KR20230060605A - Waste information analysis system and method - Google Patents

Abstract

A waste information analysis system according to an embodiment of the present invention includes: a big data analysis part that receives video data captured by a general camera when collecting waste and analyzes input video data using artificial intelligence (AI) based on a pre-built learning model; and a waste information extraction part that extracts a plurality of waste information from the captured video data based on the results of the analysis using the artificial intelligence. Therefore, it is possible to increase work efficiency.

Description

Waste information analysis system and method {WASTE INFORMATION ANALYSIS SYSTEM AND METHOD}

Embodiments of the present invention relate to a waste information analysis system and method, and more specifically, a waste collector (user) automatically measures and confirms the waste type, heterogeneous waste mixture, collected volume, etc. It relates to a waste information analysis system and method capable of increasing efficiency and providing more accurate collection information to dischargers.

As awareness of global warming and environmental destruction increases, efforts to reduce carbon dioxide, save resources, and recycle resources for environmental protection are being made in various fields. In order to make efforts for resource recycling more effective, a policy of classifying and discharging waste by type is being implemented.

In general, to separate and discharge waste, a collection box or space is provided separately according to the type of waste so that citizens can distinguish the type of waste and input it voluntarily.

When such a waste collection box or space is filled to a certain level or more, it must be transported to a waste treatment facility using a vehicle for collecting it, but it is not easy to analyze the amount of waste by type.

Conventionally, as all data is manually measured and managed during the waste collection process, demand companies try to measure the volume as accurately as possible by using scales on containers. there was a case

Accordingly, a technology has been proposed to confirm that the collection box or space is filled with waste when the collection box or space is filled to a certain height or more by using an ultrasonic sensor or the like.

However, due to the nature of the ultrasonic sensor, when waste is piled up at a certain height in a specific location, an error such as displaying that there is no more space for waste to be filled in the collection box or space may occur even though there is a space to be filled with waste in another location. . In addition, since the ultrasonic sensor does not recognize objects, there is a problem in that it cannot determine the type of waste and cannot check the cleanliness of the collection box or space.

As related prior art, there is Republic of Korea Patent Publication No. 10-2020-0138465 (Title of Invention: Waste Analysis Intelligent CCTV System, Publication Date: December 10, 2020).

According to an embodiment of the present invention, a waste collector (user) can automatically measure and check the type of waste, whether or not mixed with heterogeneous waste, and the volume of the collected amount, etc. through video recording to increase work efficiency and provide more accurate collection information to the discharger. It provides a waste information analysis system and method that can be used.

The problem to be solved by the present invention is not limited to the above-mentioned problem (s), and another problem (s) not mentioned will be clearly understood by those skilled in the art from the following description.

The waste information analysis system according to an embodiment of the present invention receives image capture data taken by a general camera during waste collection and analyzes the input image capture data with artificial intelligence (AI) based on a pre-built learning model. It includes a big data analysis unit, and a waste information extraction unit for extracting a plurality of waste information from the video recording data based on the result of the analysis by the artificial intelligence.

The plurality of pieces of waste information may include at least one of a volume amount of waste, a type of waste, and information on whether or not heterogeneous waste is mixed.

The waste information analysis system according to an embodiment of the present invention acquires a plurality of learning data related to images of waste, generates waste-related big data, and learns to build the learning model using the generated waste-related big data. A model building unit may be further included.

In relation to the image of the waste, the learning model building unit obtains data for a plurality of camera images and a plurality of depth camera images (including depth value information) photographed by a general camera and a depth camera, and Post-processing may be performed to acquire at least one of learning data for estimating the waste volume, learning data for recognizing the mixing of heterogeneous waste and the type of loaded waste.

The learning model construction unit labels a plurality of camera images taken by the general camera, labels a plurality of depth camera images taken by a depth camera, and performs volume calculation based on the depth value information to obtain the Learning data for waste volume estimation can be acquired.

The learning model building unit may obtain learning data for recognizing the mixing of heterogeneous waste and the type of stacked waste by labeling and processing a plurality of camera images captured by the general camera.

