CN107054937A - A kind of refuse classification suggestion device and system based on image recognition - Google Patents
A kind of refuse classification suggestion device and system based on image recognition Download PDFInfo
- Publication number
- CN107054937A CN107054937A CN201710179718.5A CN201710179718A CN107054937A CN 107054937 A CN107054937 A CN 107054937A CN 201710179718 A CN201710179718 A CN 201710179718A CN 107054937 A CN107054937 A CN 107054937A
- Authority
- CN
- China
- Prior art keywords
- picture
- refuse classification
- rubbish
- message handler
- classification
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B65—CONVEYING; PACKING; STORING; HANDLING THIN OR FILAMENTARY MATERIAL
- B65F—GATHERING OR REMOVAL OF DOMESTIC OR LIKE REFUSE
- B65F1/00—Refuse receptacles; Accessories therefor
- B65F1/14—Other constructional features; Accessories
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B65—CONVEYING; PACKING; STORING; HANDLING THIN OR FILAMENTARY MATERIAL
- B65F—GATHERING OR REMOVAL OF DOMESTIC OR LIKE REFUSE
- B65F2210/00—Equipment of refuse receptacles
- B65F2210/138—Identification means
Landscapes
- Engineering & Computer Science (AREA)
- Mechanical Engineering (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Processing Of Solid Wastes (AREA)
Abstract
The invention discloses a kind of refuse classification suggestion device based on image recognition and system, the device includes refuse classification prompting screen, and refuse classification prompting screen includes camera, message handler, display module and power supply module；Camera is used to send the rubbish picture photographed to message handler, the standard form that message handler is used in rubbish picture and its STL prestored carries out rubbish identifying processing, sorting result information is generated, and sorting result information is sent to display module shown；Power supply module is used to be powered for camera, message handler and display module.The advantage of the invention is that quickly carrying out rubbish identification by image recognition technology, help people quickly to carry out refuse classification in daily life, reduce the workload of the secondary sorting of rubbish.
Description
Technical field
Carried the present invention relates to the subsidiary classification identification technology of rubbish, more particularly to a kind of refuse classification based on image recognition
Showing device and system.
Background technology
How to be managed by refuse classification, realize that waste resources are utilized to greatest extent, reduce refuse disposal amount, improve life
Dis environment quality, is one of pressing issues that our times various countries are paid close attention to jointly.The unified standard formulated according to country, it is now raw
Rubbish living is divided into four classes extensively, is Recyclable, kitchen garbage, Harmful Waste and Other Waste respectively.Recyclable is represented
The suitable rubbish reclaimed with the utilization of resources, mainly including waste paper, plastics, glass, metal and the major class of cloth five, is held with blue rubbish
Device is collected, and is recycled by integrated treatment.Kitchen garbage includes the foodstuffs such as leftovers leftovers, bone, dish green tops, pericarp and given up
Thing, is collected, kitchen waste biotechnology treatment in situ compost with green refuse container.Harmful Waste, refer to containing strong to human body
The harmful heavy metal of health, poisonous material or the discarded object that reality harm or potential hazard are caused to environment.Including battery,
Fluorescent tube, bulb, mercurial thermometer, paint kettle, part household electrical appliances, Expired drug, expired cosmetics etc., use red refuse container
Collect.These rubbish need specific safety to handle, and individually reclaim.Other rubbish are to reclaim being difficult in addition to above-mentioned a few class rubbish
Discarded object, such as brick and tile ceramics, dregs, dust, with grey refuse container collect, take the processing mode of sanitary landfills, with
Effectively reduce the pollution to underground water, surface water, soil and air.
But there is the rubbish of some specific types, daily people are difficult to be accurately judged to its classification, such as paper handkerchief and lavatory paper
It is listed in " other rubbish " because water solubility is not too recyclable by force, similar also has pottery, cigarette case；Big stick-means of intimidation bone is because " hardly possible corrosion "
Also it is put into " other rubbish "；And the broken branches and fallen leaves including family are opened fresh flower for losing etc. and belonged to " garbage ".Meanwhile, have one
The affiliated classification of the discarded object each several part of a little various materials compositions is also not quite similar, it is necessary to correspondence classification is abandoned after splitting how
Problem is turned into all types of rubbish Accurate classifications.
To sum up, current most people lack the full knowledge of refuse classification, more can not accurately remember and judge complicated various
Rubbish which kind of type belonged to actually, come correspond to Accurate classification deliver discarding, cause very big pressure to urban environment and garbage disposal
Power.Therefore, how in daily life, it is that this area urgent need to resolve is asked to allow people to carry out fast and accurately classification to rubbish
Topic.
The content of the invention
In order to overcome the deficiencies in the prior art, it is an object of the invention to provide a kind of refuse classification based on image recognition
Suggestion device and system, it can realize the intelligent classification to rubbish.
