Method for monitoring input being fed into dumpster, and dumpster for executing the same

Abstract

A method for monitoring inputs into a dumpster and a dumpster for performing the same are provided. According to embodiments of the present invention, by using a camera installed in the dumpster to automatically determine the possibility that the inputs are prohibited and the overloading status of the loading compartment, and by transmitting an immediate alarm message to a control server based on the determination result, the dependency on humans in monitoring the dumpster can be minimized and an automatic reporting system can be established.

Description

Technology Field

[0001] Embodiments of the present invention relate to a technology for monitoring inputs fed into a dumpster, and more specifically, to a technology for monitoring inputs fed into a dumpster through a plurality of cameras and an artificial intelligence-based object classification model.

[0002] This invention is a research project conducted with the support of the Ministry of Climate, Energy and Environment (Project No.: B0080421002964, Research Project Title: Commercialization of Securing Exclusive U.S. Commercial Waste Collection Rights Using AIoT-based Recyclable Waste Classification Technology). Background Technology

[0004] A dumpster is a mobile container for loading waste that can be transported to a waste disposal site by a garbage truck or lifted and emptied on-site by a garbage truck. To facilitate efficient collection operations by these trucks, technology utilizing LiDAR, ultrasound, and Time of Flight (ToF) sensors to measure the internal load of dumpsters has been developed; however, this conventional load measurement technology had limitations in identifying prohibited items (e.g., waste tires, home appliances, etc.) that users indiscriminately deposit into dumpsters. Furthermore, conventional methods were limited to attaching notices regarding prohibited items around the dumpster to prevent illegal dumping, and there was a problem in that they could not automatically issue warnings or record instances of users depositing prohibited items. Additionally, conventional methods faced the problem of difficulty in simultaneously recognizing the external environment of the dumpster, the user's act of depositing prohibited items, and the internal loading status of the dumpster, as a single camera provided only a single-directional field of view. Prior art literature

[0006] Korean Patent Publication No. 10-2784265 (March 17, 2025) Korean Patent Publication No. 10-2641485 (February 22, 2024) The problem to be solved

[0007] The embodiments of the present invention are intended to automatically monitor prohibited items and overloading conditions in dumpsters to minimize reliance on humans and establish an automatic reporting system. means of solving the problem

[0009] According to one embodiment, a dumpster is provided comprising: a first camera provided on the outer surface of the dumpster to photograph an input located outside the dumpster; a second camera provided on the inner side of the dumpster to photograph the loading compartment of the dumpster; and an image processing unit that determines whether the input corresponds to a set prohibited input by analyzing a first image captured by the first camera through an artificial intelligence-based object classification model, determines the overloaded state of the loading compartment by calculating the difference between the top of the dumpster and the cumulative height of the input loaded in the loading compartment from a second image captured by the second camera, and transmits an alarm message to a control server when it is determined that the input corresponds to the prohibited input or that the loading compartment is overloaded.

[0010] The image processing unit may determine that the input is a prohibited item if the size of the input relative to the total screen size of the first image is greater than or equal to a first reference value and the shape of the input corresponds to a shape of a predefined class, and may determine that the loading box is overloaded if the difference between the top of the dumpster and the cumulative height of the input loaded in the loading box from the second image is less than a second reference value.

[0011] The image processing unit above calculates the possibility that the input is a prohibited item based on the size and shape of the input as a score within a set numerical range, and can perform at least one of a plurality of different additional measures set in relation to the input based on the score.

[0012] The image processing unit, when the score is greater than or equal to the first value, considers the input as corresponding to the prohibited input and blocks the cover of the dumpster while simultaneously transmitting an alarm message to the control server; when the score is greater than or equal to the second value but less than the first value, requests authentication from the user inserting the input, and when the input is inserted into the loading compartment, controls the operation of the second camera to allow close-up shooting of the input within a set distance from the input; and when the score is less than the second value, considers the input as not corresponding to the prohibited input and can open the cover of the dumpster.

[0013] The image processing unit can predict the next overloaded state of the loading box based on the overloaded state of the loading box before the input is fed into the dumpster and the size and shape of the input included in the first image, and then determine whether to block the cover according to the prediction result.

[0014] According to another embodiment, a method for monitoring an input to a dumpster is provided, comprising: a step of photographing an input located outside the dumpster with one or more first cameras provided on the outer surface of the dumpster; a step of photographing the loading compartment of the dumpster with a second camera provided on the inner side of the dumpster; a step of determining whether the input corresponds to a set prohibited input by analyzing a first image taken by the first camera through an artificial intelligence-based object classification model in an image processing unit; a step of determining the overloaded state of the loading compartment by calculating the difference between the top of the dumpster and the cumulative height of the input loaded in the loading compartment from a second image taken by the second camera in the image processing unit; and a step of transmitting an alarm message to a control server in the image processing unit when it is determined that the input corresponds to the prohibited input or that the loading compartment is overloaded.

