disposal point evaluation model is used, it can be trained based on the actual evaluation results of a large number of garbage disposal points by professionals as samples, so as to ensure the accuracy of the evaluation results.

[0194] In summary, the embodiments of the present invention obtain a first remote sensing image of a specified area at a first moment and a second remote sensing image of the specified area at a second moment from a preset remote sensing image database; calculate a measurement value of the water body change situation of the specified area according to the first remote sensing image and the second remote sensing image; calculate a measurement value of the vegetation change situation of the specified area according to the first remote sensing image and the second remote sensing image; obtain evaluation information of the to-be-evaluated garbage disposal point in each preset evaluation dimension from a preset garbage disposal point database; construct a feature vector of the to-be-evaluated garbage disposal point according to the measurement value of the water body change situation, the measurement value of the vegetation change situation, and the evaluation information; and use a preset garbage disposal point evaluation model to process the feature vector to obtain an evaluation result of the to-be-evaluated garbage disposal point. Through the embodiments of the present invention, it is possible to evaluate the garbage disposal point through the intelligent processing of the model instead of relying on experts, which greatly improves the efficiency.

[0195] It should be understood that the magnitudes of the sequence numbers of the steps in the above embodiments do not mean the order of execution. The order of execution of each process should be determined according to its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0196] Corresponding to the garbage disposal point evaluation method described in the above embodiments, Figure 4 FIG. shows a structural diagram of an embodiment of a garbage disposal point evaluation device provided by an embodiment of the present invention.

[0197] In this embodiment, a garbage disposal point evaluation device may include:

[0198] A remote sensing image acquisition module 401, configured to acquire a first remote sensing image of a specified area at a first moment and a second remote sensing image of the specified area at a second moment from a preset remote sensing image database; the specified area is the area where the to-be-evaluated garbage disposal point is located;

[0199] A water body change situation measurement module 402, configured to calculate a measurement value of the water body change situation of the specified area according to the first remote sensing image and the second remote sensing image;

[0200] A vegetation change situation measurement module 403, configured to calculate a measurement value of the vegetation change situation of the specified area according to the first remote sensing image and the second remote sensing image;

[0201] An evaluation information acquisition module 404, configured to acquire evaluation information of the to-be-evaluated garbage disposal point in each preset evaluation dimension from a preset garbage disposal point database;

[0202] A feature vector construction module 405, configured to construct a feature vector of the to-be-evaluated garbage disposal point according to the measurement value of the water body change situation, the measurement value of the vegetation change situation, and the evaluation information;

[0203] An evaluation result determination module 406, configured to process the feature vector by using a preset waste disposal point evaluation model to obtain an evaluation result of the to-be-evaluated waste disposal point.

[0204] In a specific implementation manner of the embodiment of the present invention, the water body change situation measurement module may include:

[0205] A first correction sub-module, configured to perform atmospheric correction processing on the first remote sensing image to obtain a first corrected image;

[0206] A first water body index image generation sub-module, configured to calculate the normalized difference water index of each pixel in the first corrected image respectively to generate a first normalized difference water index image corresponding to the first corrected image; wherein, the calculation formula of the normalized difference water index of any pixel is:

[0207]

[0208] Green is the remote sensing reflectance value of the pixel in the green band, NIR is the remote sensing reflectance value of the pixel in the near-infrared band, and NDWI is the normalized difference water index of the pixel;

[0209] A first water body segmentation image generation sub-module, configured to perform water body segmentation on the first normalized difference water index image by using a preset water body segmentation model to obtain a first water body segmentation image;

[0210] A second correction sub-module, configured to perform atmospheric correction processing on the second remote sensing image to obtain a second corrected image;

[0211] A second water body index image generation sub-module, configured to calculate the normalized difference water index of each pixel in the second corrected image respectively to generate a second normalized difference water index image corresponding to the second corrected image;

[0212] A second water body segmentation image generation sub-module, configured to perform water body segmentation on the second normalized difference water index image by using the water body segmentation model to obtain a second water body segmentation image;

[0213] A water body change situation measurement sub-module, configured to calculate a water body change situation measurement value of the specified area according to the first water body segmentation image and the second water body segmentation image.