An object of the present invention is realized using following technical scheme：
A kind of refuse classification suggestion device based on image recognition, including refuse classification prompting screen, refuse classification prompting screen
Including camera, message handler, display module and power supply module；
Camera is used to send the rubbish picture photographed to message handler, and message handler is used for according to rubbish figure
Standard form in piece and its STL prestored carries out rubbish identifying processing, generates sorting result information, and will classification
Object information sends to display module and shown；
Power supply module is used to be powered for camera, message handler and display module.
Preferably, refuse classification prompting screen also includes inductor, inductor is used for the distance for obtaining people, and according to distance
Identification solicited message is sent to message handler；Message handler controls the working condition of camera according to identification solicited message.
Preferably, message handler is additionally operable to control the working condition of display module according to identification solicited message.
Preferably, display module is touch display module, touch display module is additionally operable to be sent according to the operation of user
Operation signal is to message handler, so that message handler controls the working condition of camera according to operation signal.
Preferably, rubbish identifying processing comprises the following steps：
S1, rubbish picture to be identified pre-processed, obtain picture to be split；
S2, treat segmentation picture carry out dividing processing, obtain split picture；
S2 specifically includes following sub-step：
S21, by picture to be split carry out dividing processing, obtain several sub-pictures；
S22, the characteristic information for obtaining sub-pictures, judge whether this feature information belongs in preset threshold range, if
Then the sub-pictures are segmentation picture, otherwise perform S23；
S23, by the sub-pictures be further divided into split picture；
S3, by split picture size be normalized to pre-set dimension；
S4, the characteristic vector for obtaining segmentation picture, by the master die in the characteristic vector and standard feature storehouse of splitting picture
Plate is compared, and the classification of the segmentation picture is judged according to characteristic vector and standard form.
Preferably, obtaining the characteristic vector of segmentation picture in S4, it is specially：It is divided into N*M to divide by picture is split
Picture is cut, the rgb value of pixels all in each sub- segmentation figure piece is obtained, obtains N*M dimensional feature vectors.
Preferably, being specially in S4：The characteristic vector of segmentation picture is obtained, the characteristic vector and standard of picture will be split
Standard form in feature database is compared one by one, judges whether included in the segmentation picture according to characteristic vector and standard form
Adhesion article, if it is the segmentation picture is sub-pictures, and returns to S23, is otherwise judged according to characteristic vector and standard form
The classification of the segmentation picture.
The second object of the present invention is realized using following technical scheme：
A kind of refuse classification prompt system based on image recognition, including refuse classification suggestion device, remote management services
Device and management end, refuse classification prompting screen also include communication module；Message handler passes through communication module and remote management services
Device is connected；Management end is used to send default display information to remotely administered server, so that remote server sends default display
Information is to message handler；Default display information is sent to display module and shown by message handler.
A kind of refuse classification prompt system based on image recognition, including refuse classification suggestion device, remote management services
Device and management end, refuse classification prompting screen also include communication module；Message handler passes through communication module and remote management services
Device is connected；Management end is used to send equipment mount message to remotely administered server, so that remotely administered server is according to equipment
STL in mount message configuration information processor.
A kind of refuse classification prompt system based on image recognition, including refuse classification suggestion device and remote management services
Device, refuse classification prompting screen also includes communication module；Message handler is connected by communication module and remotely administered server；Letter
The rubbish picture that breath processor is additionally operable to will be unable to identification is sent to remotely administered server, so that remotely administered server updates
Total class template storehouse in remotely administered server.
Compared with prior art, the beneficial effects of the present invention are：Rubbish identification is quickly carried out by image recognition technology simultaneously
The prompting of classification belonging to rubbish is made, helps people are daily quickly and easily to carry out refuse classification processing, reduces secondary point of rubbish
The workload picked, allows people easily to differentiate rubbish generic and alignment processing in refuse classification, to realizing that rubbish is provided
Source is recycled, and reduces refuse disposal amount, improves living environment quality, is reduced environmental pollution and is played help.
Brief description of the drawings
Fig. 1 is the front view of the refuse classification prompting screen of the present invention；
Fig. 2 is the rearview of the refuse classification prompting screen of the present invention；
Fig. 3 is the side view of the refuse classification prompting screen of the present invention；
Fig. 4 is the module connection figure of the refuse classification prompting screen of the present invention；
Fig. 5 is the flow chart of the rubbish recognition methods of the present invention.
In figure：01st, identification marking；02nd, camera；03rd, display module；04th, inductor；05th, billboard；06th, solar energy
Plate.