The learning model builder converts original images and the post-processed information into one learning data set in relation to the plurality of camera images taken by the general camera and the plurality of depth camera images taken by the depth camera. can be saved

Waste information analysis method according to an embodiment of the present invention is artificial intelligence based on the learning model built-in learning model based on the learning model equipped server receives video data captured by a general camera when collecting waste (AI) analysis, and extracting a plurality of waste information from the image capture data based on the result of the analysis by the artificial intelligence by the server equipped with the learning model.

The waste information analysis method according to an embodiment of the present invention includes the steps of the learning model loading server acquiring a plurality of learning data related to the image of waste to generate waste-related big data, and the learning model loading server The method may further include building the learning model using waste-related big data.

The step of generating the waste-related big data includes a plurality of camera images captured by a general camera and a plurality of depth camera images (including depth value information) captured by a depth camera in relation to the image of the waste. It may include acquiring at least one of learning data for estimating waste volume and learning data for recognizing the type of mixed waste and loaded waste by performing data acquisition and post-processing.

Details of other embodiments are included in the detailed description and accompanying drawings.

According to an embodiment of the present invention, a waste collector (user) automatically measures and checks the type of waste, whether or not mixed with heterogeneous waste, and the volume of the collected amount, etc. through video recording to increase work efficiency and provide more accurate collection information to the discharger. can provide

1 is a diagram showing the network configuration of a waste information analysis system according to an embodiment of the present invention.
FIG. 2 is a block diagram illustrating a detailed configuration of a learning model loading server of FIG. 1 .
3 is a diagram showing an example of a process of acquiring learning data for waste volume estimation according to an embodiment of the present invention.
4 is a diagram showing an example of a process of acquiring learning data for recognizing heterogeneous waste mixing and loading waste types according to an embodiment of the present invention.
5 is a diagram illustrating an example of a process of extracting waste information based on image data by transmitting a picture taken with a general camera according to an embodiment of the present invention.
6 is a flowchart illustrating a waste information analysis method according to an embodiment of the present invention.

Advantages and/or features of the present invention, and methods of achieving them, will become apparent with reference to the following detailed description of the embodiments taken in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below and will be implemented in various forms different from each other, only these embodiments make the disclosure of the present invention complete, and common knowledge in the art to which the present invention pertains. It is provided to completely inform the person who has the scope of the invention, and the present invention is only defined by the scope of the claims. Like reference numbers designate like elements throughout the specification.

In addition, preferred embodiments of the present invention to be carried out below are provided in each system functional configuration in order to efficiently explain the technical components constituting the present invention, or system functions commonly provided in the technical field to which the present invention belongs. The configuration is omitted as much as possible, and the functional configuration that should be additionally provided for the present invention will be mainly described. If one of ordinary skill in the art to which the present invention pertains, one will be able to easily understand the functions of conventionally used components among the omitted functional configurations not shown below, and also the omitted configurations as described above. The relationship between elements and components added for the present invention will also be clearly understood.

In addition, in the following description, "transmission", "communication", "transmission", "reception" and other similar terms of signals or information refer to direct transmission of signals or information from one component to another. as well as passing through other components. In particular, "transmitting" or "transmitting" a signal or information as a component indicates the final destination of the signal or information, and does not mean a direct destination. The same is true for "reception" of signals or information.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a diagram showing a network configuration of a waste information analysis system according to an embodiment of the present invention, and FIG. 2 is a block diagram illustrating a detailed configuration of a learning model loading server 110 of FIG. 1 .

1 and 2, the learning model loading server 110 includes a learning model building unit 210, a big data analysis unit 220, a waste information extraction unit 230, and a control unit 240, can be configured.

The learning model builder 210 may acquire a plurality of learning data related to images of waste, generate waste-related big data, and build a learning model using the generated waste-related big data. Here, the learning model may be built in advance for analysis by the big data analyzer 220 .

That is, the learning model building unit 210 includes a plurality of camera images photographed by a general camera and a plurality of depth camera images (including depth value information) photographed by a depth camera in relation to the image of the waste. ), the learning model may be constructed by acquiring at least one of learning data for estimating waste volume and learning data for recognizing the type of heterogeneous waste mixed and loaded waste by performing data acquisition and post-processing.

