阅读说明：本技术 基于深度学习的垃圾分类投放智能监管系统及方法 (Garbage classification putting intelligent supervision system and method based on deep learning ) 是由 李伟 于 2021-06-21 设计创作，主要内容包括：本发明涉及一种基于深度学习的垃圾分类投放智能监管系统及方法,该垃圾分类投放智能监管系统,包括：用户登录模块、控制模块、图像采集模块、第一检测模块、远程数据处理模块和用户积分系统。本发明的基于深度学习的垃圾分类投放智能监管系统,能够解决传统人工值守下的垃圾分类方式,规避人工监管分类的诸多不足,设置的远程数据处理模块基于深度学习可以智能识别垃圾类别,实现了自动化垃圾分类监管,通过设置的用户积分系统,利用积分奖罚形式可以提高居民垃圾分类的自觉性。(The invention relates to a garbage classified putting intelligent supervision system and a method based on deep learning, wherein the garbage classified putting intelligent supervision system comprises: the system comprises a user login module, a control module, an image acquisition module, a first detection module, a remote data processing module and a user score system. The intelligent monitoring system for classified putting of garbage based on deep learning can solve the problems of a garbage classification mode under the traditional manual watch and avoid a plurality of defects of manual supervision and classification, the set remote data processing module can intelligently identify garbage types based on deep learning, automatic garbage classification supervision is realized, and the intuition of resident garbage classification can be improved by using an integral reward and penalty mode through the set user integral system.)

1. The utility model provides a garbage classification puts in intelligent supervisory systems based on deep learning which characterized in that includes: a user login module, a control module, an image acquisition module, a first detection module, a remote data processing module and a user score system, wherein,

the user login module is used for receiving user login information and matching the user login information with the user information in the user scoring system;

the first detection module is used for detecting the state of the throwing opening of the dustbin,

when the opening of a throwing opening of the dustbin is detected, a first control signal is sent to the control module, so that the control module controls the image acquisition module to acquire first image information in the dustbin before rubbish throwing according to the first control signal;

when the closing of the throwing-in opening of the dustbin is detected, sending a second control signal to the control module so that the control module controls the image acquisition module to acquire second image information in the dustbin after the rubbish is thrown in according to the second control signal;

the control module is further configured to transmit the first image information and the second image information to the remote data processing module, and the remote data processing module is configured to determine whether garbage classification is correct, obtain a determination result, and send the determination result to the user score system;

and the user point system is used for updating point information of the user according to the judgment result.

2. The intelligent deep learning-based spam delivery supervision system according to claim 1, wherein the user information comprises user registration information, a QR identification code associated with the user registration information, and point information.

3. The intelligent monitoring system for classified putting of garbage based on deep learning of claim 1, further comprising a second detection module, wherein the second detection module is used for detecting garbage filling quality information and garbage filling height information of a garbage can and sending the garbage filling quality information and the garbage filling height information to the control module;

and responding to the garbage filling quality information or the garbage filling height information to reach a preset threshold value, and the control module sends alarm information.

4. The intelligent monitoring system for classified putting of garbage based on deep learning of claim 3, further comprising a display module for displaying user information, alarm information and garbage classification results.

5. The intelligent deep learning-based garbage classified putting supervision system according to claim 1, wherein the remote data processing module comprises an image preprocessing unit, a garbage classification model and a judgment unit, wherein,

the image preprocessing unit is used for obtaining third image information according to the first image information and the second image information and judging whether garbage is thrown in according to the third image information;

when the garbage is judged to be thrown in, the garbage classification model classifies the third image information to obtain a garbage classification result, and the judging unit judges whether the garbage classification is correct or not according to the garbage classification result and preset garbage can information to obtain the judgment result.

6. The intelligent monitoring system for classified putting of garbage based on deep learning of claim 5, wherein the garbage classification model is obtained by training a neural network model, and the garbage classification model comprises an image feature extraction unit, an image feature attention unit and a probability calculation unit which are connected in sequence.