Embodiment
Below, with reference to accompanying drawing and embodiment, the present invention is described further：
Everybody subclass of the various rubbish of correspondence carries out refuse classification processing for convenience, reduces the pressure for the treatment of of urban garbage
Power, can come into operation situation the invention provides one kind according to existing sorting rubbish, in backstage correspondence configuration prompt taxonomic species
Class, being capable of the affiliated refuse classification automatic identification electronic cue screen classified and carry out classified prompt of automatic identification rubbish.The rubbish
Classification automatic identification electronic cue screen is matched somebody with somebody by remote back-office to be postponed, the classification garbage can class categories come into operation with having disposed
Correspond, built-in identification module is quickly carried out rubbish identification using image recognition technology and makes pair classified belonging to rubbish
Classification garbage can is answered to deliver prompting.It is of the invention to can be very good to combine with the existing classification garbage can come into operation, reduce and throw
Enter cost, it is to avoid the input of secondary repeated construction is wasted, upgrading is combined with existing classification garbage can, help the daily quick side of people
Just carry out refuse classification processing, reduces the workload of the secondary sorting of rubbish, people is easily differentiated in refuse classification
Rubbish generic and alignment processing, to realizing that waste resources are recycled, reduce refuse disposal amount, improve living environment quality,
Reduce environmental pollution and play help.
A kind of refuse classification prompt system based on image recognition, including refuse classification prompting screen, remotely administered server
And management end, refuse classification prompting screen as shown in Figures 1 to 3, including framework, identification marking 01, camera 02, message handler,
Display module 03, inductor 04, communication module, power supply module and billboard 05；Power supply module includes solar panels 06, charging control
Device and battery processed；
Display module 03 is located at the one side of framework, and billboard 05 is located at the another side of framework；Solar panels 06 are located at framework
Top.Identification marking 01 is on framework, in the present embodiment, and identification marking 01 is Quick Response Code, and every prompting screen is provided with
Unique encodings, correspondence is arranged on framework using coding to be easy at scanning.When the coding is used to dispose prompting screen with dustbin and
Deployment place correspondence is stored in remotely administered server, for the dustbin correspondence configuration equipment disposed according to dispensing mount point
Identification classification and advertisement putting management.
As shown in figure 4, camera 02, display module 03 and inductor 04 are connected with message handler；Message handler
Connected by communication module and remotely administered server.
Camera 02 is located at the one side of framework, and positioned at the lower section of display module 03；Camera 02 is used to obtain its shooting
In the range of pictorial information or video information；In the present embodiment, camera 02 is used to obtain rubbish picture, and by rubbish picture
Send to message handler.Inductor 04 is used for the distance for perceiving and judging people, and please according to the transmission identification of the range information of people
Information is sought to message handler；Inductor 04 can be infrared inductor 04, can also be microwave sensor 04, inductor 04
Embedded device framework one side, and close to camera 02 position.Inductor 04 can also be button, by user press by
Button, sends identification solicited message to message handler.
Message handler is used for the working condition that camera 02 is controlled according to identification solicited message；For example in identification request letter
When breath is less than the first pre-determined distance, control camera 02, which shoot, obtains rubbish picture, and camera 02 sends rubbish picture
To message handler.
Message handler is additionally operable to control the working condition of display module 03 according to identification solicited message；For example, in identification
When solicited message is less than the second pre-determined distance, control display module 03 is operated, the default display information of display.
Message handler connects internet by communication module, and is interacted by internet and remotely administered server connection
Information；Administrative staff are connected by management end and remotely administered server.Administrative staff can be sent default aobvious by management end
Show information to remotely administered server, so that remotely administered server sends default display information to message handler, information
Default display information is sent to display module 03 and shown by processor.Default display information can be public service ads information etc.
Content.Communication module can be WIFI module or GPRS module.
The identification module of message handler is used to be carried out judging generation sorting result information according to rubbish picture, and will classification
Object information sends to display module 03 and shown.Display module 03 can also be several display lamps, each display lamp pair
The position answered is identified provided with corresponding dustbin；For example, the corresponding position of a display lamp is identified provided with recyclable rubbish, it is another
The corresponding position of individual display lamp is identified provided with not recyclable rubbish, and the 3rd corresponding position of display lamp is provided with Harmful Waste mark
Know；The sorting result information that message handler is generated according to rubbish picture, and corresponding display is controlled according to sorting result information
Lamp carry out bright light, with point out user it is corresponding be what classification rubbish.Specifically, rubbish picture is pre-processed, feature
Extract, the standard form in the STL then prestored with it compares, and the affiliated classification of rubbish picture is judged, if treated
Identification rubbish is made up of multiple different affiliated classified parts, then each different classifications region is marked respectively, judges each
Partial affiliated classification, while each zone marker of correspondence, and sorting result information is sent to the classified prompt of message handler
Module；Sorting result information includes zone marker information and classification judged result.Message handler will be unable to by communication module
The item pictures of identification are sent to remotely administered server backup storage, and remotely administered server is periodically learned this kind of picture
Total class template storehouse is updated to again after practising training.
Classified prompt module is used to send sorting result information to display module 03 to be shown；Preferably, rubbish
Classified prompt screen also includes loudspeaker；Classified prompt module is additionally operable to send sorting result information to loudspeaker progress voice and carried
Show；The display and voice message of sorting result information are pre-set according to each classification.Such as Recyclable using blue color lump and
Word and mark simultaneously mix Recyclable icon and correspondence voice message content；Kitchen garbage green color lump and word marking are matched somebody with somebody
Kitchen garbage icon and correspondence voice message content；Harmful Waste are using red color lump and word marking, with Harmful Waste icon
With correspondence voice message content；Other Waste is carried using gray patches and word marking, with Other Waste icon and correspondence voice
Show content.So mark and the refuse classification of current main flow and mark is unified, is easy to user's correspondence Classification and Identification, reduce study with
Remember cost.
