CN-121974059-A - Intelligent garbage classification device based on differential double-layer conveying and visual recognition

CN121974059ACN 121974059 ACN121974059 ACN 121974059ACN-121974059-A

Abstract

An intelligent garbage classification device based on differential double-layer conveying and visual identification relates to the technical field of intelligent garbage classification. The device comprises an induction cover opening mechanism, a garbage dispersing and conveying mechanism, a visual identification device, a bag opening mechanism, a classifying and throwing mechanism, a storage overflow prompting mechanism and a control unit. The garbage collecting and conveying system comprises a sensing cover opening mechanism, a garbage dispersing and conveying mechanism, a visual identification device, a bag opening mechanism, a classifying and throwing mechanism, a storage overflow prompting mechanism and a prompting information output mechanism, wherein the sensing cover opening mechanism is used for controlling a throwing opening to be automatically opened and closed, the garbage dispersing and conveying mechanism is used for carrying out dispersing conveying on mixed garbage by utilizing a two-stage conveying belt which is parallel up and down and has a speed difference, the visual identification device is used for collecting garbage images and identifying types, the bag opening mechanism is arranged in a tray and used for realizing bag opening separation of bagged garbage through clamping jaws and blades, the classifying and throwing mechanism is used for throwing the garbage into a corresponding garbage can according to an identification result, and the storage overflow prompting mechanism is used for monitoring the state of the garbage can in real time and outputting prompting information. The device can reduce rubbish and pile up and shelter from, improves visual identification stability, realizes automatic opening bag separation and categorised throwing of rubbish in bags.

Inventors

TANG YIPING
ZHANG YUE
Hong Xuezhen
Lu Lanping
HUO ZIYI

Assignees

中国计量大学

Dates

Publication Date: 20260505
Application Date: 20260408

Claims (9)

1. The intelligent garbage classification device based on differential double-layer conveying and visual recognition is characterized by comprising a vertical box body structure formed by an industrial aluminum profile frame, an induction cover opening mechanism, a garbage dispersing and conveying mechanism, a visual recognition device, a bag opening mechanism, a classification throwing mechanism, a storage overflow prompting mechanism and a control unit, wherein the induction cover opening mechanism is arranged at the top of the box body and is used for detecting approaching of a human body and automatically opening or closing a throwing opening, the garbage dispersing and conveying mechanism is arranged below the throwing opening and consists of a first-stage conveying belt and a second-stage conveying belt which are parallel up and down and are connected at the front end, the two-stage conveying belt is driven by an independent stepping motor and has a speed difference and is used for realizing the dispersed conveying of mixed garbage, the visual recognition device is arranged above the tail end of the second-stage conveying belt and is used for collecting garbage images and recognizing garbage types, the bag opening mechanism is integrally arranged inside a tray of the classification throwing mechanism and consists of clamping jaws and blades and is used for carrying out bag opening separation on bagged garbage, the classification mechanism is composed of a tray, a baffle and a cloud platform rotating member and is used for carrying out bag opening recognition on the basis of a recognition result, the garbage dispersing mechanism is arranged below the throwing opening and the throwing mechanism is arranged below the garbage classification opening and is used for controlling the garbage classification unit to carry out, and the classification device is used for carrying a classification device is controlled to carry out classification and is fully, and a classification unit is used for carrying a classification device is controlled.
2. The intelligent garbage classification device based on differential double-layer conveying and visual identification is characterized in that the induction cover opening mechanism comprises a microwave induction radar, a steering engine, a gear, a rack and a cover plate, wherein the microwave induction radar is arranged beside a throwing port and is used for detecting human body approaching signals, the steering engine is in transmission connection with the gear, the gear is meshed with the rack, the rack is fixedly connected with the cover plate, when the human body approaching is detected, the control unit drives the steering engine to open the cover plate, and when garbage throwing is not detected in preset time, the cover plate is automatically closed.
3. The intelligent garbage sorting device based on differential double-layer conveying and visual identification according to claim 1 is characterized in that in the garbage dispersing and conveying mechanism, the running speed of a primary conveying belt is lower than that of a secondary conveying belt, a screw rod is connected to the bottom of the primary conveying belt in a sliding mode, the screw rod is in transmission connection with a stepping motor, the stepping motor drives the screw rod to rotate so as to drive the primary conveying belt to horizontally move, and the intelligent garbage sorting device is used for realizing one-by-one dispersing conveying when multiple garbage are put in, and for moving the primary conveying belt when single garbage does not need dispersing, so that conveying efficiency is improved.
4. The intelligent garbage classification device based on differential double-layer conveying and visual recognition, which is disclosed in claim 1, is characterized in that the visual recognition device consists of an industrial camera and a visual processing unit, wherein an improved EFFICIENTNET _B0 garbage image classification model is arranged in the visual processing unit, the garbage classification model comprises a trunk feature extraction network, a feature enhancement module and a class discrimination module, the feature enhancement module is a weighted convolution enhancement unit, and the response capability of a key region in a feature map is enhanced by introducing a preset weight template into a convolution kernel, so that the separability between garbage classes with similar appearance is improved.
5. The intelligent garbage classification device based on differential double-layer conveying and visual recognition according to claim 4, wherein the garbage image classification model outputs garbage types including four types of recyclable materials, kitchen garbage, harmful garbage and other garbage.
6. The intelligent garbage classification device based on differential double-layer conveying and visual identification according to claim 1 is characterized in that in the classification throwing mechanism, a tray is arranged below the tail end of a secondary conveying belt and used for receiving garbage sent by the garbage dispersing and conveying mechanism, a cradle head rotating piece is in transmission connection with the bottom of the tray and used for driving the tray to rotate or topple in a preset direction, a baffle is arranged on the periphery of the tray and in transmission connection with a steering engine, and the baffle is controlled to open and close by a control unit and used for preventing the garbage from falling off a path in the transferring process.
7. The intelligent garbage classification device based on differential double-layer conveying and visual identification of claim 1 is characterized in that in the bag opening mechanism, clamping jaws are used for clamping and fixing garbage bags, a blade is used for cutting the garbage bags to separate garbage in the bags from plastic garbage bags, separated kitchen garbage falls into a kitchen garbage can, and plastic garbage bags fall into a recyclable garbage can.
8. The intelligent garbage classification device based on differential double-layer conveying and visual recognition according to claim 1 is characterized in that in the storage overflow prompt mechanism, classified garbage cans comprise four types of recyclable materials, kitchen garbage, harmful garbage and other garbage, the full-load detection sensor is an infrared ranging sensor, is arranged near the bin mouths of the classified garbage cans, is electrically connected with the control unit and the display screen, outputs overflow prompt electric signals when the ranging value reaches a preset threshold value, and displays overflow prompt information.
9. The intelligent garbage classification device based on differential double-layer conveying and visual recognition according to claim 1, wherein the control unit is a controller or a main control chip and is electrically connected with the induction cover opening mechanism, the garbage dispersing and conveying mechanism, the visual recognition device, the bag opening mechanism, the classified throwing mechanism and the storage overflow prompting mechanism respectively, and is used for completing the whole flow control of cover opening, dispersed conveying, recognition, bag opening, classified throwing and overflow prompting.