[0015] The step of determining whether the input corresponds to a set prohibited input is to determine that the input corresponds to the prohibited input if the size of the input relative to the total screen size of the first image is greater than or equal to a first threshold and the shape of the input corresponds to the shape of a predefined class, and the step of determining the overloaded state of the cargo box is to determine that the cargo box is overloaded if the difference between the top of the dumpster and the cumulative height of the input loaded in the cargo box from the second image is less than a second threshold.

[0016] The step of determining whether the above-mentioned input corresponds to a set prohibited input is to calculate the probability of the above-mentioned input corresponding to a prohibited input as a score within a set numerical range based on the size and shape of the above-mentioned input, and the method for monitoring the above-mentioned input may further include, after the step of determining whether the above-mentioned input corresponds to a set prohibited input, a step of performing at least one of a plurality of different additional measures set in relation to the above-mentioned input based on the score in the image processing unit.

[0017] The step of performing at least one of the plurality of different additional measures is as follows: if the score is greater than or equal to the first value, the input is considered to be the prohibited input, and the cover of the dumpster is blocked while simultaneously sending an alarm message to the control server; if the score is greater than or equal to the second value but less than the first value, authentication is requested from the user inserting the input, and when the input is inserted into the loading compartment, the operation of the second camera is controlled so that the input can be photographed at close range within a set distance from the input; and if the score is less than the second value, the input is considered not to be the prohibited input, and the cover of the dumpster can be opened.

[0018] The method for monitoring an input to be fed into the dumpster may further include, after the step of photographing an input intended to be fed into the dumpster, the image processing unit predicts the next overloaded state of the cargo compartment based on the overloaded state of the cargo compartment before the input is fed into the dumpster and the size and shape of the input included in the first image, and then determines whether to block the cover according to the prediction result. Effects of the invention

[0020] According to embodiments of the present invention, by using a camera installed on a dumpster to automatically determine the possibility of the input being prohibited and the overloading status of the loading compartment, and by transmitting an immediate alarm message to a control server based on the determination result, the dependency on humans in monitoring dumpsters can be minimized and an automatic reporting system can be established.

[0021] In addition, according to embodiments of the present invention, the possibility of an input object being a prohibited object is calculated as a score within a set numerical range based on an image captured by the first camera of the dumpster, and additional measures such as transmitting an alarm message to a control server, user authentication, control of the operation of the cover, and control of the operation of the second camera are automatically performed based on the score, thereby more accurately determining the possibility of an input object being a prohibited object and increasing the utility of monitoring the dumpster by immediately recording and notifying an administrator of acts that may be illegal dumping.

[0022] In addition, according to embodiments of the present invention, the accuracy of the judgment regarding the possibility of an input being a prohibited item and the overloading condition of the loading container can be improved based on the analysis results of each of the first image captured by the first camera and the second image captured by the second camera, and in particular, the utility of monitoring the dumpster can be increased by controlling the first camera and the second camera to operate complementarily. Brief explanation of the drawing

[0024] FIG. 1 is a block diagram showing the detailed configuration of a dumpster monitoring system according to embodiments of the present invention. FIG. 2 is a block diagram showing the detailed configuration of a dumpster according to embodiments of the present invention. FIG. 3 is an example of an implementation of a dumpster according to an embodiment of the present invention. FIG. 4 is an example of an implementation of a dumpster according to another embodiment of the present invention. FIG. 5 is an example of an implementation of a first camera and a second camera according to embodiments of the present invention. FIG. 6 is a drawing for explaining the image processing process in an image processing unit according to embodiments of the present invention. FIG. 7 is a flowchart illustrating a method for monitoring an input according to an embodiment of the present invention. FIG. 8 is a flowchart illustrating step S106 of FIG. 7 FIG. 9 is a flowchart illustrating a method for monitoring an input according to another embodiment of the present invention. FIG. 10 is a flowchart illustrating a method for monitoring an input according to another embodiment of the present invention. FIG. 11 is a block diagram illustrating a computing environment including a computing device suitable for use in exemplary embodiments. Specific details for implementing the invention

[0025] Hereinafter, specific embodiments of the present invention will be described with reference to the drawings. The following detailed description is provided to facilitate a comprehensive understanding of the methods, apparatuses, and / or systems described herein. However, this is merely illustrative and the present invention is not limited thereto.

[0026] In describing the embodiments of the present invention, detailed descriptions of known technologies related to the present invention are omitted if it is determined that such detailed descriptions may unnecessarily obscure the essence of the present invention. Furthermore, the terms described below are defined in consideration of their functions within the present invention, and these may vary depending on the intentions or practices of the user or operator. Therefore, such definitions should be based on the content throughout this specification. Terms used in the detailed description are intended merely to describe the embodiments of the present invention and should not be limiting in any way. Unless explicitly stated otherwise, expressions in the singular form include the meaning of the plural form. In this description, expressions such as "include" or "comprise" are intended to refer to certain characteristics, numbers, steps, actions, elements, parts thereof, or combinations thereof, and should not be interpreted to exclude the existence or possibility of one or more other characteristics, numbers, steps, actions, elements, parts thereof, or combinations thereof other than those described.