Remotely administered server is used to managing, configuring the refuse classification automatic identification electronic cue screen installed throughout, right
Electronic cue screen carries out configuration Classification Management, advertisement putting management.When equipment is installed, obtain administrative staff and pass through in management end
Corresponding classification garbage can picture, equipment Quick Response Code and the infield of the installation refuse classification automatic identification electronic cue screen of biography
Information, and the said equipment mount message is stored one by one.Management end is managed for prompting screen, in installation suggestion screen, will pacify
The details of dustbin photo, the Quick Response Code on prompting screen and infield that dress place has been set are sent to remote management
Server；The dustbin photo set needs to include obvious classification garbage can all angles comprehensively, so that remote management takes
Business device can recognize that the corresponding color of each classification garbage can, mark or corresponding specific name, in order to remotely administered server
The refuse classification that correspondence configuration needs after identification, carries out the STL and identification classification in configuration information processor.
In the present embodiment, management end is divided into configuration management end and display management end；Administrative staff pass through configuration management end
Equipment mount message is sent to remotely administered server；And display information is preset to remote by display management end transmission/configuration
Thread management server.But may have the new standard form regularly updated in total class template storehouse, information processing can be updated to
In STL in device.
In configuration standard ATL, remotely administered server is classified according to the conventional garbage of the classification garbage can set
Photo, recognizes the correspondence classification of dustbin, then configures the master die in the message handler of each refuse classification prompting screen
The classification of classification in plate storehouse, i.e. each refuse classification prompting screen correspondence prompting.Such as the classification garbage can installed, it can only return
Receipts, not recyclable two class, then only classified prompt is carried out when automatically configuring according to the two major classes；Such as the classification garbage can installed
Subclass it is more, then be configured according to existing subclass, accordingly, the classification that can classify during prompting is also thinner more.Together
When, remotely administered server and refuse classification automatic identification electronic cue screen are by network connection, periodically by the standard trained
STL of the template renewal into message handler, and obtain the rubbish for the None- identified that the camera 02 of prompting screen is photographed
Rubbish item pictures, so that further total class template storehouse of upgrading remote management server., can basis when advertisement putting is managed
The numbering of equipment and corresponding installation region carry out subregion advertisement putting.
It should be noted that there is all rubbish classes involved by the system in total class template storehouse in remotely administered server
Other classified catalogue；The catalogue in STL in message handler also includes the classification of involved all rubbish classifications
Catalogue, but remotely administered server carries out configuration identification classification according to equipment mount message, may be only several in all catalogues
Class catalogue is called to be identified, it is also possible to reminded according to the classification configurations that the situation of actual classification dustbin carries out group.
Display module 03 is touch display module, and touch display module is additionally operable to the operation according to user, sends operation letter
Number to message handler, message handler controls the working condition of camera 02 according to operation signal.For example, user is recognizing
After one rubbish, in addition it is also necessary to carry out second and recognize, the control signal again identified that can be sent by touch display module
To message handler, message handler control camera 02 is shot again.
Solar panels 06 are connected by charge controller and battery；Power supply module is camera 02, display module 03, sense
Device 04 and message handler is answered to be powered, its operation principle is prior art, be will not be repeated here.Solar panels 06 are located at frame
The top of frame, it is triangular in shape to be fixedly installed.
Billboard 05 is located at the another side of framework, can be used for showing that periodically delivering the open air updated puts up billboard,
It can be used for the benefit for publicizing refuse classification.
Remotely administered server sets up refuse classification catalogue, is obtained according to catalogue by search engine under a large amount of correspondence catalogues
Waste items picture pre-processed, feature extraction, recognition training, so as to set up total class template storehouse.Such as：Rubbish is divided into four
Major class, be respectively：Recyclable, kitchen garbage, Harmful Waste and Other Waste.The article included during Recyclable is classified is such as
Article involved by carton, beverage bottle, vial, the pop can for filling beer etc. obtains corresponding diagram by search engine
Piece, the picture of these articles is pre-processed, feature extraction, recognition training is so as to setting up the total of corresponding Recyclable identification
Class model storehouse；Accordingly, the article lower correspondence of each classification of kitchen garbage, Harmful Waste and Other Waste included passes through same
Mode sets up total class template storehouse.
Installation personnel opens the management end APP on mobile phone when installing refuse classification prompting screen, and starting device installs management
Quick Response Code, shooting in system, scanning device correspond to the arranged classification garbage can photo of mount point, submit the ground of mount point
Positional information is managed, the said equipment mount message is transferred to remotely administered server, and completes equipment and installs, enable.Long-range pipe
Reason server receives the equipment mount message of management end APP transmission, and correspondence is gone out according to the classification garbage can photo array of shooting
The classification of rubbish, carries out the correspondence configuration of message handler, and corresponding device encodes the corresponding infield of record storage equipment
Information, while sending default display information to the message handler of mounted prompting screen.
