阅读说明：本技术 垃圾分类回收方法及系统 (Garbage classification recovery method and system ) 是由 陈志强 张丽 彭刚 孙运达 黄清萍 杨帆 于 2019-12-16 设计创作，主要内容包括：本公开披露了一种垃圾分类回收方法和系统。该方法包括：获取待分类垃圾的投放人的身份信息；对待分类的垃圾进行检测,得到待分类垃圾的检测数据；分析检测数据,得到垃圾查验结论：即确定待分类垃圾属于混装垃圾或者非混装垃圾中的一种；根据垃圾查验结论处理待分类的垃圾；以及绑定所述身份信息、所述检测数据以及所述垃圾查验结论。上述技术方案能够从垃圾出发追溯到投放人,尤其是在查验结果为混装垃圾的情况下,能够追溯到投放人,有利于落实垃圾分类措施。(The present disclosure discloses a garbage classification and recovery method and system. The method comprises the following steps: acquiring identity information of a person throwing garbage to be classified; detecting the garbage to be classified to obtain detection data of the garbage to be classified; analyzing the detection data to obtain a spam inspection conclusion: namely determining that the garbage to be classified belongs to one of mixed garbage or non-mixed garbage; processing the garbage to be classified according to a garbage checking conclusion; and binding the identity information, the detection data and the spam verification conclusion. According to the technical scheme, the throwing persons can be traced from the starting of the garbage, particularly, the throwing persons can be traced under the condition that the inspection result is the mixed garbage, and the garbage classification measures are favorably implemented.)

1. A garbage classification and recovery method is characterized by comprising the following steps:

acquiring identity information of a person throwing garbage to be classified;

detecting the garbage to be classified to obtain detection data of the garbage to be classified;

analyzing the detection data to obtain a rubbish inspection conclusion, namely determining that the rubbish to be classified belongs to one of mixed rubbish or non-mixed rubbish;

processing the garbage to be classified according to a garbage inspection conclusion; and

and binding the identity information, the detection data and the spam verification conclusion.

2. The garbage classification and recycling method according to claim 1, wherein the step of obtaining the identity information of the placer of the garbage to be classified comprises at least one of:

reading the identification information of the garbage thrower from the card by using a card reader;

reading a one-dimensional code, a two-dimensional code or a character string on the garbage bag to obtain identification information of a garbage thrower;

and identifying the biological characteristics of the garbage thrower to obtain the identification information of the thrower.

3. The garbage classification recycling method according to claim 1, characterized in that the detection data at least comprises a ray-scanned image of the garbage to be classified.

4. The garbage classification and recycling method according to claim 3, wherein the detection data further includes weight information of the garbage to be classified.

5. The garbage classification and recycling method according to claim 3, wherein the ray scan image comprises a gray scale transmission image, a pseudo color transmission image, a single view X-ray image, a double view X-ray image, an atomic number image, an electron density image, or a stereo image.

6. The garbage classification and recycling method according to claim 1, wherein the detection data is a ray scanned image of the garbage to be classified, and the ray scanned image of the garbage to be classified is analyzed by using a semantic segmentation model to obtain a garbage inspection conclusion.

7. The garbage classification recycling method according to claim 1, wherein the detection data is a ray-scanned image of the garbage to be classified, and the step of analyzing the detection data includes:

analyzing a ray scanning image of the garbage to be classified by utilizing a semantic segmentation model;

assigning a semantic label to each pixel or voxel on the ray scanning image, wherein the pixel or voxel belongs to one of mixed garbage or non-mixed garbage; and

and outputting the semantic tag graph.

8. The garbage classification and recycling method according to claim 7, wherein the semantic segmentation model is specifically a semantic segmentation convolutional neural network, wherein the semantic segmentation convolutional neural network is trained by the following steps:

performing pixel or voxel level labeling on the multiple garbage ray scanning images;

and inputting the marked ray scanning image into a semantic segmentation convolution neural network for training.

9. The garbage classification recycling method according to claim 7, wherein the step of analyzing the detection data further comprises:

and counting the semantic label graph, if the statistical value shows that only one kind of garbage or one kind of garbage accounts for more than a set proportion in the garbage to be classified, judging that the garbage to be classified belongs to non-mixed garbage, and if the statistical value shows that a plurality of kinds of garbage exist in the garbage to be classified and the proportions of various kinds of garbage in the garbage to be classified are less than the set proportion, judging that the garbage to be classified belongs to mixed garbage.

10. The garbage classification recycling method according to claim 9, further comprising the steps of: and for mixed garbage, sending an inspection conclusion and detection data to the garbage thrower in an electronic communication mode.