For example, as shown in FIG. 3, the learning model building unit 210 labels a plurality of camera images taken by the general camera (pre-processing including labeling), and displays a plurality of camera images taken by the depth camera. A learning model (for estimating the volume of waste) may be constructed by obtaining learning data for estimating the waste volume by performing labeling processing (pre-processing) on the depth camera image and volume calculation based on the depth value information.

At this time, the learning model building unit 210 proceeds with learning a 3D shape estimation model necessary for estimating the volume of waste based on the collected data, that is, the plurality of camera images and the plurality of depth camera images, and this model is It can be configured by additionally connecting a model that estimates the volume of waste from the feature from 3D shape estimation.

In addition, as shown in FIG. 4, the learning model building unit 210 labels a plurality of camera images taken by the general camera (pre-processing including labeling) for mixing the heterogeneous waste and recognizing the type of the stacked waste. By acquiring learning data, it is possible to build the learning model (for recognizing the types of heterogeneous waste mixture and stacked waste).

At this time, the learning model builder 210 may proceed with classification of waste and learning of a model specialized for the inclusion of foreign substances based on a classification and segmentation model using the collected data, that is, the plurality of camera images.

Here, a CNN-based classification model can be used as an algorithm to classify the waste, and the position and type of the foreign object can be determined by using an X-ray foreign object detection algorithm and an external damage recognition algorithm model to determine the contamination of the foreign object. there is.

Meanwhile, the learning model building unit 210 learns the original image and the post-processed information in relation to the plurality of camera images captured by the general camera and the plurality of depth camera images captured by the depth camera. It can be saved as a data set. The learning model builder 210 may generate and update waste-related big data by continuously collecting the stored learning data set.

The big data analysis unit 220 may receive image capture data captured by a general camera during waste collection and analyze the input image capture data with artificial intelligence (AI) based on the learning model.

For example, as shown in FIG. 5, when video data captured by a general camera mounted on the user terminal 120 is transmitted to the learning model loading server 110 (transferring a photograph), the big data The analysis unit 220 may receive the transmitted image capturing data and perform artificial intelligence analysis based on the learning model. Accordingly, the learning model loading server 110 may return the artificial intelligence analysis result to the user terminal 120 .

The waste information extraction unit 230 may extract a plurality of waste information from the image capture data based on the result of analysis by the big data analysis unit 220 by artificial intelligence.

Here, the plurality of pieces of waste information may include at least one of information about the volume of waste, the type of waste, and whether or not heterogeneous waste is mixed. As the type of waste, food, plastic, can, glass, paper, vinyl, etc. can be recognized, and whether or not the heterogeneous waste is mixed can be recognized by detecting whether or not different types of waste are wet.

5 as an example, when the learning model loading server 110 returns the analysis result of the image capture data to the user terminal 120, the waste information extractor 230 extracts the image capture data. Through the analysis result, information such as the plurality of waste information, that is, waste volume (volume estimation), waste type (type recognition), heterogeneous waste mixing (mixing recognition), etc. may be extracted.

Accordingly, the user terminal 120 may display the plurality of waste information on the screen and provide the information to the user. The user can check waste information displayed on the screen of the user terminal 120 .

Accordingly, according to an embodiment of the present invention, it is possible to automatically obtain waste information that was manually entered by a user (manpower), thereby increasing the efficiency of waste collection and measuring the existing waste collection volume by relying on time and estimation. It is possible to accurately check the amount of waste volume where there was an error, and furthermore, it is possible to reduce the cost of the discharger and improve the reliability between the discharger and the collection carrier.

On the other hand, the server 110 equipped with the learning model may provide various services related to waste in connection with services such as monitoring waste discharge trends for each discharge customer and billing service costs.

The control unit 240 controls the overall operation of the learning model loading server 110, that is, the learning model building unit 210, the big data analysis unit 220, and the waste information extraction unit 230. can

The waste information analysis system 100 according to an embodiment of the present invention has the following effects in terms of technology.