7. The intelligent monitoring system for classified putting of garbage based on deep learning of claim 1, wherein the image acquisition module comprises a camera and an annular light supplement lamp.

8. The intelligent deep learning-based garbage classified putting supervision system according to claim 3, wherein the second detection module comprises an ultrasonic ranging sensor and a quality sensor.

9. The intelligent monitoring method for classified garbage throwing based on deep learning is characterized by comprising the following steps:

matching the user login information with the user information in the user scoring system, and locking the user after the matching is successful;

controlling an image acquisition module to acquire first image information in the dustbin before garbage throwing according to the received first control signal;

controlling an image acquisition module to acquire second image information in the garbage bin after garbage is thrown according to the received second control signal;

transmitting the first image information and the second image information to a remote data processing module to judge whether the garbage classification is correct or not and obtain a judgment result;

updating the integral information of the user in response to the obtained judgment result;

the first control signal is a signal sent when the first detection module detects that the input port of the dustbin is opened, and the second control signal is a signal sent when the first detection module detects that the input port of the dustbin is closed.

10. The intelligent supervision method for garbage classified putting based on deep learning according to claim 9, characterized by further comprising:

acquiring the garbage filling quality information and the garbage filling height information of the garbage can, responding to the garbage filling quality information or the garbage filling height information reaching a preset threshold value, and sending alarm information.

Technical Field

The invention belongs to the technical field of garbage classification, and particularly relates to an intelligent monitoring system and method for garbage classification putting based on deep learning.

Background

The garbage classification is a series of activities for classifying, storing, throwing and transporting daily domestic garbage according to a certain regulation or standard so as to convert the daily domestic garbage into public resources. The classification aims to improve the resource value and the economic value of the garbage, make the best use of the garbage, reduce the garbage treatment amount and the use of treatment equipment, reduce the treatment cost, reduce the consumption of land resources, and have social, economic, ecological and other benefits. The garbage classification problem is solved, the utilization level of garbage resources can be improved, the garbage disposal amount can be reduced, and the method is an effective way for realizing garbage reduction, recycling and harmlessness and avoiding garbage enclosing.

However, since knowledge of garbage classification is incomplete and popular, a habit is not formed yet, people generally have a weak garbage classification consciousness, so that a garbage classification error phenomenon frequently occurs, and in order to reduce the garbage classification error phenomenon, most of people supervise and manage garbage classification by adding a manager, but the manager is difficult to supervise and manage everyone, and a large amount of manpower resources are wasted.

Disclosure of Invention

In order to solve the problems in the prior art, the invention provides an intelligent monitoring system and method for garbage classification putting based on deep learning. The technical problem to be solved by the invention is realized by the following technical scheme:

the invention provides a garbage classification putting intelligent supervision system based on deep learning, which comprises: a user login module, a control module, an image acquisition module, a first detection module, a remote data processing module and a user score system, wherein,

the user login module is used for receiving user login information and matching the user login information with the user information in the user scoring system;

the first detection module is used for detecting the state of the throwing opening of the dustbin,

when the opening of a throwing opening of the dustbin is detected, a first control signal is sent to the control module, so that the control module controls the image acquisition module to acquire first image information in the dustbin before rubbish throwing according to the first control signal;

when the closing of the throwing-in opening of the dustbin is detected, sending a second control signal to the control module so that the control module controls the image acquisition module to acquire second image information in the dustbin after the rubbish is thrown in according to the second control signal;

the control module is further configured to transmit the first image information and the second image information to the remote data processing module, and the remote data processing module is configured to determine whether garbage classification is correct, obtain a determination result, and send the determination result to the user score system;

and the user point system is used for updating point information of the user according to the judgment result.

In one embodiment of the present invention, the user information includes user registration information, a QR identification code associated with the user registration information, and point information.

In an embodiment of the present invention, the intelligent monitoring system for classified garbage dumping further includes a second detection module, and the second detection module is configured to detect garbage filling quality information and garbage filling height information of a garbage can, and send the garbage filling quality information and the garbage filling height information to the control module;

and responding to the garbage filling quality information or the garbage filling height information to reach a preset threshold value, and the control module sends alarm information.