Description

Intelligent garbage classification device based on differential double-layer conveying and visual recognition Technical Field The invention relates to the technical field of intelligent garbage classification, in particular to a novel efficient classified delivery device based on machine vision and automatic control. Background At present, household garbage classification is an important link of resource recovery and environmental protection, but two general problems generally exist in actual throwing, namely, firstly, residents often mix various garbage together and throw the garbage at the same time, so that garbage is stacked and shielded mutually, and the accurate judgment of the garbage classification is affected by a visual recognition device, secondly, kitchen garbage is often mixed with a plastic bag and thrown, the kitchen garbage and the plastic bag are difficult to automatically separate by traditional classification equipment, and kitchen garbage pollution or plastic is mishandled. In view of the above problems, some automatic sorting devices have attempted to adopt a single-stage conveyor belt or a mechanical sorting structure, but it is still difficult to solve the problems of garbage stacking shielding and automatic separation of kitchen garbage bags. Therefore, it is necessary to provide an intelligent garbage classification device capable of effectively dispersing and conveying mixed garbage, automatically opening bags and separating bagged garbage, and accurately recognizing and classifying and putting the mixed garbage so as to solve the problems of unstable recognition effect, difficult separation of the bagged garbage and the like in the prior art. Disclosure of Invention Aiming at the problems that the existing automatic garbage classification device has low recognition rate and is difficult to automatically separate kitchen garbage from plastic bags due to stacking shielding when dealing with mixed garbage and throwing the mixed garbage, the invention provides an intelligent garbage classification device based on differential double-layer conveying and visual recognition. The device is provided with a double-layer conveying structure which is parallel up and down and has a speed difference, so that mixed garbage is scattered and conveyed piece by piece, a bag opening mechanism is integrated in the tray, in-situ separation of bagged kitchen garbage and plastic bags is realized, the recognition capability of garbage with similar appearance is improved through an improved visual recognition model, and therefore, automatic and high-precision classified delivery of garbage is realized. The intelligent garbage classification device based on differential double-layer conveying and visual recognition is characterized by being of a vertical box structure formed by an industrial aluminum profile frame, wherein an induction uncovering mechanism, a garbage dispersing and conveying mechanism, a visual recognition device, a bag opening mechanism, a classification throwing mechanism, a storage overflow prompting mechanism and a control unit are respectively arranged from top to bottom. The mechanisms work cooperatively to realize the full-automatic treatment from throwing, dispersing and conveying, recognition to classification and storage of garbage. The induction cover opening mechanism comprises a throwing opening, a microwave induction radar and a switch cover driving device. The microwave induction radar is arranged beside the throwing port and used for detecting whether a person enters a preset garbage throwing range, the switch cover driving device consists of a steering engine, a gear, a rack and a cover plate, the steering engine is connected with the gear in a transmission mode, the gear is meshed with the rack, the rack is fixedly connected with the cover plate, the throwing port is opened after the switch cover driving device receives an induction signal, and the throwing port is closed when the garbage throwing is not detected within a preset time. The garbage dispersing and conveying mechanism comprises a primary conveying belt, a secondary conveying belt, a stepping motor and a screw rod. The first-stage conveyer belt is located under the throwing opening, and is parallel from top to bottom with the second-stage conveyer belt, and the head end links up, and the second-stage conveyer belt end links up with the tray. The two-stage conveyor belts are independently driven by the stepping motor, have different running speeds and are used for receiving input garbage, and the input mixed garbage is separated one by one and conveyed to the tray through speed difference matching, so that the problem that garbage is mutually blocked is avoided, and meanwhile, the continuity of classification operation is enhanced. The stepping motor is connected with the screw rod in a transmission way, the screw rod is connected with the bottom of the primary conveying belt in a sliding way, the stepping motor