[0214] In a specific implementation manner of the embodiment of the present invention, the first water body segmentation image generation sub-module may include:

[0215] A downsampling layer processing unit, configured to perform convolution and downsampling processing on the first normalized difference water index image by using the downsampling layer to obtain a first processing result;

[0216] An intermediate processing layer processing unit, configured to perform feature extraction on the first processing result by using a preset residual module in the intermediate processing layer, and superimpose the feature extraction result on the first processing result to obtain a second processing result;

[0217] An upsampling layer processing unit, configured to perform convolution and upsampling processing on the second processing result by using the upsampling layer to obtain a first water body segmentation image corresponding to the first normalized water body index image.

[0218] In a specific implementation manner of an embodiment of the present invention, the garbage disposal point evaluation device may further include:

[0219] A training sample set acquisition module, configured to acquire a training sample set from a preset training sample database; the training sample set includes each training sample, and each training sample includes a normalized water body index sample image and an expected water body segmentation image, wherein, the expected water body segmentation image in each training sample has a one-to-one correspondence with the normalized water body index sample image;

[0220] A model processing module, configured to input the normalized water body index sample images in each training sample into the water body segmentation model for processing to obtain an actual water body segmentation image;

[0221] A difference degree calculation module, configured to calculate the difference degree between the expected water body segmentation image and the actual water body segmentation image in each training sample;

[0222] A model adjustment module, configured to, if the difference degree is greater than a preset first threshold, adjust the model parameters of the water body segmentation model, and return to execute the step of inputting the normalized water body index sample images in each training sample into the water body segmentation model for processing until the difference degree is less than or equal to the first threshold.

[0223] In a specific implementation manner of an embodiment of the present invention, the water body change situation measurement sub-module may include:

[0224] A first binarization processing unit, configured to perform binarization processing on the first water body segmentation image to obtain a first binarized image; perform binarization processing on the second water body segmentation image to obtain a second binarized image;

[0225] A first statistics unit, configured to respectively count the number of water body pixels in the first binarized image and the number of water body pixels in the second binarized image; wherein, the value of the water body pixel is a preset first value, and the value of the non-water body pixel is a preset second value;

[0226] A first calculation unit for calculating a metric value of the water body change in the specified area according to the following formula:

[0227]

[0228] where WaterPixN1 is the number of water body pixels in the first binary image, WaterPixN2 is the number of water body pixels in the second binary image, and WaterIndex is the metric value of the water body change situation.

[0229] In a specific implementation manner of the embodiment of the present invention, the vegetation change situation metric module may include:

[0230] A third correction sub-module for performing atmospheric correction processing on the first remote sensing image to obtain a third corrected image;

[0231] A first vegetation index image generation sub-module for calculating the normalized vegetation index of each pixel in the third corrected image respectively to generate a first normalized vegetation index image corresponding to the third corrected image; wherein, the calculation formula of the normalized vegetation index of any pixel is:

[0232]

[0233] Red is the remote sensing reflectance value of the pixel in the red band, NIR is the remote sensing reflectance value of the pixel in the near-infrared band, and NDVI is the normalized vegetation index of the pixel;

[0234] A first vegetation segmentation image generation sub-module for performing vegetation segmentation on the first normalized vegetation index image using a preset vegetation segmentation model to obtain a first vegetation segmentation image;

[0235] A fourth correction sub-module for performing atmospheric correction processing on the second remote sensing image to obtain a fourth corrected image;

[0236] A second vegetation index image generation sub-module for calculating the normalized vegetation index of each pixel in the fourth corrected image respectively to generate a second normalized vegetation index image corresponding to the fourth corrected image;

[0237] A second vegetation segmentation image generation sub-module for performing vegetation segmentation on the second normalized vegetation index image using the vegetation segmentation model to obtain a second vegetation segmentation image;

[0238] A vegetation change situation metric sub-module for calculating the metric value of the vegetation change situation in the specified area according to the first vegetation segmentation image and the second vegetation segmentation image.