[0028] FIG. 1 is a block diagram showing the detailed configuration of a dumpster monitoring system (100) according to embodiments of the present invention. As shown in FIG. 1, the dumpster monitoring system (100) according to embodiments of the present invention includes a plurality of dumpsters (110) and a control server (130). At this time, the plurality of dumpsters (110) and the control server (130) may be interconnected through a network (120). Here, the network (120) may include, for example, a mobile network such as the Internet, a Wi-Fi network, a 3G network, an LTE network, a wire area network, etc.

[0029] A dumpster (110) is a movable container for loading inputs and may be equipped with a loading space of a predetermined size, i.e., a loading box, inside. Here, inputs are a general term for objects fed into the dumpster, i.e., waste, and are used in a broad sense to include not only commercial waste discharged from offices, shops, buildings, etc., but also industrial waste such as piles of scrap metal waste and piles of construction cement, marine waste, household waste, etc.

[0030] Each dumpster (110) may be installed in one or more different areas and may have various sizes and shapes. These dumpsters (110) may be transported to a waste disposal site by a garbage collection vehicle or lifted and emptied by a garbage collection vehicle at the site. At this time, the dumpster (110) is provided with a dumpster hole for docking a front loader. A worker may operate an input means (e.g., a joystick) provided on the garbage collection vehicle to dock the front loader into the dumpster hole and collect the input loaded in the dumpster (110).

[0031] Additionally, as described below, the dumpster (110) may be equipped with multiple cameras, and the exterior and interior of the dumpster (110) can be photographed respectively through the cameras. Additionally, the dumpster (110) can analyze the video (or image frame) captured through the cameras to determine whether the input material corresponds to a set prohibited material or whether the cargo compartment is overloaded. If the dumpster (110) determines that the input material corresponds to a prohibited material or that the cargo compartment is overloaded, it can send an alarm message to the control server (130).

[0032] The control server (130) is connected to each dumpster (110) via a network (120) and monitors each dumpster (110). The control server (130) can manage the size, shape, identification number, installation location, user information, etc. of each dumpster (110) by storing them in an internal database (not shown). In addition, the control server (130) can determine the operation method of the garbage collection vehicle and whether a manager is dispatched to the site by monitoring the current status of each dumpster (110), for example, whether prohibited items are being inserted, the overloading status of the cargo container, etc. In particular, the control server (130) can receive an alarm message from at least one of the dumpsters (110), and can determine measures corresponding to the alarm message, such as connecting to a manager, sending out a guidance message, imposing a fine, or adjusting the collection route of the garbage collection vehicle, according to the alarm message. At this time, each dumpster (110) may be equipped with an embedded system, and through this embedded system, it may determine whether prohibited items are being inserted, the overloaded state of the cargo box, etc., and then send an alarm message to the control server (130).

[0033] Below, we will examine the detailed configuration and operation of the dumpster (110) in more detail with reference to FIGS. 2 to 10.

[0035] FIG. 2 is a block diagram showing the detailed configuration of a dumpster (110) according to embodiments of the present invention. As shown in FIG. 2, the dumpster (110) according to embodiments of the present invention includes a first camera (210), a second camera (220), a cover (230), an image processing unit (240), a user authentication unit (250), and a database (260).

[0036] One or more first cameras (210) are provided on the outer surface of the dumpster (110) to photograph an input located outside the dumpster (110). For example, one or more first cameras (210) may be provided on the upper outer surface of the dumpster (110). The first camera (210) can photograph the user and the input during the process of the user approaching the dumpster (110) to insert the input into the dumpster (110). Accordingly, the first camera (210) acquires a first image containing the user and the input, and can identify the user and the input respectively by removing a previously stored background area from the first image. As described below, the image processing unit (240) can determine whether the input corresponds to a set prohibited input by analyzing the first image captured by the first camera (210) through an artificial intelligence-based object classification model.

[0037] A second camera (220) is provided on the inside of the dumpster (110) to photograph the loading compartment of the dumpster (110). For example, one or more second cameras (220) may be provided on the inner surface of the top side of the dumpster (110). The second camera (220) may be configured to photograph the loading compartment of the dumpster (110) at set intervals, immediately before the input is fed into the loading compartment, and immediately after the input is fed into the loading compartment, respectively. Accordingly, the second camera (220) can acquire a second image including the input loaded into the loading compartment and calculate the cumulative height of the input loaded into the loading compartment from the second image. As described below, the image processing unit (240) can determine the overloaded state of the loading compartment by calculating the difference between the top of the dumpster (110) and the cumulative height of the input loaded into the loading compartment from the second image captured by the second camera (220).