First, the waste collection volume measurement technology through AI of the demand company is the first technology to be introduced to accurately measure the amount of waste collection.

Second, it was difficult to accurately calculate the volume of waste collection, which is difficult to measure, unless a plastic bag with a fixed volume is used.

Third, when the technology is introduced into living systems managed by local governments in the future, the collection amount can be accurately measured by volume measurement without introducing expensive equipment that measures weight.

In addition, the waste information analysis system 100 according to an embodiment of the present invention also has the following effects in terms of business value.

First, the waste collection, transportation, and treatment services of demand companies innovate the existing opaque and low-quality waste services. It is continuously developing to improve the efficiency of simple repetitive tasks such as checking the type of waste, whether mixed with heterogeneous waste, and checking the amount of waste collected manually by waste collectors, and to provide more accurate services. In particular, reliable and accurate measurement of the amount (volume) of waste collected can improve the reliability of the customer service as a pain point for customers who cannot be satisfied with the current service.

Second, the waste discharger can reduce costs by calculating the accurate amount of treatment, and the collector can improve the service satisfaction by providing the accurate amount of collection, thereby increasing the value of the service.

Third, waste collectors can contribute to efficient waste collection by reducing the difficulty of manually classifying waste types, checking whether heterogeneous waste is mixed, and estimating the amount of unclear waste discharged.

On the other hand, the waste information analysis system according to an embodiment of the present invention can be used in the waste collection step or treatment step. At this time, when the density of each waste type and the carbon reduction information during recycling are combined, the waste type and volume estimation information discharged by the emitter It is possible to predict the amount of carbon reduction, and it can also be used by the emitter to secure carbon credits.

The devices described above may be implemented as hardware components, software components, and/or a combination of hardware components and software components. For example, devices and components described in the embodiments may include, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable array (FPA), It may be implemented using one or more general purpose or special purpose computers, such as a programmable logic unit (PLU), microprocessor, or any other device capable of executing and responding to instructions. A processing device may run an operating system (OS) and one or more software applications running on the operating system. A processing device may also access, store, manipulate, process, and generate data in response to execution of software. For convenience of understanding, there are cases in which one processing device is used, but those skilled in the art will understand that the processing device includes a plurality of processing elements and/or a plurality of types of processing elements. It can be seen that it can include. For example, a processing device may include a plurality of processors or a processor and a controller. Other processing configurations are also possible, such as parallel processors.

Software may include a computer program, code, instructions, or a combination of one or more of the foregoing, which configures a processing device to operate as desired or processes independently or collectively. The device can be commanded. Software and/or data may be any tangible machine, component, physical device, virtual equipment, computer storage medium or device, intended to be interpreted by or provide instructions or data to a processing device. , or may be permanently or temporarily embodied in a transmitted signal wave. Software may be distributed on networked computer systems and stored or executed in a distributed manner. Software and data may be stored on one or more computer readable media.

6 is a flowchart illustrating a waste information analysis method according to an embodiment of the present invention.

The waste information analysis method described here is only one embodiment of the present invention, and other than that, various steps may be added as follows, and the following steps may be performed by changing the order, so that the present invention It is not limited to each step and its order described below.

Referring to FIGS. 1 and 6 , in step 610, the learning model loading server 110 may acquire a plurality of learning data related to an image of waste and generate big data related to waste.

At this time, the learning model loading server 110 displays a plurality of camera images captured by a general camera mounted on the user terminal 120 and a plurality of images captured by a depth camera in relation to the image of the waste. Data acquisition and post-processing may be performed on the depth camera image (including depth value information) to obtain at least one of learning data for estimating waste volume, learning data for mixing heterogeneous waste and recognizing types of loaded waste.

Next, in step 620, the learning model loading server 110 may build a learning model using the generated waste-related big data.

Next, in step 630, the learning model loading server 110 may receive image data captured by a general camera during waste collection.

Next, in step 640, the learning model loading server 110 may analyze the input image capture data with artificial intelligence (AI) based on the learning model.