In an embodiment of the present invention, the intelligent monitoring system for classified delivery of garbage further includes a display module, and the display module is configured to display user information, alarm information, and a garbage classification result.

In one embodiment of the invention, the remote data processing module comprises an image pre-processing unit, a garbage classification model and a determination unit, wherein,

the image preprocessing unit is used for obtaining third image information according to the first image information and the second image information and judging whether garbage is thrown in according to the third image information;

when the garbage is judged to be thrown in, the garbage classification model classifies the third image information to obtain a garbage classification result, and the judging unit judges whether the garbage classification is correct or not according to the garbage classification result and preset garbage can information to obtain the judgment result.

In an embodiment of the present invention, the garbage classification model is obtained by training a neural network model, and the garbage classification model includes an image feature extraction unit, an image feature attention unit, and a probability calculation unit, which are connected in sequence.

In one embodiment of the invention, the image acquisition module comprises a camera and an annular fill light.

In one embodiment of the invention, the second detection module comprises an ultrasonic ranging sensor and a quality sensor.

The invention provides a garbage classification putting intelligent supervision method based on deep learning, which comprises the following steps:

matching the user login information with the user information in the user scoring system, and locking the user after the matching is successful;

controlling an image acquisition module to acquire first image information in the dustbin before garbage throwing according to the received first control signal;

controlling an image acquisition module to acquire second image information in the garbage bin after garbage is thrown according to the received second control signal;

transmitting the first image information and the second image information to a remote data processing module to judge whether the garbage classification is correct or not and obtain a judgment result;

updating the integral information of the user in response to the obtained judgment result;

the first control signal is a signal sent when the first detection module detects that the input port of the dustbin is opened, and the second control signal is a signal sent when the first detection module detects that the input port of the dustbin is closed.

In one embodiment of the invention, the method further comprises:

acquiring the garbage filling quality information and the garbage filling height information of the garbage can, responding to the garbage filling quality information or the garbage filling height information reaching a preset threshold value, and sending alarm information.

Compared with the prior art, the invention has the beneficial effects that:

1. the intelligent monitoring system for classified putting of garbage based on deep learning can solve the problem of a garbage classification mode under the traditional manual watch and avoid many defects of manual supervision and classification, and the set remote data processing module can intelligently identify garbage categories based on deep learning, so that automatic garbage classification supervision is realized.

2. The intelligent monitoring system for garbage classification putting based on deep learning can improve the self-perception of resident garbage classification by using an integral rewarding and punishing form through the arranged user integral system.

The foregoing description is only an overview of the technical solutions of the present invention, and in order to make the technical means of the present invention more clearly understood, the present invention may be implemented in accordance with the content of the description, and in order to make the above and other objects, features, and advantages of the present invention more clearly understood, the following preferred embodiments are described in detail with reference to the accompanying drawings.

Drawings

Fig. 1 is a block diagram of a garbage classification putting intelligent supervision system based on deep learning according to an embodiment of the present invention;

fig. 2 is a block diagram of another intelligent monitoring system for garbage classification and delivery based on deep learning according to an embodiment of the present invention;

fig. 3 is a schematic structural diagram of another intelligent monitoring system for garbage classification and delivery based on deep learning according to an embodiment of the present invention;

FIG. 4 is a block diagram of a remote data processing module provided by an embodiment of the present invention;

FIG. 5 is a block diagram of a garbage classification model according to an embodiment of the present invention;

FIG. 6 is a schematic structural diagram of a garbage classification model according to an embodiment of the present invention;

fig. 7 is a flowchart of an intelligent monitoring method for garbage classification putting based on deep learning according to an embodiment of the present invention;

fig. 8 is a diagram of an evaluation index result of the garbage classification model according to the embodiment of the present invention.

Detailed Description

In order to further explain the technical means and effects of the present invention adopted to achieve the predetermined invention purpose, the following describes in detail a garbage classification and delivery intelligent supervision system and method based on deep learning according to the present invention with reference to the accompanying drawings and the detailed implementation.