When user prepares that the waste items of discarding are carried out putting on by classification identification, close to prompting screen, message handler connects
Receive inductor 04 and sense that someone enters the identification solicited message of the distance of reaction scope of setting, prompting screen stops advertisement and broadcast
The identification module of refuse classification is put and starts, into identification process.
Display module 03 by voice and picture point out user by waste items to be identified close to camera 02 in order to
Camera 02 can clearly get the image of waste items；Obtained simultaneously in the middle section displaying camera 02 of display module 03
The image content got.The position of user's adjustment waste items and the distance with camera 02 are waited, camera 02, which is captured, to be treated
After the picture rich in detail of the waste items of identification, identification module is identified；
Identification module gets waste items picture to be identified, picture is pre-processed, after feature extraction with system
Each grouped data in the refuse classification model library prestored compares, and judges the affiliated classification of rubbish to be identified.If to be identified
Rubbish is made up of some, belongs to different classifications, then is corresponded to after each zone marker, affiliated point that each several part is judged respectively
Class, meanwhile, sorting result information is sent to classified prompt module.
The sorting result information that classified prompt module transmits identification module, according to prompt rules in rubbish picture to be identified
On mark one by one, and shown on display module 03, can also by the form of voice combination word, color and icon,
Point out user that refuse classification is delivered to the classification garbage can to side by the classification marked.For example, " blue region part is can
Rubbish is reclaimed, green area part is rubbish from cooking, please puts on by classification rubbish by cue mark, thank you to environmentally friendly support ".
Complete rubbish identification classified prompt process is completed once above.Button, Yong Huke are re-recognized in screen lower right simultaneously
To select to re-recognize the waste items for starting to recognize that other need to classify.Such as user operates off without any, inductor
04 senses that user leaves, and feeds back information to after message handler, and message handler is then switched to advertisement module control display
The broadcasting advertisement content of module 03.
Specifically, as shown in figure 5, message handler is used to be carried out judging generation sorting result information according to rubbish picture
Rubbish identifying processing is comprised the following specific steps that：
Step 101, acquisition rubbish picture to be identified；
Step 102, rubbish picture to be identified is pre-processed, obtain picture to be split；
Specifically, pretreatment includes following sub-step：Rubbish picture is decoded, noise reduction and removal noise spot are treated
Split picture.
Rubbish picture is filtered and removes the noise that pixel is highly 1 and 2, so that follow-up processing can be more preferable
Carry out.The process that filtering removes noise spot is prior art, be will not be repeated here.
Step 103, treat segmentation picture carry out dividing processing, obtain split picture；
Dividing processing specifically includes following sub-step：
Step 31, by seed fill algorithm picture to be split is split, obtain several sub-pictures；
Some of which sub-pictures only include a kind of article wherein, as split picture；But have in some sub-pictures
Potentially include more than one article wherein, i.e. article adhesion then needs to carry out these sub-pictures further segmentation
Final segmentation picture can be obtained.
Step 32, the characteristic information for obtaining sub-pictures, judge whether this feature information belongs in preset threshold range, if
It is that the sub-pictures are segmentation picture, otherwise performs step 33.
Characteristic information is always counted including article in sub-pictures, article the ratio of width to height, contoured article, article transparency, article face
Color, article texture and article pattern etc., the total points of article are the quantity of the pixel of article in the sub-pictures, and the ratio of width to height is the thing
The ratio of width to height in the region of product.Picture feature difference with specific reference to every kind of rubbish carries out different characteristic informations judgements.
Such as：The article of each type, there is some various sizes of Aspect Ratio numerical value, but all can be in a scope
It is interior to float, corresponding threshold values is set up according to the length-width ratio sliding scales of this kind of article, if not in preset threshold range, said
This bright object part is blocked, and length or width are inadequate, so deviate from predetermined threshold value.
Predetermined threshold value is according to obtained by the characteristic information statistical analysis of each rubbish of article.
It should be noted that the different characteristic information in order to obtain sub-pictures, is located in advance to rubbish picture to be identified
The processing of reason is different, such as the characteristic information for obtaining contoured article, and preprocessing process also includes rim detection；Specifically
Different characteristic acquisition of information can know from prior art with preprocessing process, will not be repeated here.
Step 33, the vertical projection diagram for obtaining the sub-pictures, the method by finding valley point in vertical projection diagram, are obtained
The position of article adhesion, and sub-pictures are further divided into by segmentation picture according to this.
Step 104, by barycenter align and linear interpolation amplification by split picture size be normalized to pre-set dimension, with
In anti-picture article existence position skew, it is not of uniform size and rotate it is indefinite the problem of.