Next, in step 650, the learning model loading server 110 may extract a plurality of pieces of waste information from the image capture data based on the result of analysis by the artificial intelligence.

The method according to the embodiment may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer readable medium. The computer readable medium may include program instructions, data files, data structures, etc. alone or in combination. Program commands recorded on the medium may be specially designed and configured for the embodiment or may be known and usable to those skilled in computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks and magnetic tapes, optical media such as CDROMs and DVDs, and magnetic-optical media such as floptical disks. Included are hardware devices specially configured to store and execute program instructions, such as magneto-optical media and ROM, RAM, flash memory, and the like. Examples of program instructions include high-level language codes that can be executed by a computer using an interpreter, as well as machine language codes such as those produced by a compiler. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the embodiments, and vice versa.

As described above, although the embodiments have been described with limited examples and drawings, those skilled in the art can make various modifications and variations from the above description. For example, the described techniques may be performed in an order different from the method described, and/or components of the described system, structure, device, circuit, etc. may be combined or combined in a different form than the method described, or other components may be used. Or even if it is replaced or substituted by equivalents, appropriate results can be achieved.

Therefore, other implementations, other embodiments, and equivalents of the claims are within the scope of the following claims.

110: learning model loading server
120: user terminal
210: learning model building unit
220: big data analysis unit
230: waste information extraction unit
240: control unit

Claims (10)

A big data analysis unit that receives video data taken by a general camera during waste collection and analyzes the input video data with artificial intelligence (AI) based on a pre-built learning model; and
A waste information extraction unit for extracting a plurality of waste information from the image capturing data based on the result of the analysis by the artificial intelligence.
Waste information analysis system comprising a.
According to claim 1,
The plurality of waste information
A waste information analysis system comprising at least one of information on the volume of waste, the type of waste, and whether or not heterogeneous waste is mixed.
According to claim 1,
A learning model construction unit that acquires a plurality of learning data related to images of waste, generates waste-related big data, and builds the learning model using the generated waste-related big data.
Waste information analysis system characterized in that it further comprises.
According to claim 3,
The learning model building unit
In relation to the image of the waste, data acquisition and post-processing are performed on a plurality of camera images taken by a general camera and a plurality of depth camera images (including depth value information) taken by a depth camera, A waste information analysis system characterized by acquiring at least one of learning data for estimating waste volume and learning data for recognizing the type of mixed and loaded waste with heterogeneous waste.
According to claim 4,
The learning model building unit
Learning for estimating the waste volume by labeling a plurality of camera images taken by the general camera, labeling a plurality of depth camera images taken by a depth camera, and performing volume calculation based on the depth value information A waste information analysis system characterized in that it acquires data.
According to claim 4,
The learning model building unit
Waste information analysis system, characterized in that for labeling and processing a plurality of camera images taken by the general camera to acquire learning data for recognizing the mixed waste and the type of stacked waste.
According to claim 4,
The learning model building unit
In relation to the plurality of camera images photographed by the general camera and the plurality of depth camera images photographed by the depth camera, the original image and the post-processed information are stored as one learning data set. Waste information analysis system characterized in that to do.
A server equipped with a learning model receives video data taken by a general camera when collecting waste, and analyzes the input video data with artificial intelligence (AI) based on a pre-built learning model; and
Extracting a plurality of waste information from the image capture data based on the result of the analysis by the learning model loaded server with the artificial intelligence
Waste information analysis method comprising a.
According to claim 8,
generating big data related to waste by acquiring a plurality of learning data related to images of waste by the learning model loading server; and
Building, by the learning model loading server, the learning model using the generated waste-related big data
Waste information analysis method characterized in that it further comprises.
According to claim 9,
The step of generating the waste-related big data is
In relation to the image of the waste, data acquisition and post-processing are performed on a plurality of camera images taken by a general camera and a plurality of depth camera images (including depth value information) taken by a depth camera, Acquiring at least one of learning data for estimating waste volume and learning data for recognizing the type of mixed and loaded waste with heterogeneous waste
Waste information analysis method comprising a.

Images