[0038] The cover (230) is provided on the upper side of the dumpster and is used to open and close the cargo compartment. The cover (230) may be a manual type that opens and closes according to the user's operation, but is not limited thereto. The cover (230) may be connected to the image processing unit (240) via a wired network or a wireless network, for example, and may be opened and closed, or switched to a locked state or an unlocked state according to the control of the image processing unit (240).

[0039] The image processing unit (240) analyzes the first image and the second image captured by the first camera (210) and the second camera (220), respectively, to determine whether the input corresponds to the set prohibited input and whether the loading container is overloaded. In the embodiments, the prohibited input is an input that cannot be collected by a garbage collection vehicle or requires to be collected separately from other inputs, and may be, for example, large waste including household appliances such as waste tires, refrigerators, and air conditioners; hazardous / flammable substances such as paint, fuel, chemical products, adhesives, pesticides, and herbicides; electronic waste such as automobile batteries, rechargeable batteries, computers, TVs, and monitors; propane tanks, industrial drums, contaminated soil, animal carcasses, food waste, etc.

[0040] First, the image processing unit (240) may determine that an input object is a prohibited object if the size of the input object relative to the total screen size of the first image is greater than or equal to a first reference value and the shape of the input object corresponds to a shape of a predefined class. At this time, the image processing unit (240) may calculate the probability that the input object is a prohibited object as a score within a set numerical range based on the size and shape of the input object. In addition, during the process of calculating the score, the image processing unit (240) may additionally consider the user's access pattern to the dumpster (110) before the input object is inserted, the color of the input object, etc. As described below, the image processing unit (240) may perform at least one of a plurality of different additional measures set in relation to the input object based on the score.

[0041] Next, the image processing unit (240) can determine that the loading compartment is overloaded if the difference between the top of the dumpster (110) and the cumulative height of the input loaded in the loading compartment is less than the second reference value from the second image. The image processing unit (240) can determine whether the loading compartment is overloaded by calculating the difference between the predefined top side point of the dumpster (110) and the cumulative height of the input loaded in the loading compartment, and then comparing the calculation result with the second reference value.

[0042] In this way, the image processing unit (240) can send an alarm message to the control server (130) when it is determined that the input is a prohibited item or that the loading box is overloaded.

[0043] Additionally, the image processing unit (240) can analyze consecutive image frames of the first image and, if the user's input behavior corresponds to a learned abnormal input behavior, generate a warning message and then transmit an alarm message to the control server (130). Here, the learned abnormal input behavior may be, for example, an act of a user approaching the dumpster (110) holding an input with a size greater than or equal to the first standard and forcibly opening the lid (230), or an act of a user leaving an input with a size greater than or equal to the first standard near the dumpster (110).

[0044] Meanwhile, the image processing unit (240) may exist as a separate hardware configuration from the first camera (210) and the second camera (220), but is not limited thereto, and may exist as a component of the first camera (210) and the second camera (220). That is, the image processing unit (240) may be embedded in various hardware devices as a component of the embedded system.

[0045] The user authentication unit (250) performs user authentication in accordance with the request of the image processing unit (240). As described above, the image processing unit (240) may perform at least one of a plurality of different additional measures set in relation to the input object based on the possibility of the input object being prohibited, i.e., the score. Here, the additional measures may be, for example, sending an alarm message to the control server (130), user authentication, control of the operation of the cover (230), control of the operation of the second camera (120), etc. In accordance with the user authentication request of the image processing unit (240), the user authentication unit (250) may authenticate the user by transmitting and receiving data with an application within a user terminal (not shown) held by the user, or by performing near-field data communication with the user terminal through NFC (Near Field Communication), QR code, etc. Through this user authentication process, the user authentication unit (250) can identify the user's name, gender, age, address, contact information, etc. If user authentication fails in the user authentication unit (250), the image processing unit (240) can block the cover (230) and request a connection of the administrator to the control server (130).

[0046] The database (260) is a storage facility where various information related to the dumpster (110) is stored. In the database (260), for example, the size, shape, identification number, installation location, user information, top edge information of the dumpster (110), a first image, a second image, the first image analysis result from the image processing unit (230), the second image analysis result, whether the cover (230) is open or closed, etc., can be stored in real time.

[0048] FIG. 3 is an example of implementation of a dumpster (110) according to one embodiment of the present invention, and FIG. 4 is an example of implementation of a dumpster (110) according to another embodiment of the present invention.

[0049] Referring to FIGS. 3 and 4, a first camera (210) may be installed on the outer surface of the dumpster (110) and a second camera (220) may be installed on the inner surface of the dumpster (110). Additionally, a cover (230) may be installed on the upper surface of the dumpster (110). At this time, there may be one or more covers (230). Furthermore, an image processing unit (240) may be provided on the inner surface of the dumpster (110), and a user authentication unit (250) may be provided on the outer surface of the dumpster (110).

[0050] Meanwhile, it should be noted that the size, shape, and installation positions of each component constituting the dumpster (110) shown in FIGS. 3 and 4 are merely examples and are not limited to those shown in FIGS. 3 and 4.