Step 105, the characteristic vector for obtaining segmentation picture；Specially：Segmentation picture is divided into N*M segmentation figure
Piece, obtains the rgb value of pixels all in each sub- segmentation figure piece, obtains N*M dimensional feature vectors, and N and M are present count
Word, each characteristic vector includes the RBG values of pixels all in corresponding sub- segmentation picture.
Step 106, by split picture characteristic vector and STL in standard form compared one by one, according to
Characteristic vector and standard form judge the classification of the segmentation picture, and generate sorting result information.
There is the standard form of various rubbish classifications in STL, each rubbish classification there are hundreds of master dies
Plate, after pretreatment, dividing processing and feature extraction, the characteristic vector of training set sorting rubbish article is stored in file.
, it is necessary to indicate the right value of each sorting rubbish article during training.The standard form of each rubbish is all the standard of Individual Items
Characteristic vector, when setting up STL, the waste items picture of adhesion is added without STL.
Netscape messaging server Netscape sets up refuse classification catalogue first, and a large amount of correspondence mesh are obtained by search engine according to catalogue
The lower rubbish picture of record is pre-processed, dividing processing, normalization and acquisition characteristic vector, so as to set up what refuse classification was recognized
STL.Such as：Rubbish is divided into 4 major classes, is respectively：Recyclable, kitchen garbage, Harmful Waste and Other Waste.
The institute such as the article included during Recyclable is classified such as carton, beverage bottle, vial, the pop can for filling beer
The article being related to obtains correspondence picture by search engine, the picture of these articles is pre-processed, dividing processing, normalizing
Change and obtain characteristic vector to set up the STL of corresponding Recyclable identification；
Accordingly, the article lower correspondence of each classification of kitchen garbage, Harmful Waste and Other Waste included passes through same
Mode sets up the STL of identification.
Get rubbish picture to be identified, picture is pre-processed, after dividing processing and feature extraction with it is pre- in system
Each grouped data in the STL for the refuse classification deposited compares, and judges the affiliated classification of rubbish to be identified.If treated
Identification rubbish is made up of some, belongs to different classifications, then corresponds to after each zone marker, the affiliated of each several part is judged respectively
Classification, meanwhile, above-mentioned zone label information, classification judged result are sent to refuse classification reminding module.
The rubbish that classified prompt module transmits rubbish identification module corresponds to each zone marker, sorting result information, according to
Prompt rules mark and show user one by one on rubbish picture to be identified, use voice combination word, color and icon
Form, points out user to carry out refuse classification dispensing processing by the classification marked.With reference to voice message, " blue region part is can
Rubbish is reclaimed, green area part is rubbish from cooking, is please classified by cue mark, thank you to environmentally friendly support ".
Further, it is preliminary in step 32 to judge sub-pictures to split after picture, in order to prevent misjudgment, step 106
In also include judging in the segmentation picture whether to include adhesion article according to characteristic vector and standard form, if it is return to step
Rapid 33, the segmentation picture is split, the classification of the segmentation picture is otherwise judged according to characteristic vector and standard form.
It will be apparent to those skilled in the art that technical scheme that can be as described above and design, make other various
It is corresponding to change and deformation, and all these change and deformation should all belong to the protection domain of the claims in the present invention
Within.
Claims (10)
1. a kind of refuse classification suggestion device based on image recognition, it is characterised in that including refuse classification prompting screen, rubbish point
Class prompting screen includes camera, message handler, display module and power supply module；
Camera is used to send the rubbish picture photographed to message handler, message handler be used for according to rubbish picture and
Standard form in its STL prestored carries out rubbish identifying processing, generates sorting result information, and by classification results
Information sends to display module and shown；
Power supply module is used to be powered for camera, message handler and display module.
2. refuse classification suggestion device as claimed in claim 1, it is characterised in that refuse classification prompting screen also includes sensing
Device, inductor is used for the distance for obtaining people, and is sent according to distance transmission identification solicited message to message handler；Information processing
Device controls the working condition of camera according to identification solicited message.
3. refuse classification suggestion device as claimed in claim 2, it is characterised in that message handler is additionally operable to please according to identification
Information is asked to control the working condition of display module.
4. refuse classification suggestion device as claimed in claim 1, it is characterised in that display module is touch display module, is touched
Touch display module to be additionally operable to send operation signal to message handler according to the operation of user, so that message handler is according to operation
Signal controls the working condition of camera.
5. refuse classification suggestion device as claimed in claim 1, it is characterised in that rubbish identifying processing comprises the following steps：
S1, rubbish picture to be identified pre-processed, obtain picture to be split；
S2, treat segmentation picture carry out dividing processing, obtain split picture；
S2 specifically includes following sub-step：
S21, by picture to be split carry out dividing processing, obtain several sub-pictures；
S22, the characteristic information for obtaining sub-pictures, judge whether this feature information belongs in preset threshold range, if it is should
Sub-pictures are segmentation picture, otherwise perform S23；
S23, by the sub-pictures be further divided into split picture；
S3, by split picture size be normalized to pre-set dimension；
S4, the characteristic vector for obtaining segmentation picture, the standard form in the characteristic vector and standard feature storehouse of splitting picture is entered
Row is compared, and the classification of the segmentation picture is judged according to characteristic vector and standard form.