[0239] In a specific implementation manner of the embodiment of the present invention, the vegetation change situation measurement sub-module may include:

[0240] A second binarization processing unit, configured to perform binarization processing on the first vegetation segmentation image to obtain a third binarized image; and perform binarization processing on the second vegetation segmentation image to obtain a fourth binarized image;

[0241] A second statistics unit, configured to respectively count the number of vegetation pixels in the third binarized image and the number of vegetation pixels in the fourth binarized image; wherein, the value of a vegetation pixel is a preset third value, and the value of a non-vegetation pixel is a preset fourth value;

[0242] A second calculation unit, configured to calculate the measurement value of the vegetation change situation in the specified area according to the following formula:

[0243]

[0244] wherein, VegePixN1 is the number of vegetation pixels in the third binarized image, VegePixN2 is the number of vegetation pixels in the fourth binarized image, and VegeIndex is the measurement value of the vegetation change situation.

[0245] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described devices, modules, and units can refer to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0246] In the above embodiments, the descriptions of the respective embodiments have their own focuses. For parts not detailed or recorded in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0247] Figure 5 FIG. shows a schematic block diagram of a server provided by an embodiment of the present invention. For the sake of convenience of description, only parts related to the embodiment of the present invention are shown.

[0248] The server 5 may include: a processor 50, a memory 51, and computer-readable instructions 52 stored in the memory 51 and executable on the processor 50, such as computer-readable instructions for executing the above-mentioned garbage disposal point evaluation method. When the processor 50 executes the computer-readable instructions 52, the steps in the foregoing embodiments of various garbage disposal point evaluation methods are implemented, such as Figure 1 the steps S101 to S106 shown. Alternatively, when the processor 50 executes the computer-readable instructions 52, the functions of the respective modules / units in the foregoing device embodiments are implemented, such as Figure 4 the functions of the modules 401 to 406 shown.

[0249] Exemplarily, the computer-readable instructions 52 may be divided into one or more modules / units, which are stored in the memory 51 and executed by the processor 50 to implement the present invention. The one or more modules / units may be a series of computer-readable instruction segments capable of performing specific functions, and these instruction segments are used to describe the execution process of the computer-readable instructions 52 in the server 5.

[0250] The processor 50 may be a central processing unit (CPU), or may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0251] The memory 51 may be an internal storage unit of the server 5, such as the hard disk or memory of the server 5. The memory 51 may also be an external storage device of the server 5, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. provided on the server 5. Further, the memory 51 may also include both the internal storage unit and the external storage device of the server 5. The memory 51 is used to store the computer-readable instructions and other instructions and data required by the server 5. The memory 51 may also be used to temporarily store data that has been output or is to be output.

[0252] In various embodiments of the present invention, the functional units may be integrated in one processing unit, or each unit may exist physically alone, or two or more units may be integrated in one unit. The above-mentioned integrated units may be implemented in the form of hardware or in the form of software functional units.

[0253] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or 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 and includes several computer-readable instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store computer-readable instructions.

[0254] The above-described embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A method for evaluating a garbage disposal point, characterized in that, it includes: Obtain the first remote sensing image of a specified area at the first moment and the second remote sensing image of the specified area at the second moment from a preset remote sensing image database; The specified area is the area where the garbage disposal point to be evaluated is located; Calculate the measurement value of the water body change situation in the specified area according to the first remote sensing image and the second remote sensing image; Calculate the measurement value of the vegetation change situation in the specified area according to the first remote sensing image and the second remote sensing image; Obtain the evaluation information of the garbage disposal point to be evaluated in each preset evaluation dimension from a preset garbage disposal point database, including cleanliness, attitude of management personnel, garbage classification publicity, supervision, activity organization, sign setting, hardware equipment, citizen participation rate, classification accuracy rate of garbage disposal, and trash can setting; Construct a feature vector of the garbage disposal point to be evaluated according to the measurement value of the water body change situation, the measurement value of the vegetation change situation and the evaluation information; Use a preset garbage disposal point evaluation model to process the feature vector to obtain the evaluation result of the garbage disposal point to be evaluated.