11. The garbage classification recycling method according to claim 9, further comprising the steps of: and further outputting the specific composition proportion and position distribution of various wastes in the wastes to be classified under the condition that the wastes to be classified belong to the mixed wastes in the waste inspection conclusion.

12. The method for sorting and recycling garbage according to claim 1, wherein the non-mixed garbage is one of the following:

can recover garbage, harmful garbage, kitchen garbage and other garbage.

13. A waste sorting and recycling system, the waste sorting and recycling system comprising:

the identity information acquisition device is used for acquiring the identity information of a person who throws the garbage to be classified;

the garbage detection device is used for detecting garbage to be classified to obtain detection data of the garbage to be classified;

the garbage analysis device analyzes the detection data to obtain a garbage inspection conclusion, namely determining that the garbage to be classified belongs to one of mixed garbage or non-mixed garbage;

the garbage storage device processes the garbage to be classified according to the garbage inspection conclusion; and

the spam analysis device binds the identity information, the detection data and the spam inspection conclusion.

14. The garbage classification recycling system of claim 13, wherein the identity information obtaining means comprises at least one of:

the card reader reads the identification information of the garbage thrower from the card;

the code scanning or photographing device reads the one-dimensional code, the two-dimensional code or the character string on the garbage bag to obtain the identification information of the garbage thrower;

and identifying the biological characteristics of the garbage thrower to obtain the identification information of the thrower.

15. The garbage classification recycling system of claim 13, wherein the detection data includes at least a ray-scan image of the garbage to be classified.

16. The garbage collection system of claim 15, wherein the radiographic image comprises a grayscale transmission image, a pseudo-color transmission image, a single-view X-ray image, a dual-view X-ray image, an atomic number image, an electron density image, or a stereo image.

17. The garbage classification recycling system according to claim 13, wherein the detection data is a ray-scanned image of the garbage to be classified, and the garbage analysis device is specifically configured to: and analyzing the ray scanning image of the garbage to be classified by utilizing a semantic segmentation model to obtain a garbage inspection conclusion.

18. The garbage classification recycling system according to claim 13, wherein the detection data is a ray-scanned image of the garbage to be classified, and the garbage analysis device is specifically configured to:

analyzing a ray scanning image of the garbage to be classified by utilizing a semantic segmentation model;

assigning a semantic label to each pixel or voxel on the ray scanning image, namely determining whether the pixel or voxel belongs to mixed garbage or non-mixed garbage; and

and outputting the semantic tag graph.

19. The garbage classification and recycling system of claim 18, wherein the semantic segmentation model is specifically a semantic segmentation convolutional neural network, wherein the semantic segmentation convolutional neural network is trained by:

performing pixel or voxel level labeling on the multiple garbage ray scanning images;

and inputting the marked ray scanning image into a semantic segmentation convolution neural network for training.

20. The garbage classification recycling system of claim 18, wherein the garbage analysis device is further configured to:

and counting the semantic label graph, if the statistical value shows that only one kind of garbage or the total share of one kind of garbage in the garbage to be classified is more than a set proportion, judging that the garbage to be classified belongs to non-mixed garbage, and if the statistical value shows that a plurality of kinds of garbage exist in the garbage to be classified and the total shares of various kinds of garbage are less than the set proportion, judging that the garbage to be classified belongs to mixed garbage.

21. The garbage classification recycling system of claim 20, wherein the garbage analysis device is further configured to: and further outputting the specific composition proportion and position distribution of various wastes in the wastes to be classified under the condition that the wastes to be classified belong to the mixed wastes in the waste inspection conclusion.

Technical Field

The embodiment of the disclosure relates to a garbage classification recycling technology, in particular to a garbage classification recycling method and system.

Background

China is the first country of refuse manufacture in the world, the annual refuse yield is 4 hundred million tons, the year is 8 percent ~ 10 percent in all years, the environment is overwhelmed, the refuse classification is urgent, 6 months in 2019, nine departments such as a housing department, a development and improvement commission, an ecological environment department and the like jointly transmit' notice that departments such as a housing department, a town and country construction department and the like comprehensively start the refuse classification work in the country and above cities from the year 2019 to the year 2020, 46 key cities which are tried in advance are to be basically built into a refuse classification treatment system, other land-level cities realize the full coverage of the refuse classification of public institutions, at least 1 region of each land-level city realizes the full coverage from the year 2022, and the land-level cities and above cities are to be basically built into the refuse classification treatment system before 2025 years.

In the city of the prior trial garbage classification, the problem is that the classified delivery behavior of residents is not standard and the mixed loading is serious. On one hand, a good garbage classification delivery habit needs to be developed for a long time, on the other hand, an effective supervision means for classification delivery is also lacked at the present stage, and the garbage classification delivery habit is simply based on a small proportion of manual sampling inspection. Therefore, technical means are urgently needed to fill the monitoring blank at present.