[0052] FIG. 5 is an example of an implementation of a first camera (210) and a second camera (220) according to embodiments of the present invention.

[0053] Referring to FIG. 5, one or more first cameras (210) are provided on the outer surface of the dumpster (110) to photograph the contents located outside the dumpster (110). Additionally, a second camera (220) is provided on the inner side of the dumpster (110) to photograph the contents of the dumpster (110). That is, the first camera (210) and the second camera (220) can be installed at positions capable of photographing the exterior and interior of the dumpster (110), respectively.

[0054] The first camera (210) can monitor the approach of garbage collection vehicles and illegal dumping by users by, for example, capturing the front of the dumpster (110). The image processing unit (230) can determine whether the input object corresponds to a set prohibited input object by analyzing the first image captured by the first camera (210) through an artificial intelligence-based object classification model.

[0055] Additionally, the second camera (220) can photograph the input loaded in the cargo compartment to monitor the loading amount of the cargo compartment. The image processing unit (230) can, for example, create a virtual line connecting a point set on the upper side of the dumpster (110) and the second camera (220), and calculate the difference between the line and the cumulative height of the input loaded in the cargo compartment to determine the overloading condition of the cargo compartment.

[0057] FIG. 6 is a diagram illustrating the image processing process in an image processing unit (230) according to embodiments of the present invention.

[0058] Referring to FIG. 6, the image processing unit (230) is equipped with an artificial intelligence-based object classification model and can analyze a first image captured by a first camera (210) through the object classification model. Here, the object classification model can be implemented as a deep learning model such as a CNN (Convolutional Neural Network), an RNN (Recurrent Neural Network), etc. As described above, the image processing unit (230) can determine that the input is a prohibited object if the size of the input relative to the total screen size of the first image is greater than or equal to a first threshold and the shape of the input corresponds to the shape of a predefined class.

[0059] At this time, the image processing unit (230) can calculate the probability that the input is a prohibited item based on the size and shape of the input as a score within a set numerical range. Specifically, the image processing unit (230) can calculate the score in a direction where the probability of it being a prohibited item increases as the difference between the size of the input and the first reference value increases, and as the similarity between the shape of the input and the shape of the class stored in the database (260) increases. At this time, the image processing unit (230) can assign a first weight and a second weight, respectively, to the size and shape of the input during the process of calculating the score. In addition, the image processing unit (230) can dynamically change the first weight and the second weight according to the type of class among the classes stored in the database (260) that is the subject of comparison with the input included in the first image. For example, if the type of class that is the subject of comparison is a refrigerator, the image processing unit (230) can set the second weight for the shape of the input to be greater than the first weight for the size of the input. In the case of a refrigerator, since the diversity of its shape is not as great as the diversity of its size, the image processing unit (230) can determine whether the input corresponds to a refrigerator by relatively increasing the weight given to the shape of the input. These first weights and second weights may be stored in the database (260) for each type of class. Additionally, the image processing unit (240) may additionally consider the user's access pattern to the dumpster (110) and the color of the input before the input is inserted during the calculation of the score. For example, during the calculation of the score, if the user's access pattern to the dumpster (110) corresponds to a learned abnormal insertion behavior, or if the size, shape, and color of the input match all of the size, shape, and color of the set prohibited input, the image processing unit (240) may determine the possibility of it being a prohibited input with a score of a set numerical value, generate a warning message, and then transmit an alarm message to the control server (130).

[0060] The image processing unit (230) can express the probability that the input determined by this method corresponds to a prohibited input as a probability value and calculate a score between 0 and 100. Subsequently, the image processing unit (230) can perform at least one of a plurality of different additional measures set in relation to the input based on the score. Here, the additional measures may be, for example, sending an alarm message to the control server (130), user authentication, controlling the operation of the cover (230), controlling the operation of the second camera (120), etc.

[0061] Specifically, if the score is greater than or equal to a first value (e.g., 0.8 or greater), the image processing unit (230) may consider the input material to be a prohibited material and block the cover (230) while simultaneously transmitting an alarm message to the control server (130). At this time, the image processing unit (230) may output a guidance message through a speaker (not shown) provided in the dumpster (110) stating that the input material is a prohibited material and that input into the dumpster (110) is prohibited.