6. refuse classification suggestion device as claimed in claim 5, it is characterised in that obtained in S4 the feature of segmentation picture to
Amount, be specially：It is divided into N*M to split picture by picture is split, obtains pixels all in each sub- segmentation figure piece
Rgb value, obtain N*M dimensional feature vectors.
7. refuse classification suggestion device as claimed in claim 5, it is characterised in that be specially in S4：Obtain segmentation picture
Characteristic vector, by split picture characteristic vector and standard feature storehouse in standard form compared one by one, according to feature to
Amount and standard form judge whether include adhesion article in the segmentation picture, and if it is the segmentation picture is sub-pictures, and is returned
S23 is returned, the classification of the segmentation picture is otherwise judged according to characteristic vector and standard form.
8. a kind of refuse classification prompt system based on image recognition, it is characterised in that including any one of such as claim 1 to 7
Described refuse classification suggestion device, remotely administered server and management end, refuse classification prompting screen also includes communication module；Letter
Processor is ceased to connect by communication module and remotely administered server；
Management end is used to send default display information to remotely administered server, so that remote server sends default display information
To message handler；Default display information is sent to display module and shown by message handler.
9. a kind of refuse classification prompt system based on image recognition, it is characterised in that including any one of such as claim 1 to 7
Described refuse classification suggestion device, remotely administered server and management end, refuse classification prompting screen also includes communication module；Letter
Processor is ceased to connect by communication module and remotely administered server；
Management end is used to send equipment mount message to remotely administered server, so that remotely administered server is installed according to equipment
STL in information configuration message handler.
10. a kind of refuse classification prompt system based on image recognition, it is characterised in that including any one of such as claim 1 to 7
Described refuse classification suggestion device and remotely administered server, refuse classification prompting screen also include communication module；Information processing
Device is connected by communication module and remotely administered server；
The rubbish picture that message handler is used to will be unable to identification is sent to remotely administered server, so that remotely administered server
Update total class template storehouse in remotely administered server.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710179718.5A CN107054937A (en) | 2017-03-23 | 2017-03-23 | A kind of refuse classification suggestion device and system based on image recognition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710179718.5A CN107054937A (en) | 2017-03-23 | 2017-03-23 | A kind of refuse classification suggestion device and system based on image recognition |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107054937A true CN107054937A (en) | 2017-08-18 |
Family
ID=59620389
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710179718.5A Pending CN107054937A (en) | 2017-03-23 | 2017-03-23 | A kind of refuse classification suggestion device and system based on image recognition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107054937A (en) |
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107909072A (en) * | 2017-09-29 | 2018-04-13 | 广东数相智能科技有限公司 | A kind of vegetation type recognition methods, electronic equipment, storage medium and device |
| CN108171255A (en) * | 2017-11-22 | 2018-06-15 | 广东数相智能科技有限公司 | Picture association intensity ratings method and device based on image identification |
| CN108910340A (en) * | 2018-05-25 | 2018-11-30 | 四川斐讯信息技术有限公司 | A kind of control method and system of intelligent garbage bin |
| CN109898265A (en) * | 2017-12-08 | 2019-06-18 | 青岛海尔洗衣机有限公司 | A kind of detergent additive put-on method and washing machine |
| CN110428065A (en) * | 2019-07-25 | 2019-11-08 | 北京思迪环保科技服务有限公司 | A kind of fast receipts cloud management plateform system |
| CN110525821A (en) * | 2019-08-20 | 2019-12-03 | 北京精英系统科技有限公司 | A kind of system promoting separate waste collection site operation efficiency |
| CN110569874A (en) * | 2019-08-05 | 2019-12-13 | 深圳大学 | Garbage classification method and device, intelligent terminal and storage medium |
| CN110626662A (en) * | 2019-10-12 | 2019-12-31 | 张颢宸 | Image recognition-based garbage self-classification method and device |
| CN110647833A (en) * | 2019-09-18 | 2020-01-03 | 上海合合信息科技发展有限公司 | Garbage classification judgment device and method |