In addition, the establishment of the electronic standing book for garbage delivery registration management by taking a family as a unit is expected to be realized by all levels of government regulatory levels. Information such as the weight of each type of garbage thrown at each time by each family and the standard degree of classified throwing can be recorded in the standing book, so that reward and punishment measures for garbage classification behaviors are more rational and powerful, and official statistics and policy and regulations can have credible big data as a basis. However, the prior art does not address this aspect of the technology.

Disclosure of Invention

Aiming at one or more problems in the prior art, a garbage classification recycling method and a garbage classification recycling system are provided, and garbage throwers can be traced.

According to one aspect of the present disclosure, a garbage classification and recycling method is provided, which includes the steps of: acquiring identity information of a person throwing garbage to be classified; detecting the garbage to be classified to obtain detection data of the garbage to be classified; analyzing the detection data to obtain a rubbish inspection conclusion, namely determining that the rubbish to be classified belongs to one of mixed rubbish or non-mixed rubbish; processing the garbage to be classified according to a garbage inspection conclusion; and binding the identity information, the detection data and the spam verification conclusion.

According to the embodiment of the disclosure, the step of acquiring the identity information of the thrower of the garbage to be classified comprises at least one of the following steps: reading the identification information of the garbage thrower from the card by using a card reader; reading a one-dimensional code, a two-dimensional code or a character string on the garbage bag to obtain identification information of a garbage thrower; and identifying the biological characteristics of the garbage thrower to obtain the identification information of the thrower.

According to an embodiment of the present disclosure, the detection data includes at least a ray-scanned image of the trash to be classified.

According to an embodiment of the present disclosure, the detection data further includes weight information of the garbage to be classified.

According to an embodiment of the present disclosure, the radiographic image includes a grayscale transmission image, a pseudo-color transmission image, a single-view X-ray image, a dual-view X-ray image, an atomic number image, an electron density image, or a stereoscopic image.

According to the embodiment of the disclosure, the detection data is the ray scanning image of the garbage to be classified, and the ray scanning image of the garbage to be classified is analyzed by utilizing a semantic segmentation model to obtain a garbage inspection conclusion.

According to an embodiment of the present disclosure, the detection data is a ray-scanned image of the garbage to be classified, and the step of analyzing the detection data includes: analyzing a ray scanning image of the garbage to be classified by utilizing a semantic segmentation model; assigning a semantic label to each pixel or voxel on the ray scanning image, namely which type of mixed garbage and non-mixed garbage the pixel or voxel belongs to; and outputting the semantic tag graph.

According to an embodiment of the present disclosure, the semantic segmentation model is specifically a semantic segmentation convolutional neural network, wherein the semantic segmentation convolutional neural network is trained by the following steps: performing pixel or voxel level labeling on the multiple garbage ray scanning images; and inputting the marked ray scanning image into a semantic segmentation convolution neural network for training.

According to an embodiment of the present disclosure, the step of analyzing the detection data further comprises: and counting the semantic tag graph, if the statistical value shows that only one kind of garbage or the share of one kind of garbage in the garbage to be classified accounts for more than a set proportion, judging that the garbage to be classified belongs to non-mixed garbage, and if the statistical value shows that a plurality of kinds of garbage exist in the garbage to be classified and the shares of various kinds of garbage in the garbage to be classified account for less than the set proportion, judging that the garbage to be classified belongs to mixed garbage.

According to the embodiment of the present disclosure, the garbage classification and recycling method further includes the steps of: and for mixed garbage, sending an inspection conclusion and detection data to the garbage thrower in an electronic communication mode.

According to an embodiment of the present disclosure, the garbage classification recycling method further includes the steps of: and further outputting the specific composition proportion and position distribution of various wastes in the wastes to be classified under the condition that the wastes to be classified belong to the mixed wastes in the waste inspection conclusion.

According to an embodiment of the present disclosure, the non-mixed garbage is specifically one of the following: can recover garbage, harmful garbage, kitchen garbage and other garbage.

According to another aspect of the present disclosure, a garbage classification and recycling system is provided, including: the identity information acquisition device is used for acquiring the identity information of a person who throws the garbage to be classified; the garbage detection device is used for detecting garbage to be classified to obtain detection data of the garbage to be classified; the garbage analysis device analyzes the detection data to obtain a garbage inspection conclusion, namely determining that the garbage to be classified belongs to one of mixed garbage or non-mixed garbage; the garbage storage device processes the garbage to be classified according to the garbage inspection conclusion; and the garbage analysis device binds the identity information, the detection data and the garbage checking conclusion.