[0062] Additionally, the image processing unit (230) may request authentication from the user inserting the material when the score is greater than or equal to the second value and less than the first value (e.g., 0.4 or greater and less than 0.8), and then control the operation of the second camera (220) so that the material can be photographed at close range within a set distance from the material when the material is inserted into the loading container. That is, when the score is greater than or equal to the second value and less than the first value, the image processing unit (230) may determine that the material is in a somewhat ambiguous situation to be judged as prohibited material and perform additional measures including user authentication and additional shooting by the second camera (120). The user authentication unit (250) may authenticate the user by transmitting and receiving data with an application within a user terminal (not shown) held by the user in accordance with the user authentication request of the image processing unit (240), or by performing near-field data communication with the user terminal through NFC, QR code, etc. At this time, the image processing unit (230) may keep the cover (230) locked until user authentication is completed, and may switch the locked state of the cover (230) to an unlocked state after user authentication is completed. Subsequently, when an input is placed into the loading container, the image processing unit (230) may control the operation of the second camera (220) to photograph the input more closely through the second camera (220). In this case, the image processing unit (230) may acquire additional images of the input from multiple angles at a closer position through the second camera (220), and may secondarily determine the possibility that the input is a prohibited item by analyzing the additional images through an object classification model. If, after user authentication is completed, the score regarding the possibility that the input is a prohibited item is calculated to be greater than or equal to the first value, the image processing unit (230) may send an alarm message to the control server (130). In this case, the control server (130) may impose a fine or additional charge on the user who has completed user authentication.

[0063] Additionally, the image processing unit (230) may open the cover (230) by considering that the input material does not correspond to an input prohibited material when the score is less than the second value (e.g., less than 0.4).

[0064] In this way, according to embodiments of the present invention, the possibility of an input object being a prohibited object is calculated as a score within a set numerical range based on an image captured by the first camera (210) of the dumpster (110), and additional measures such as transmitting an alarm message to the control server (130), user authentication, control of the operation of the cover (230), and control of the operation of the second camera (120) are automatically performed based on the score, thereby more accurately determining the possibility of an input object being a prohibited object, and at the same time, increasing the utility of monitoring the dumpster (110) through immediate recording of acts that may be illegal dumping and notification to the manager.

[0065] Additionally, the image processing unit (240) can predict the next overloaded state of the cargo compartment based on the overloaded state of the cargo compartment before the input is fed into the dumpster (110) and the size and shape of the input included in the first image, and then determine whether to block the cover (230) according to the prediction result. That is, the image processing unit (240) can determine the overloaded state of the cargo compartment based on the second image captured by the second camera (220) before the input is fed into the dumpster (110), and can predict the next overloaded state of the cargo compartment when the input is fed into the cargo compartment based on the size and shape of the input included in the first image.

[0066] To this end, the image processing unit (240) can calculate the cumulative height of the input loaded in the loading compartment before the input is fed into the dumpster (110), and calculate the expected height at the time of input based on the size and shape of the input included in the first image. Subsequently, the image processing unit (240) can predict the next overloaded state of the loading compartment when the input is fed into the loading compartment by adding the expected height at the time of input to the cumulative height of the input loaded in the loading compartment before the input is fed into the dumpster (110). At this time, the image processing unit (240) can calculate the expected height differently according to the size of the input by considering the shape of the input. Generally, the input is not loaded into the loading compartment while standing upright, but may be loaded into the loading compartment lying down in different shapes depending on its shape. In this case, the expected height at the time of input may vary depending on the size (e.g., width, length, height, etc.), shape (e.g., rectangular shape, cylindrical shape, rod shape, etc.), and the loading form of the input loaded in the loading container. The image processing unit (240) can predict the expected height at the time of input according to the size, shape, and loading form of the input loaded in the loading container based on previously learned training data. Here, the training data may include information regarding the size of the input, the shape of the input, the loading form of the input loaded in the loading container, and the height of the input loaded in the loading container when the input is inserted into the loading container. For example, the image processing unit (240) may select training data corresponding to the size, shape, and loading form of the input loaded in the loading container from the training data, and determine the average of the heights of the inputs inserted into the loading container when the input is inserted into the loading container corresponding to each of the selected training data as the expected height at the time of input.Afterward, the image processing unit (240) can predict the next overloaded state of the loading compartment when the input is loaded by adding the expected height at the time of input to the cumulative height of the input loaded in the loading compartment before the input is loaded into the dumpster (110), and can determine whether to block the cover (230) according to the prediction result. If the next overloaded state of the loading compartment, that is, the difference between the top of the dumpster (110) and the cumulative height of the input loaded in the loading compartment is less than the second threshold, the image processing unit (240) can block the cover (230) regardless of whether the input corresponds to a prohibited input. At this time, the image processing unit (240) can output a guidance message stating that the input cannot be loaded into the loading compartment due to its size and shape.

[0067] In this way, according to embodiments of the present invention, the accuracy of the judgment regarding the possibility of the input being a prohibited item and the overloading condition of the loading container can be improved based on the analysis results of each of the first image captured by the first camera (210) and the second image captured by the second camera (220), and in particular, the utility of monitoring the dumpster (110) can be increased by controlling the first camera (210) and the second camera (220) to operate complementarily.

[0069] FIG. 7 is a flowchart illustrating a method for monitoring an input according to an embodiment of the present invention. In the illustrated flowchart, the method is described by dividing it into a plurality of steps, but at least some of the steps may be performed in a different order, combined with other steps and performed together, omitted, divided into detailed steps, or performed with one or more steps not illustrated added.