The foregoing and other technical matters, features and effects of the present invention will be apparent from the following detailed description of the embodiments, which is to be read in connection with the accompanying drawings. The technical means and effects of the present invention adopted to achieve the predetermined purpose can be more deeply and specifically understood through the description of the specific embodiments, however, the attached drawings are provided for reference and description only and are not used for limiting the technical scheme of the present invention.

Example one

Referring to fig. 1, fig. 1 is a block diagram of a system for intelligently supervising garbage classification and delivery based on deep learning according to an embodiment of the present invention. As shown in the figure, the intelligent monitoring system for garbage classification putting based on deep learning of the embodiment includes: a user login module 1, a control module 2, an image acquisition module 3, a first detection module 4, a remote data processing module 5 and a user score system 6, wherein,

the user login module 1 is used for receiving user login information and matching the user login information with user information in the user score system 6;

a first detection module 4 for detecting the state of the throwing opening of the dustbin,

when the opening of a throwing opening of the dustbin is detected, a first control signal is sent to the control module 2, so that the control module 2 controls the image acquisition module 3 to acquire first image information in the dustbin before rubbish throwing according to the first control signal;

when the closing of the throwing-in opening of the dustbin is detected, a second control signal is sent to the control module 2, so that the control module 2 controls the image acquisition module 3 to acquire second image information in the dustbin after the rubbish is thrown in according to the second control signal;

the control module 2 is further configured to transmit the first image information and the second image information to the remote data processing module 5, and the remote data processing module 5 is configured to determine whether the garbage classification is correct, obtain a determination result, and send the determination result to the user score system 6;

and the user point system 6 is used for updating point information of the user according to the judgment result.

In the present embodiment, the user information includes user registration information, a QR identification code associated with the user registration information, and point information.

Optionally, the user login module 1 logs in by using a QR identification code, before the user puts in the garbage, the QR identification code is scanned to generate user login information, the user login information is matched with the user registration information in the user score system 6, and after the matching is successful, the user is locked.

Optionally, the control module 2 is composed of Smart200PLC and an electromagnetic switch.

Optionally, the image capturing module 3 includes a camera 301 and an annular light supplement lamp 302, in this embodiment, the camera 301 and the annular light supplement lamp 302 are installed above the dustbin body, so as to conveniently capture image information in the dustbin.

Optionally, the first detection module 4 may adopt a switch circuit module, when the input port of the dustbin is opened, the switch is turned off, an electric signal in the circuit changes, and the sensor sends a first control signal. When the throwing opening of the dustbin is detected to be closed, the switch is closed, the electric signal in the circuit is changed, and the sensor sends out a second control signal.

In other embodiments, the first detection module 4 may also adopt an optical sensor or an infrared sensor, and the state of the input port of the dustbin may be detected, and the specific structure is not limited herein.

Further, referring to fig. 4, fig. 4 is a structural block diagram of a remote data processing module according to an embodiment of the present invention, and as shown in the drawing, in the embodiment, the remote data processing module 5 includes an image preprocessing unit 501, a garbage classification model 502, and a determining unit 503.

The image preprocessing unit 501 is configured to obtain third image information according to the first image information and the second image information, and determine whether to put garbage according to the third image information. When the garbage is judged to be thrown in, the garbage classification model 502 performs classification processing on the third image information to obtain a garbage classification result, and the judgment unit 503 judges whether the garbage classification is correct or not according to the garbage classification result and preset garbage can information to obtain a judgment result.

Specifically, in this embodiment, the image preprocessing unit 501 subtracts the first image information and the second image information to detect image difference information of two images, and performs filtering and image enhancement processing on the subtracted image difference information to obtain third image information. And when the first image information and the second image information have no difference, judging that the garbage is thrown.