According to an embodiment of the present disclosure, the identity information acquiring apparatus includes at least one of: the card reader reads the identification information of the garbage thrower from the card; and the code scanning or photographing device reads the one-dimensional code, the two-dimensional code or the character string on the garbage bag to obtain the identification information of the garbage throwing person, or identifies the biological characteristics of the garbage throwing person to obtain the identification information of the throwing person.

According to an embodiment of the present disclosure, the detection data includes at least a ray-scanned image of the trash to be classified.

According to an embodiment of the present disclosure, the radiographic image includes a grayscale transmission image, a pseudo-color transmission image, a single-view X-ray image, a dual-view X-ray image, an atomic number image, an electron density image, or a stereoscopic image.

According to an embodiment of the present disclosure, the detection data is a ray-scanned image of the garbage to be classified, and the garbage analysis device is specifically configured to: and analyzing the ray scanning image of the garbage to be classified by utilizing a semantic segmentation model to obtain a garbage inspection conclusion.

According to an embodiment of the present disclosure, the detection data is a ray-scanned image of the garbage to be classified, and the garbage analysis device is configured to: analyzing a ray scanning image of the garbage to be classified by utilizing a semantic segmentation model; assigning a semantic label to each pixel or voxel on the ray scanning image, wherein the pixel or voxel belongs to one of mixed garbage or non-mixed garbage; and outputting the semantic tag graph.

According to an embodiment of the present disclosure, the semantic segmentation model is specifically a semantic segmentation convolutional neural network, wherein the semantic segmentation convolutional neural network is trained by the following steps: performing pixel or voxel level labeling on the multiple garbage ray scanning images; and inputting the marked ray scanning image into a semantic segmentation convolution neural network for training.

According to an embodiment of the present disclosure, the garbage analysis apparatus is further configured to: and counting the semantic tag graph, if the statistical value shows that only one kind of garbage or the share of one kind of garbage in the garbage to be classified accounts for more than a set proportion, judging that the garbage to be classified belongs to non-mixed garbage, and if the statistical value shows that a plurality of kinds of garbage exist in the garbage to be classified and the shares of various kinds of garbage in the garbage to be classified account for less than the set proportion, judging that the garbage to be classified belongs to mixed garbage.

According to an embodiment of the present disclosure, the garbage analysis apparatus is further configured to: and further outputting the specific composition proportion and position distribution of various wastes in the wastes to be classified under the condition that the wastes to be classified belong to the mixed wastes in the waste inspection conclusion.

The scheme of the embodiment overcomes the defect that the garbage throwing cannot be supervised in the prior art. By utilizing the technical scheme, the garbage thrower can be traced. In other embodiments, the semantic segmentation model is utilized to accurately classify the garbage to be classified into mixed garbage and non-mixed garbage and clearly indicate the specific situation of mixed garbage.

Drawings

For a better understanding of the present disclosure, reference will be made to the following detailed description taken in conjunction with the accompanying drawings in which:

fig. 1 shows a schematic structural diagram of a garbage classification recycling system according to an embodiment of the present disclosure;

FIG. 2 is a schematic structural diagram of a garbage detection apparatus according to an embodiment of the present disclosure;

FIG. 3 is a schematic view of a waste analysis device according to an embodiment of the present disclosure;

FIG. 4 shows a schematic flow diagram of a garbage classification recycling method according to an embodiment of the present disclosure; and

FIG. 5 shows a schematic flow diagram of analyzing a scanned image of trash to be classified using a semantic segmentation model in accordance with an embodiment of the present disclosure.

Detailed Description

Specific embodiments of the present disclosure will be described in detail below, with the understanding that the embodiments described herein are illustrative only and are not intended to limit the present disclosure. In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present disclosure. However, it will be apparent to one of ordinary skill in the art that: these specific details need not be employed to practice the present disclosure. In other instances, well-known structures, materials, or methods have not been described in detail in order to avoid obscuring the present disclosure.

Throughout the specification, reference to "one embodiment," "an embodiment," "one example," or "an example" means: the particular features, structures, or characteristics described in connection with the embodiment or example are included in at least one embodiment of the present disclosure. Thus, the appearances of the phrases "in one embodiment," "in an embodiment," "one example" or "an example" in various places throughout this specification are not necessarily all referring to the same embodiment or example. Furthermore, the particular features, structures, or characteristics may be combined in any suitable combination and/or sub-combination in one or more embodiments or examples. Further, as used herein, the term "and/or" will be understood by those of ordinary skill in the art to include any and all combinations of one or more of the associated listed items.