[0070] In step S102, the first camera (210) photographs the input located outside the dumpster (110).

[0071] In step S104, the image processing unit (230) determines whether the input object corresponds to a set prohibited object by analyzing the first image captured by the first camera (210) through an artificial intelligence-based object classification model. At this time, the image processing unit (230) can calculate the probability that the input object corresponds to a prohibited object as a score within a set numerical range based on the size and shape of the input object.

[0072] In step S106, the image processing unit (230) performs at least one of a plurality of different additional measures set in relation to the input based on the score.

[0074] Figure 8 is a flowchart illustrating step S106 of Figure 7.

[0075] In steps S202, S204, and S206, the numerical range of the score regarding the possibility that the input is a prohibited item is determined.

[0076] If the above score is greater than or equal to the first value, in steps S208 and S210, the image processing unit (230) may consider the input material to be a prohibited input material and block the cover (230) of the dumpster (110) while simultaneously sending an alarm message to the control server (130).

[0077] If the score is greater than or equal to the second value and less than the first value, in steps S212 and S214, the image processing unit (230) may request authentication from the user inserting the input, and then, when the input is inserted into the loading container, control the operation of the second camera (220) so that the input can be photographed at close range within a set distance from the input.

[0078] If the score is less than the second value, in step S216, the image processing unit (230) may open the cover (230) of the dumpster (110) by considering that the input does not correspond to an input prohibited material.

[0080] FIG. 9 is a flowchart illustrating a method for monitoring an input according to another embodiment of the present invention.

[0081] In step S302, the second camera (220) photographs the cargo compartment of the dumpster (110).

[0082] In step S304, the image processing unit (230) calculates the difference between the top of the dumpster (110) and the accumulated height of the input loaded in the storage compartment from the second image captured by the second camera (220) to determine the overloaded state of the storage compartment.

[0083] In step S306, if the image processing unit (230) determines that the cargo box is overloaded, it sends an alarm message to the control server (130).

[0085] FIG. 10 is a flowchart illustrating a method for monitoring an input according to another embodiment of the present invention.

[0086] In step S402, the image processing unit (230) predicts the next overloaded state of the loading box based on the overloaded state of the loading box before the input is fed into the dumpster, and the size and shape of the input included in the first image.

[0087] In step S404, the image processing unit (230) determines whether to block the cover (230) according to the next overloaded state of the predicted cargo box.

[0089] FIG. 11 is a block diagram illustrating a computing environment including a computing device suitable for use in exemplary embodiments. In the illustrated embodiments, each component may have different functions and capabilities in addition to those described below, and may include additional components in addition to those not described below.

[0090] The illustrated computing environment (10) includes a computing device (12). In one embodiment, the computing device (12) may be one or more components included in a dumpster monitoring system (100) or a dumpster (110).

[0091] The computing device (12) includes at least one processor (14), a computer-readable storage medium (16), and a communication bus (18). The processor (14) can cause the computing device (12) to operate according to the exemplary embodiment described above. For example, the processor (14) can execute one or more programs stored in the computer-readable storage medium (16). The one or more programs may include one or more computer-executable instructions, and the computer-executable instructions may be configured to cause the computing device (12) to perform operations according to the exemplary embodiment when executed by the processor (14).

[0092] A computer-readable storage medium (16) is configured to store computer-executable instructions or program code, program data and / or other suitable forms of information. A program (20) stored in the computer-readable storage medium (16) includes a set of instructions executable by a processor (14). In one embodiment, the computer-readable storage medium (16) may be memory (volatile memory such as random access memory, non-volatile memory, or a suitable combination thereof), one or more magnetic disk storage devices, optical disk storage devices, flash memory devices, or other forms of storage media that are accessed by a computing device (12) and capable of storing desired information, or a suitable combination thereof.

[0093] The communication bus (18) interconnects various other components of the computing device (12), including the processor (14) and the computer-readable storage medium (16).

[0094] The computing device (12) may also include one or more input / output interfaces (22) and one or more network communication interfaces (26) that provide interfaces for one or more input / output devices (24). The input / output interfaces (22) and network communication interfaces (26) are connected to a communication bus (18). The input / output devices (24) may be connected to other components of the computing device (12) through the input / output interfaces (22). An exemplary input / output device (24) may include an input device such as a pointing device (such as a mouse or trackpad), a keyboard, a touch input device (such as a touchpad or touchscreen), a voice or sound input device, various types of sensor devices and / or imaging devices, and / or an output device such as a display device, a printer, a speaker and / or a network card. An exemplary input / output device (24) may be included inside the computing device (12) as a component constituting the computing device (12), or it may be connected to the computing device (12) as a separate device distinct from the computing device (12).