Further, the third image information is input into the garbage classification model 502 to obtain a garbage classification result, in this embodiment, the garbage classification result includes recyclable garbage, harmful garbage, kitchen garbage or other garbage. Specifically, please refer to fig. 5, wherein fig. 5 is a block diagram of a garbage classification model according to an embodiment of the present invention. As shown in the figure, the garbage classification model 502 is obtained by training a neural network model, and the garbage classification model 502 includes an image feature extraction unit 5021, an image feature attention unit 5022 and a probability calculation unit 5023, which are connected in sequence.

In this implementation, the image feature extraction unit 5021 is used for learning specific features of corresponding garbage and simultaneously extracting the features to the image feature attention unit 5022, the image feature attention unit 5022 is used for performing feature splicing on the extracted features and performing attention learning on the extracted features to ensure that a model pays more attention to the features of the corresponding garbage, and simultaneously eliminates image background noise interference, and the probability calculation unit 5023 performs probabilistic statistical calculation on the learned garbage attention features to finally obtain a garbage classification result and a probability value of accurate classification.

It should be noted that, in the model training process, a random gradient descent method is used as a network optimizer, a cross entropy function is used as a network loss function, and the model is iterated continuously until the model converges, so that a set of solidified neural network models with optimal weights is finally obtained. The specific training method is similar to the existing model training method, and is not described herein again.

Further, please refer to fig. 6, where fig. 6 is a schematic structural diagram of a garbage classification model according to an embodiment of the present invention. As shown in the figure, the garbage classification model of this embodiment is composed of 3 convolution layers, 3 pooling layers, 1 feature splicing layer, 4 intensive jump connection layers, 3 attention mechanism layers, 6 BN layers, and 1 softmax layer, where 1 attention mechanism layer is included between 2 intensive jump connection layers, and each convolution layer is followed by 1 pooling layer and 1 normalization layer.

The sizes of convolution kernels of the 3 independent convolution layers are 3x3, 5x5 and 7x7 respectively, and garbage with different sizes is subjected to targeted sign extraction according to the difference of the sizes of the convolution kernels. The feature splicing layer performs feature fusion on the outputs of the 3 independent convolution layers in a channel superposition mode. The pooling layer is mainly responsible for screening a large number of extracted characteristic parameters, obvious characteristic parameters are reserved, operation parameters are reduced, and operation speed is improved. The normalization layer is used for normalizing the characteristic parameters, so that the characteristic parameters have similar characteristic distribution, rapid convergence of the model is facilitated, and the generalization capability of the model is improved.

The dense jump connection layer is composed of 4 convolutional layers, and the convolutional layers are followed by 1 ReLU activation function and 1 normalization layer. And the output of each convolution layer after passing through the ReLU activation function is connected with the convolution layer of the subsequent jump connection layer through a direct connection channel, and the output of the two convolution layers is subjected to characteristic fusion by the characteristic connection layer. The deep features and the shallow features are efficiently fused through the dense jump connection layer, so that the features have the advanced information of the deep features and the image global property of the shallow features, and the gradient explosion and gradient dispersion of the model caused by the fact that the depth of the network is too deep can be avoided.

The attention mechanism layer extracts the space attention and the channel attention of the feature by using an attention algorithm, and adds the space attention and the channel attention to the feature map, so that the space attention and the channel attention are added to the feature map. The learning direction of the model can be guided through the added space and channel attention, and the special characteristics of the garbage image can be studied.

Further, the determination unit 503 is pre-stored with trash bin information, where the trash bin information is a type of the trash bin, and the type of the trash bin includes a recyclable trash bin, a harmful trash bin, a kitchen waste bin, and other trash bins. If the garbage classification result is consistent with the type of the garbage can, the throwing is correct, and if the garbage classification result is consistent with the type of the garbage can, the throwing is wrong.

Further, if the delivery is correct, the user is awarded with 200 points, the user point system 6 updates the point information of the user, and if the delivery is wrong, the user point system 6 deducts the 200 points of the user and updates the point information of the user.

It should be noted that the remote data processing module 5 and the user scoring system 6 are both disposed in the cloud server.