2. The garbage disposal point evaluation method according to claim 1, characterized in that, The calculating the measurement value of the water body change situation in the specified area according to the first remote sensing image and the second remote sensing image includes: Perform atmospheric correction processing on the first remote sensing image to obtain a first corrected image; Calculate the normalized difference water index of each pixel in the first corrected image respectively to generate a first normalized difference water index image corresponding to the first corrected image; wherein, the calculation formula of the normalized difference water index of any pixel is: Green is the remote sensing reflectance value of the pixel in the green band, NIR is the remote sensing reflectance value of the pixel in the near-infrared band, and NDWI is the normalized difference water index of the pixel; Use a preset water body segmentation model to perform water body segmentation on the first normalized difference water index image to obtain a first water body segmentation image; Perform atmospheric correction processing on the second remote sensing image to obtain a second corrected image; Calculate the normalized difference water index of each pixel in the second corrected image respectively to generate a second normalized difference water index image corresponding to the second corrected image; Use the water body segmentation model to perform water body segmentation on the second normalized difference water index image to obtain a second water body segmentation image; Calculate the measurement value of the water body change situation in the specified area according to the first water body segmentation image and the second water body segmentation image.

3. The garbage disposal point evaluation method according to claim 2, characterized in that, The water body segmentation model includes a downsampling layer, an intermediate processing layer and an upsampling layer. The using a preset water body segmentation model to perform water body segmentation on the first normalized difference water index image to obtain a first water body segmentation image includes: Use the downsampling layer to perform convolution and downsampling processing on the first normalized difference water index image to obtain a first processing result; Feature extraction is performed on the first processing result using a preset residual module in the intermediate processing layer, and the feature extraction result is superimposed on the first processing result to obtain a second processing result; The second processing result is subjected to convolution and upsampling processing using the upsampling layer to obtain a first water body segmentation image corresponding to the first normalized water body index image.

4. The garbage disposal point evaluation method according to claim 2, wherein, the training process of the water body segmentation model includes: obtaining a training sample set from a preset training sample database; the training sample set includes each training sample, and each training sample includes a normalized water body index sample image and an expected water body segmentation image, wherein the expected water body segmentation image in each training sample has a one-to-one correspondence with the normalized water body index sample image; inputting the normalized water body index sample images in each training sample into the water body segmentation model for processing to obtain an actual water body segmentation image; calculating the difference degree between the expected water body segmentation image and the actual water body segmentation image in each training sample; if the difference degree is greater than a preset first threshold, adjusting the model parameters of the water body segmentation model, and returning to execute the step of inputting the normalized water body index sample images in each training sample into the water body segmentation model for processing until the difference degree is less than or equal to the first threshold.

5. The garbage disposal point evaluation method according to claim 2, wherein, calculating the water body change situation measurement value of the specified area according to the first water body segmentation image and the second water body segmentation image includes: performing binarization processing on the first water body segmentation image to obtain a first binarized image; performing binarization processing on the second water body segmentation image to obtain a second binarized image; respectively counting the number of water body pixels in the first binarized image and the number of water body pixels in the second binarized image; wherein, the value of the water body pixel is a preset first numerical value, and the value of the non-water body pixel is a preset second numerical value; calculating the water body change situation measurement value of the specified area according to the following formula: where WaterPixN1 is the number of water body pixels in the first binarized image, WaterPixN2 is the number of water body pixels in the second binarized image, and WaterIndex is the water body change situation measurement value.