[0096] Although the present invention has been described in detail above through representative embodiments, those skilled in the art will understand that various modifications can be made to the aforementioned embodiments without departing from the scope of the present invention. Therefore, the scope of the present invention should not be limited to the described embodiments, but should be defined by the claims set forth below as well as equivalents thereof. Explanation of the symbols

[0098] 100: Dumpster surveillance system 110 : Dumpster 120 : Network 130 : Control Server 210: 1st Camera 220: Second camera 230 : Cover 240 : Image processing unit 250 : User Authentication Section 260 : Database

Claims

Claim 1 A dumpster comprising: one or more first cameras provided on the outer surface of the dumpster to photograph an input located outside the dumpster; a second camera provided on the inner side of the dumpster to photograph the loading compartment of the dumpster; and an image processing unit that determines whether the input corresponds to a set prohibited input by analyzing a first image captured by the first camera through an artificial intelligence-based object classification model, determines the overloaded state of the loading compartment by calculating the difference between the top of the dumpster and the cumulative height of the input loaded in the loading compartment from a second image captured by the second camera, and transmits an alarm message to a control server when it is determined that the input corresponds to the prohibited input or that the loading compartment is overloaded. Claim 2 A dumpster according to claim 1, wherein the image processing unit determines that the input is a prohibited input if the size of the input relative to the total screen size of the first image is greater than or equal to a first reference value and the shape of the input corresponds to a shape of a predefined class, and determines that the loading compartment is overloaded if the difference between the top of the dumpster and the cumulative height of the input loaded in the loading compartment from the second image is less than a second reference value. Claim 3 A dumpster according to claim 2, wherein the image processing unit calculates the probability of the input being a prohibited item based on the size and shape of the input as a score within a set numerical range, and performs at least one of a plurality of different additional measures set in relation to the input based on the score. Claim 4 In claim 3, the image processing unit considers the input to be the prohibited input when the score is greater than or equal to the first value, blocks the cover of the dumpster, and simultaneously transmits an alarm message to the control server; when the score is greater than or equal to the second value and less than the first value, requests authentication from the user inserting the input, and when the input is inserted into the loading compartment, controls the operation of the second camera to allow close-up shooting of the input within a set distance from the input; and when the score is less than the second value, considers the input not to be the prohibited input and opens the cover of the dumpster. Claim 5 A dumpster according to claim 1, wherein the image processing unit predicts the next overloaded state of the loading compartment based on the overloaded state of the loading compartment before the input is fed into the dumpster and the size and shape of the input included in the first image, and then determines whether to block the cover of the dumpster according to the prediction result. Claim 6 A method for monitoring an input to a dumpster, comprising: a step of photographing an input located outside the dumpster using a first camera provided on the outer surface of the dumpster; a step of photographing the loading compartment of the dumpster using a second camera provided on the inner side of the dumpster; a step of determining whether the input corresponds to a set prohibited input by analyzing a first image captured by the first camera through an artificial intelligence-based object classification model in an image processing unit; a step of determining the overloaded state of the loading compartment by calculating the difference between the top of the dumpster and the cumulative height of the input loaded in the loading compartment from a second image captured by the second camera in the image processing unit; and a step of transmitting an alarm message to a control server in the image processing unit when it is determined that the input corresponds to the prohibited input or that the loading compartment is overloaded. Claim 7 A method for monitoring an input into a dumpster according to claim 6, wherein the step of determining whether the input corresponds to a set prohibited input is to determine that the input corresponds to the prohibited input if the size of the input relative to the total screen size of the first image is greater than or equal to a first reference value and the shape of the input corresponds to a shape of a predefined class, and the step of determining the overloaded state of the loading compartment is to determine that the loading compartment is overloaded if the difference between the top of the dumpster and the cumulative height of the input loaded in the loading compartment from the second image is less than a second reference value. Claim 8 A method for monitoring an input into a dumpster according to claim 7, wherein the step of determining whether the input corresponds to a set prohibited input is to calculate the probability of the input corresponding to a prohibited input as a score within a set numerical range based on the size and shape of the input, and the method for monitoring the input further comprises, after the step of determining whether the input corresponds to a set prohibited input, the step of performing at least one of a plurality of different additional measures set in relation to the input based on the score in the image processing unit. Claim 9 A method for monitoring an input into a dumpster according to claim 8, wherein the step of performing at least one of the plurality of different additional measures comprises: if the score is greater than or equal to a first value, considering the input as corresponding to the prohibited input, blocking the cover of the dumpster and simultaneously transmitting an alarm message to the control server; if the score is greater than or equal to a second value but less than the first value, requesting authentication from the user inserting the input, and then controlling the operation of the second camera so that the input can be photographed up close within a set distance from the input when the input is inserted into the loading compartment; and if the score is less than the second value, considering the input as not corresponding to the prohibited input, and opening the cover of the dumpster. Claim 10 A method for monitoring an input to be fed into a dumpster according to claim 6, further comprising, after the step of photographing an input to be fed into the dumpster, the step of predicting the next overloaded state of the loading compartment based on the overloaded state of the loading compartment before the input is fed into the dumpster and the size and shape of the input included in the first image, and determining whether to block the cover of the dumpster according to the prediction result.

Images