The embodiment of the disclosure provides a garbage classification and recovery method and a garbage classification and recovery system, which can purposefully solve the above defects of the prior art. According to the embodiment of the disclosure, the identity information of the throwing person of the garbage to be classified is obtained through card swiping or identity recognition technology, then the detection data of the garbage to be classified is obtained, and a garbage inspection conclusion is further obtained, for example, the detection data is analyzed, the classification of the garbage is judged, the garbage is classified into one of various non-mixed garbage (such as recoverable garbage, harmful garbage, kitchen garbage, other garbage and the like) or mixed garbage, and the mixed garbage is analyzed to obtain the specific situation of mixed loading. And binding and storing the 'person identity information of the throwing person, the garbage detection data and the garbage inspection conclusion'. According to the scheme, people who throw garbage can be traced.

According to the embodiment, the non-mixed garbage is respectively conveyed to the corresponding garbage storage devices through the conveying devices. And returning the throwing port and feeding the concrete situation of the mixed loading back to a throwing person or feeding the mixed loading into a mixed loading garbage storage device for disposal by workers of a garbage disposal station according to the currently selected working mode.

Fig. 1 shows a schematic structural diagram of a garbage classification recycling system according to an embodiment of the present disclosure. As shown in fig. 1, the garbage sorting and recycling system according to the embodiment of the present disclosure includes a conveying device 110 for conveying garbage to be sorted, a card reader 120 (optional), a code scanning or photographing device 115 (optional), a garbage detecting device 125, a garbage analyzing device 130, a recyclable garbage storage device 135, a harmful garbage storage device 140, a kitchen garbage storage device 145, an other garbage storage device 150, and a mixed garbage storage device 155. In addition, the "person identity information of the throwing person-spam detection data-spam inspection conclusion" obtained by the spam classification and recovery system is bound and stored in the spam analysis device or the remote central storage 190.

For example, the card reader 120 is used to obtain the player identity information, which may be player information stored on a magnetic card. Alternatively, the barcode may be a one-dimensional barcode on the trash bag, a two-dimensional barcode on the trash bag, or a character string on the trash bag. The information on the identity of the person being delivered is obtained by a code scanning or photographing device 115. The player identity information may be, for example, a name/code (id card)/room number/telephone number, etc. that identifies the player.

For this embodiment, obtaining the identity information of the throwing person who throws the garbage to be classified is realized by requiring the throwing person to swipe a card before throwing the garbage, the card reader 120 is installed near the throwing port of the garbage classification recycling system, and the garbage thrown in after swiping the card is bound with the holder of the magnetic card by default. And (4) canceling binding by swiping the card again or canceling binding without garbage throwing after a certain time length, wherein the binding time length can be reasonably configured as required.

For another embodiment, acquiring the identity information of the garbage thrower to be classified is achieved by a code scanning or photographing device 115, which has an optical character recognition function to scan codes or photograph garbage bags. The code scanning or photographing device 115 may be installed before the garbage detection device 125 or may be installed after the garbage detection device 125. In this embodiment, the player places the customized garbage bag with the player identity information toward the code scanning or photographing device 115, for example, when the code scanning or photographing device 115 is installed above the conveying device 110, the garbage bag with the player identity information should be placed on the conveying device upward, otherwise, once the code scanning or photographing device 115 cannot acquire the player identity information, the garbage can be automatically returned to the player mouth and the player is prompted to place the garbage correctly.

The trash detection device 125 obtains detection data of the trash to be classified, which at least includes a scanned image of the trash to be classified. The detection data of the garbage to be classified can also contain weight information of the garbage to be classified.

Fig. 2 shows a schematic view of a debris detecting device. The apparatus shown in fig. 2 includes a conveyor 110, such as a belt or the like, for carrying the refuse to be sorted, an X-ray source 210, a detector 220, an acquisition circuit 225, a controller 230, a data processing computer 240 (which may be a refuse analyzer), and the like. The radiation source 210 includes one or more X-ray generators and may perform a single energy scan or a dual energy scan.

As shown in fig. 2, the conveyor 110 carries the waste to be sorted through a scanning region between the source 210 and the detector 220. In some embodiments, the detector 220 and the acquisition circuit 225 are, for example, a detector and data acquisition device with an integral module structure, such as a multi-row detector, for detecting the rays transmitted through the inspected object, obtaining an analog signal, and converting the analog signal into a digital signal, thereby outputting the projection data of the garbage to be classified for the X-rays. For example, one row of detectors is provided for high-energy radiation and another row of detectors is provided for low-energy radiation, or the high-energy radiation and the low-energy radiation are used in a time-shared manner. The controller 230 is used to control the synchronous operation of the various parts of the overall system. The data processing computer 240 is used to process the data collected by the data collector, analyze the data using a semantic segmentation model such as a trained artificial neural network, and output a spam inspection result.

According to this embodiment, the detector 220 and acquisition circuitry 225 are used to acquire transmission data of the waste to be sorted. The acquisition circuit 225 includes a data amplification and shaping circuit that can operate in either a (current) integration mode or a pulse (count) mode. The cable of the acquisition circuit 225 is connected to the controller 230 and the data processing computer 240, and the acquired data is stored in the data processing computer 240 according to the trigger command.