The intelligent monitoring system for classified garbage putting based on deep learning can solve the garbage classification mode under the traditional manual watch, avoid a plurality of defects of manual supervision and classification, and the set remote data processing module can intelligently identify garbage categories based on deep learning, so that automatic garbage classification supervision is realized. Through the user score system, the self-consciousness of resident garbage classification can be improved by adopting a score rewarding and punishing mode.

Example two

Further, please refer to fig. 2 and fig. 3 in combination, where fig. 2 is a block diagram of a structure of another intelligent monitoring system for garbage classification and delivery based on deep learning according to an embodiment of the present invention, and fig. 3 is a schematic structural diagram of another intelligent monitoring system for garbage classification and delivery based on deep learning according to an embodiment of the present invention. As shown in the figure, compared with the first embodiment, the intelligent monitoring system for classified putting of garbage based on deep learning of the present embodiment further includes a second detection module 7 and a display module 8, wherein,

the second detection module 7 is used for detecting the garbage filling quality information and the garbage filling height information of the garbage can and sending the garbage filling quality information and the garbage filling height information to the control module 2; and responding to the situation that the garbage filling quality information or the garbage filling height information reaches a preset threshold value, and sending alarm information by the control module 2. The display module 8 is used for displaying user information, alarm information and garbage classification results.

In this embodiment, the second detecting module 7 includes an ultrasonic ranging sensor 701 and a quality sensor 702, wherein the ultrasonic ranging sensor 701 is used for detecting the garbage filling height information of the garbage bin, and the quality sensor 702 is used for detecting the garbage filling quality information of the garbage bin. Alternatively, the ultrasonic ranging sensor 701 is an AJ-SR04M ultrasonic ranging sensor, and the mass sensor 702 is a HZC-H1 high-precision planar weight sensor.

In this embodiment, the ultrasonic distance measuring sensor 701 is installed above the dustbin body, and the quality sensor 702 is installed at the bottom of the dustbin body. When the detection value of the ultrasonic ranging sensor 701 reaches 10cm (+ -0.2 cm) or the detection value of the mass sensor 702 reaches 300kg (+ -0.5%), the control module 2 alarms and displays the alarm information on the display module 8.

It should be noted that the display module 8 may also display a QR identification code, which is convenient for the user to scan and log in.

It should be noted that, the intelligent monitoring system for classified garbage putting of the first embodiment and the second embodiment further includes a power module (not shown in the figure), and optionally, the power module mainly includes an air switch, a fuse, an ac contactor and a switching power supply, and is configured to provide dc power for the intelligent monitoring system for classified garbage putting.

Further, the working process of the intelligent monitoring system for garbage classified putting based on deep learning of the embodiment is described as follows:

the first step is as follows: initializing an intelligent garbage classification throwing monitoring system;

specifically, the power supply module is started and starts to supply power, then the control module performs self-checking operation, hardware equipment initialization is completed, the second detection module enters a resident working state and is used for measuring garbage filling quality information and garbage filling height information of the garbage can in real time, the cloud server is started and performs power-on self-checking, and preloading of a user integral system is completed.

When the detection value of the ultrasonic ranging sensor reaches 10cm (+ -0.2 cm) or the detection value of the weight sensor reaches 300kg (+ -0.5%), the control module sends alarm information.

The second step is that: a user logs in;

specifically, before the user puts garbage, the QR identification code on the display module is scanned, the cloud server receives the user login information system, confirms and locks the user in the user score system, and after the locking is completed, the cloud server is placed in a waiting state to prepare for receiving image data.

The third step: throwing garbage;

specifically, a feeding port of the dustbin is opened, when the opening of the feeding port of the dustbin is detected, a first detection module sends a first control signal to a control module, the control module triggers an image detection module, a camera and an annular light supplement lamp start to work, the camera shoots pictures in the dustbin, and first image information is collected;

after the garbage is put in, the putting opening of the garbage can is closed, when the putting opening of the garbage can is closed, a second control signal is sent to the control module, the control module triggers the image detection module again, the camera and the annular light supplement lamp start to work, the camera shoots the garbage in the put-in box, and second image information is collected.