6. The garbage disposal point evaluation method according to any one of claims 1 to 5, wherein, calculating the vegetation change situation measurement value of the specified area according to the first remote sensing image and the second remote sensing image includes: performing atmospheric correction processing on the first remote sensing image to obtain a third corrected image; respectively calculating the normalized vegetation index of each pixel in the third corrected image to generate a first normalized vegetation index image corresponding to the third corrected image; wherein, the calculation formula of the normalized vegetation index of any one pixel is: Red is the remote sensing reflectance value of the pixel in the red band, NIR is the remote sensing reflectance value of the pixel in the near-infrared band, and NDVI is the normalized difference vegetation index of the pixel; Perform vegetation segmentation on the first normalized difference vegetation index image using a preset vegetation segmentation model to obtain a first vegetation segmentation image; Perform atmospheric correction processing on the second remote sensing image to obtain a fourth corrected image; Calculate the normalized difference vegetation index of each pixel in the fourth corrected image respectively, and generate a second normalized difference vegetation index image corresponding to the fourth corrected image; Perform vegetation segmentation on the second normalized difference vegetation index image using the vegetation segmentation model to obtain a second vegetation segmentation image; Calculate the vegetation change situation measurement value of the specified area according to the first vegetation segmentation image and the second vegetation segmentation image.

7. The garbage disposal point evaluation method according to claim 6, characterized in that, The calculating the vegetation change situation measurement value of the specified area according to the first vegetation segmentation image and the second vegetation segmentation image includes: Perform binarization processing on the first vegetation segmentation image to obtain a third binarized image; perform binarization processing on the second vegetation segmentation image to obtain a fourth binarized image; Count the number of vegetation pixels in the third binarized image and the number of vegetation pixels in the fourth binarized image respectively; wherein, the value of the vegetation pixel is a preset third value, and the value of the non-vegetation pixel is a preset fourth value; Calculate the vegetation change situation measurement value of the specified area according to the following formula: wherein, VegePixN1 is the number of vegetation pixels in the third binarized image, VegePixN2 is the number of vegetation pixels in the fourth binarized image, and VegeIndex is the vegetation change situation measurement value.

8. A garbage disposal point evaluation device, characterized in that, comprising: A remote sensing image acquisition module, configured to acquire a first remote sensing image of a specified area at a first moment and a second remote sensing image of the specified area at a second moment from a preset remote sensing image database; The specified area is the area where the garbage disposal point to be evaluated is located; A water body change situation measurement module, configured to calculate the water body change situation measurement value of the specified area according to the first remote sensing image and the second remote sensing image; A vegetation change situation measurement module, configured to calculate the vegetation change situation measurement value of the specified area according to the first remote sensing image and the second remote sensing image; An evaluation information acquisition module, configured to acquire evaluation information of the garbage disposal point to be evaluated in various preset evaluation dimensions from a preset garbage disposal point database, including cleanliness, attitude of management personnel, garbage classification publicity, supervision, activity organization, signpost setting, hardware equipment, citizen participation rate, classification placement accuracy rate, and trash can setting; A feature vector construction module, configured to construct a feature vector of the garbage disposal point to be evaluated according to the water body change situation measurement value, the vegetation change situation measurement value, and the evaluation information; An evaluation result determination module, configured to process the feature vector by using a preset garbage disposal point evaluation model to obtain an evaluation result of the garbage disposal point to be evaluated.

9. A computer-readable storage medium storing computer-readable instructions, wherein, when the computer-readable instructions are executed by a processor, the steps of the garbage disposal point evaluation method according to any one of claims 1 to 7 are implemented.

10. A server, comprising a memory, a processor, and computer-readable instructions stored in the memory and executable on the processor, wherein, when the processor executes the computer-readable instructions, the steps of the garbage disposal point evaluation method according to any one of claims 1 to 7 are implemented.

Images