In some embodiments, the detector 220 includes a plurality of detection units that receive X-rays that have penetrated the object under examination. The data acquisition circuitry 225 is coupled to the detector 220 to convert signals generated by the detector 220 into detection data. The controller 230 is coupled to the source 210 via control lines CTRL1, the detector 220 via control lines CTRL2, and the data acquisition circuitry, and controls one or more X-ray generators in the source 210 to generate single-energy or dual-energy X-rays to emit radiation that penetrates the object under examination as the object under examination moves. In addition, the controller 230 controls the detector 220 and the data acquisition circuit 225 to obtain detection data corresponding to the X-ray generator at a single energy or at least two energies. The data processing computer 240 reconstructs an image of the garbage to be classified based on the detection data, analyzes the data (image) using a semantic segmentation model such as a trained artificial neural network, and outputs a garbage inspection conclusion.

Fig. 3 shows a block diagram of the configuration of a garbage analysis apparatus (data processing computer). As shown in fig. 3, the data acquired by the acquisition circuit 225 is stored in the memory 310 through the interface unit 340 and the bus 345. A Read Only Memory (ROM) 315 stores configuration information and programs. The Random Access Memory (RAM) 320 is used for temporarily storing various data during the operation of the processor 330. In addition, the memory 310 also stores a computer program for performing data processing. The internal bus 345 connects the memory 310, the read only memory 315, the random access memory 320, the input device 325, the processor 330, the display device 335, and the interface unit 340.

According to one embodiment, after an operation command is input by a user through an input device 325 such as a keyboard and a mouse, an instruction code of the computer program instructs the processor 330 to execute a predetermined data reconstruction algorithm, and after a data processing result is obtained, it is displayed on a display device 335 such as an LCD display or detected image data is directly output in the form of a hard copy such as printing, and furthermore, the data is analyzed using a semantic segmentation model such as a trained artificial neural network, and a spam conclusion is output. In other embodiments, the computer program may be automatically executed without the user inputting an operation instruction.

For example, the radiation source 210 may be a radioisotope (e.g., cobalt-60), a low energy X-ray machine, a high energy X-ray accelerator, or the like.

For example, the detector 220 may be divided into material, gas, scintillator, or solid detectors, or array, single, double, or multiple rows, single-layer, or double-layer high-low energy detectors.

Preferably, the scanned image of the garbage to be classified can be a gray or pseudo-color plane image, and the corresponding garbage detection device is a single-view X-ray detection device, wherein the color comprises the material information of the garbage; the scanned images of the garbage to be classified can also be two or more gray or pseudo-color plane images, the corresponding garbage detection devices are X-ray detection devices with double visual angles or more visual angles, and the garbage penetrates from more directions to acquire information, so that the accuracy of garbage classification judgment is improved; the scanned image of the garbage to be classified can also be a three-dimensional image, and the corresponding garbage detection device is an X-ray detection device based on the CT technology, so that the information is further enriched to improve the accuracy of garbage classification judgment. In other embodiments, the scanned image may be an atomic number image or an electron density image obtained based on a dual energy technique.

The weight information of the garbage to be classified is obtained, a weight measuring unit can be installed in the conveying device 110, a weight measuring unit can be installed in the garbage detection device, and weight estimation can also be carried out through a scanned image of the garbage to be classified.

The spam detection data is analyzed using the spam analysis device 130 or the data processing computer 240 to obtain spam ping conclusions. Such as applying techniques in the fields of computer vision and machine learning. Preferably, a deep learning technology is applied to design a special convolutional neural network suitable for the characteristics of the garbage scanning image. Chinese patent publication CN 109201514 a also proposes to process garbage scan images using deep learning neural networks to calculate the matching rate between the garbage to be classified and the recyclable garbage.

However, this technique can only obtain the similarity that the garbage to be classified belongs to certain recoverable garbage, and cannot accurately indicate the specific situation of garbage mixed loading. The semantic segmentation model used in the embodiment of the present disclosure, such as the semantic segmentation convolutional neural network, can analyze the scanned image of the garbage to be classified finely by learning a large number of scanned images of the garbage with pixel or voxel level labels, and assign a semantic label to each pixel or voxel on the scanned image, that is, to which garbage of various non-mixed garbage (such as recoverable garbage, harmful garbage, kitchen garbage, other garbage, etc.) and mixed garbage the pixel or voxel belongs, and output a semantic label map. The specific situation of garbage mixed loading can be given by carrying out filtering denoising, connected domain analysis and area/volume/weight estimation on the semantic label graph, wherein the specific situation comprises the types of garbage contained in the garbage bag, the statistical values of the area/volume/weight ratio of each type of garbage and the like, and the position distribution of each type of garbage in the garbage bag. If the statistical value shows that only one kind of garbage or the share of one kind of garbage in the garbage to be classified in the total is more than a set proportion (such as the area/volume/weight ratio exceeds a threshold), the garbage inspection concludes that the garbage to be classified belongs to the non-mixed garbage. If the statistical value shows that the garbage to be classified has various types of garbage and the total share of various types of garbage is smaller than the set proportion, the garbage inspection conclusion is that the garbage to be classified belongs to mixed garbage and the specific composition proportion and position distribution of the various types of garbage are indicated. The spam verification conclusion can also contain the identification number of the spam classification recycling system.