And then the control module transmits the first image information and the second image information to the cloud server.

The fourth step: processing data;

specifically, after the cloud server receives the first image information and the second image information, the remote data processing module obtains third image information according to the first image information and the second image information to determine whether to throw garbage or not, if yes, the third image information is subjected to garbage classification processing, the type of the thrown garbage is accurately identified, whether to throw the garbage is determined, and a determination result is transmitted to the user score system.

The fifth step: the user scoring system updates the data.

And the user point system obtains a final judgment result, if the releasing is correct, the user point system gives a reward of 200 points to the user, and if the releasing is wrong, the user point system deducts 200 points until the reward returns to 0. And after the data is updated, the user score system returns the updated user information to the display module for display. And after the user confirms, the system quits and returns to the initialization state.

EXAMPLE III

The embodiment provides an intelligent monitoring method for garbage classification putting based on deep learning, please refer to fig. 7, where fig. 7 is a flowchart of the intelligent monitoring method for garbage classification putting based on deep learning according to the embodiment of the present invention, and as shown in the figure, the method of the embodiment includes:

matching the user login information with the user information in the user scoring system, and locking the user after the matching is successful;

controlling an image acquisition module to acquire first image information in the dustbin before garbage throwing according to a received first control signal, wherein the first control signal is a signal sent by a first detection module when detecting that a throwing opening of the dustbin is opened,

and controlling the image acquisition module to acquire second image information in the garbage can after garbage is thrown in according to the received second control signal, wherein the second control signal is a signal sent when the first detection module detects that the throwing port of the garbage can is closed.

Transmitting the first image information and the second image information to a remote data processing module to judge whether the garbage classification is correct or not and obtain a judgment result;

and updating the point information of the user in response to the obtained judgment result.

In one embodiment of the invention, the method further comprises: the garbage loading quality information and the garbage loading height information of the garbage can are obtained, and alarm information is sent out in response to the fact that the garbage loading quality information or the garbage loading height information reaches a preset threshold value.

Example four

The effectiveness of the garbage classification model of the intelligent monitoring system for garbage classification delivery in the first embodiment and the second embodiment is verified, in the present embodiment, the model is evaluated by the evaluation index accuracy (Acc), the accuracy (P), and the recall rate (R), and the calculation formula is as follows:

the accuracy represents the proportion of correctly classified samples in the classification result to the total samples, the recall rate represents the proportion of correctly classified samples in the positive examples in the total samples, and the accuracy represents the proportion of the positive examples in the samples predicted to be in the positive examples. And Acc, P and R are all detected by detecting TP in the sample confusion matrix: true positive, TN: true negative, FP: false positive, FN: and determining a false negative index.

The garbage classification model training is carried out aiming at four categories of recoverable garbage, harmful garbage, kitchen garbage and other garbage, 300 four categories of garbage images are used for testing to obtain the following confusion matrix,

the overall accuracy of the garbage classification model obtained by calculation is 96%, the accuracy is 95%, and the recall rate is 96%. Specific accuracy and recall of each category are shown in fig. 8, and fig. 8 is an evaluation index result diagram of the garbage classification model provided by the embodiment of the present invention.

In addition, the intelligent monitoring system for garbage classification putting is put in a residential district, 1800 images of garbage with multiple categories are measured, and the accuracy of the garbage classification model is about 90% through manual review and inspection.

It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that an article or device that comprises a list of elements does not include only those elements but may include other elements not expressly listed. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of additional like elements in the article or device comprising the element. The terms "connected" or "coupled" and the like are not restricted to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The directional or positional relationships indicated by "upper", "lower", "left", "right", etc., are based on the directional or positional relationships shown in the drawings, and are only for convenience of describing the present invention and simplifying the description, but do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as limiting the present invention.

The foregoing is a more detailed description of the invention in connection with specific preferred embodiments and it is not intended that the invention be limited to these specific details. For those skilled in the art to which the invention pertains, several simple deductions or substitutions can be made without departing from the spirit of the invention, and all shall be considered as belonging to the protection scope of the invention.