It can be understood by those skilled in the art that, in addition to the convolutional neural network, the garbage analysis apparatus in the embodiment of the present disclosure may also use a neural network structure such as a self-coding network, a deep belief network, and the like to perform deep learning, that is, any neural network structure capable of outputting a semantic tag map with the same resolution as that of the garbage scan image is included in the technical solution of the present disclosure.

For example, the garbage analysis device 130 may perform deep learning by using a deep neural network structure such as a stacked self-coding neural network (SAE) or a deep belief neural network (DBN), may also use a classical machine learning method based on a probabilistic graph model such as a Conditional Random Field (CRF) or a Markov Random Field (MRF), or may also use a method in which a deep neural network is combined with a probabilistic graph model; any semantic segmentation mathematical model that can output a semantic label map with the same resolution as the spam scan image can be included in the solution of the present disclosure.

After the verification is obtained, the garbage analysis device 130 sends a control signal CTRL3 to the conveying device 110 to convey the garbage to be classified to the corresponding garbage storage devices, such as the recyclable garbage storage device 135, the hazardous garbage storage device 140, the kitchen garbage storage device 145, the other garbage storage devices 150, and the mixed garbage storage device 155. In other embodiments, the information may be sent to the personnel in the garbage station, and the personnel transfers the garbage to be classified into the corresponding garbage storage device.

The garbage analysis device 130 binds the above-mentioned deposited person identity information, garbage detection data, and garbage check conclusion together to form a complete record of "deposited person identity information-garbage detection data-garbage check conclusion", and this record can be uploaded to the database of the central storage 190 for storage in a wired or wireless manner, or can be stored locally in the garbage classification recycling system, for example, in the memory of the garbage analysis device 130.

In addition, the camera, the temperature sensor, the humidity sensor, the combustible gas sensor, the weight measuring unit and the volume measuring unit can be used for monitoring the conditions in various garbage storage devices in real time, so that the generation of putrefactive peculiar smell and toxic and harmful gases is prevented, and the fireproof, explosion-proof and flame-retardant capabilities are achieved. The garbage storage device can be additionally provided with a virus killing and sterilizing device which can be an ultraviolet lamp tube, a high-temperature sterilizing device, a chemical sterilizing device and a radiation sterilizing device. The garbage storage device can be internally and additionally provided with an air filtering device, can be used for electromagnetic adsorption filtration, can also be used for filtering a filter screen, and can also be used for filtering activated carbon, so that the diffusion of harmful gases is prevented on the premise of ensuring air circulation.

In order to quickly dispose of unqualified and classified mixed garbage bags, workers in the garbage disposal station can obtain detection data and corresponding inspection conclusions of the mixed garbage bags again, and if the garbage bags have identity information of a throwing person, the personnel can scan codes or take pictures to obtain identity information of the throwing person, and then send query requests to a central memory database or a corresponding garbage classification recycling system to call the previous detection data and the corresponding inspection conclusions; otherwise, the garbage classification and recovery system can detect the garbage again. For the situation that the mixed garbage bag is disposed by the garbage disposal station staff, the specific situation of the mixed garbage bag can still be notified to the throwing person through APP or short messages and the like, the garbage disposal station staff can also choose to take pictures or record videos of the disposal process while disposing, and the pictures or videos can be uploaded, stored or sent to the throwing person for education reminding.

Fig. 4 shows a schematic flow chart of a garbage classification recycling method according to an embodiment of the present disclosure. As shown in fig. 4, in step S410, identity information of the shooter who is to sort the garbage is obtained, for example, identity information of the shooter is obtained through the card reader 120 or the code scanning or photographing device 115.

In step S415, the garbage to be classified is detected to obtain the detection data of the garbage to be classified, for example, the garbage to be classified is detected by the detection device shown in fig. 2 to obtain the image data of the garbage to be classified. According to other embodiments, the detection data may also contain information on the weight of the waste to be sorted.

In step S420, the detection data is analyzed to obtain a garbage checking result, i.e. it is determined that the garbage to be classified belongs to one of mixed garbage or non-mixed garbage, for example, one of mixed garbage or non-mixed garbage (recoverable garbage, harmful garbage, kitchen garbage, other garbage, etc.).

At step S425, the garbage to be classified is processed according to the garbage checking result, for example, the garbage to be classified is transferred to the corresponding garbage storage device according to the checking result. At step S430, the identity information, the detection data and the spam verification result are bound.

FIG. 5 shows a schematic flow diagram of analyzing a scanned image of trash to be classified using a semantic segmentation model in accordance with an embodiment of the present disclosure.

The disclosed embodiments use semantic segmentation models, such as a semantic segmentation convolutional neural network, to scan images by learning a large number of garbage rays with pixel or voxel level labels. In step S510, a ray-scanned image of the garbage to be classified is analyzed by using a semantic segmentation model, and in step S515, a semantic label is assigned to each pixel or voxel on the ray-scanned image, that is, the pixel or voxel belongs to one of various non-mixed garbage (such as recoverable garbage, harmful garbage, kitchen garbage, other garbage, etc.) and mixed garbage, and in step S520, a semantic label map is output.

According to some embodiments, the semantic tag map can be subjected to filtering and denoising, connected domain analysis and area/volume/weight estimation, so that specific situations of garbage mixed loading can be given, including statistics of which types of garbage are contained in the garbage bag, area/volume/weight ratios of each type of garbage, and the position distribution of each type of garbage in the garbage bag.

In step S525, if the statistical value indicates that only one kind of garbage or a certain kind of garbage occupies an absolute main body (for example, the area/volume/weight ratio exceeds a threshold), the garbage inspection concludes that the garbage to be classified belongs to the kind of non-mixed garbage. If the statistical value shows that the garbage to be classified contains various types of garbage and no garbage accounts for the absolute main body, the garbage inspection conclusion is that the garbage to be classified belongs to mixed garbage. In step S530, as another embodiment, in the case that the spam verification result indicates that the spam to be classified belongs to the mixed spam, the specific composition ratios and location distributions of various types of spam can be further indicated. The spam verification conclusion can also contain the identification number of the spam classification recycling system. For example, in the case of mixing recyclable waste with kitchen waste, specific positions of the recyclable waste and the kitchen waste in the waste (waste bag) to be sorted are given, such as the upper part, the middle part or the lower part of the waste bag. Therefore, the garbage station is convenient for workers to handle and can also conveniently indicate the mixed loading condition of the garbage thrown by the throwing personnel.

The foregoing detailed description has set forth numerous embodiments of garbage classification based recycling methods and systems using schematics, flowcharts, and/or examples. Where such diagrams, flowcharts, and/or examples contain one or more functions and/or operations, it will be understood by those within the art that each function and/or operation within such diagrams, flowcharts, or examples can be implemented, individually and/or collectively, by a wide range of structures, hardware, software, firmware, or virtually any combination thereof. In one embodiment, portions of the subject matter described in embodiments of the present disclosure may be implemented by Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs), Digital Signal Processors (DSPs), or other integrated formats. However, those skilled in the art will recognize that some aspects of the embodiments disclosed herein, in whole or in part, can be equivalently implemented in integrated circuits, as one or more computer programs running on one or more computers (e.g., as one or more programs running on one or more computer systems), as one or more programs running on one or more processors (e.g., as one or more programs running on one or more microprocessors), as firmware, or as virtually any combination thereof, and that designing the circuitry and/or writing the code for the software and or firmware would be well within the skill of one of skill in the art in light of this disclosure. In addition, those skilled in the art will appreciate that the mechanisms of the subject matter described herein are capable of being distributed as a program product in a variety of forms, and that an illustrative embodiment of the subject matter described herein applies regardless of the particular type of signal bearing media used to actually carry out the distribution. Examples of signal bearing media include, but are not limited to: recordable type media such as floppy disks, hard disk drives, Compact Disks (CDs), Digital Versatile Disks (DVDs), digital tape, computer memory, etc.; and a transmission type medium such as a digital and/or an analog communication medium (e.g., a fiber optic cable, a waveguide, a wired communications link, a wireless communication link, etc.).

While the present disclosure has been described with reference to several exemplary embodiments, it is understood that the terminology used is intended to be in the nature of words of description and illustration, rather than of limitation. As the present disclosure may be embodied in several forms without departing from the spirit or essential characteristics thereof, it should also be understood that the above-described embodiments are not limited by any of the details of the foregoing description, but rather should be construed broadly within its spirit and scope as defined in the appended claims, and therefore all changes and modifications that fall within the meets and bounds of the claims, or equivalences of such meets and bounds are therefore intended to be embraced by the appended claims.