| CN111086797A (en) * | 2019-11-25 | 2020-05-01 | 徐州立方机电设备制造有限公司 | Dustbin with recognition function |
| CN111086796A (en) * | 2019-11-05 | 2020-05-01 | 广东拜登网络技术有限公司 | Perfecting method for garbage type identification and background server |
| CN111178329A (en) * | 2020-01-17 | 2020-05-19 | 小圾(上海)环保科技有限公司 | Big data collection method for garbage image |
| CN111217055A (en) * | 2018-11-27 | 2020-06-02 | 深圳利万联科技有限公司 | Garbage putting supervision method, device, server and system |
| CN111242236A (en) * | 2020-01-19 | 2020-06-05 | 北京声智科技有限公司 | Rubbish type identification method, server, client, system and medium |
| CN111241967A (en) * | 2020-01-06 | 2020-06-05 | 珠海格力电器股份有限公司 | Garbage classification method, system, terminal equipment and storage medium |
| CN111237765A (en) * | 2019-11-28 | 2020-06-05 | 武汉九象云大数据科技有限公司 | Garbage pretreatment system and method for garbage proportioning based on information technology |
| CN111400583A (en) * | 2020-03-11 | 2020-07-10 | 北京字节跳动网络技术有限公司 | Information searching method and device, electronic equipment and storage medium |
| CN111559586A (en) * | 2020-04-29 | 2020-08-21 | 南京信息职业技术学院 | Household intelligent garbage classification and identification system and method |
| CN112364758A (en) * | 2020-11-10 | 2021-02-12 | 湖北惠立网络科技有限公司 | Garbage classification recovery method and system based on multi-target image recognition |
| CN112767196A (en) * | 2021-01-15 | 2021-05-07 | 森罗云科技(广州)有限公司 | Intelligent property management information sharing platform for smart community |
| CN112837491A (en) * | 2020-12-31 | 2021-05-25 | 苏州农业职业技术学院 | Residential quarter is with intelligent collection system of rubbish from cooking |
| CN112866647A (en) * | 2021-01-15 | 2021-05-28 | 森罗云科技(广州)有限公司 | Intelligent property management system based on smart community |
| CN113936251A (en) * | 2021-11-01 | 2022-01-14 | 深圳绿达环境工程技术有限公司 | Image recognition's categorised intelligent supervisory equipment of rubbish |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101193261A (en) * | 2007-03-28 | 2008-06-04 | 腾讯科技（深圳）有限公司 | Video communication system and method |
| CN101571875A (en) * | 2009-05-05 | 2009-11-04 | 程治永 | Realization method of image searching system based on image recognition |
| CN101788497A (en) * | 2009-12-30 | 2010-07-28 | 深圳先进技术研究院 | Embedded bean classifying system based on image recognition technology |
| CN102799854A (en) * | 2011-05-23 | 2012-11-28 | 株式会社摩如富 | Image identification device and image identification method |
| CN103143509A (en) * | 2013-03-21 | 2013-06-12 | 电子科技大学 | Garbage classification robot and garbage identification and classification method |
| US20140025589A1 (en) * | 2012-07-20 | 2014-01-23 | Chen Yang | Waste compression device with fee calculation function |
| CN204938038U (en) * | 2015-09-25 | 2016-01-06 | 蒋曙 | A kind of refuse classification equipment |
| CN205289011U (en) * | 2015-12-17 | 2016-06-08 | 赵凌宇 | Building rubbish sorting device based on machine vision |
| CN105730930A (en) * | 2016-02-04 | 2016-07-06 | 杭州睿生环保科技有限公司 | Intelligent environment-friendly garbage classifying and recycling system and method for recycling and classifying garbage by using same |
| WO2016157825A1 (en) * | 2015-03-30 | 2016-10-06 | 日本電気株式会社 | Monitoring system, monitored device, control method, and recording medium |
| CN106241106A (en) * | 2016-08-31 | 2016-12-21 | 合肥工业大学 | A kind of refuse receptacle |
| CN106275945A (en) * | 2016-08-17 | 2017-01-04 | 浪潮电子信息产业股份有限公司 | A kind of garbage recovery device, garbage classification system and using method |
| CN106395186A (en) * | 2016-12-05 | 2017-02-15 | 上海商溱环保节能科技有限公司 | Intelligent urban waste bin |
-
2017
- 2017-03-23 CN CN201710179718.5A patent/CN107054937A/en active Pending
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101193261A (en) * | 2007-03-28 | 2008-06-04 | 腾讯科技（深圳）有限公司 | Video communication system and method |
| CN101571875A (en) * | 2009-05-05 | 2009-11-04 | 程治永 | Realization method of image searching system based on image recognition |
| CN101788497A (en) * | 2009-12-30 | 2010-07-28 | 深圳先进技术研究院 | Embedded bean classifying system based on image recognition technology |
| CN102799854A (en) * | 2011-05-23 | 2012-11-28 | 株式会社摩如富 | Image identification device and image identification method |
| US20140025589A1 (en) * | 2012-07-20 | 2014-01-23 | Chen Yang | Waste compression device with fee calculation function |
| CN103143509A (en) * | 2013-03-21 | 2013-06-12 | 电子科技大学 | Garbage classification robot and garbage identification and classification method |
| WO2016157825A1 (en) * | 2015-03-30 | 2016-10-06 | 日本電気株式会社 | Monitoring system, monitored device, control method, and recording medium |
| CN204938038U (en) * | 2015-09-25 | 2016-01-06 | 蒋曙 | A kind of refuse classification equipment |
| CN205289011U (en) * | 2015-12-17 | 2016-06-08 | 赵凌宇 | Building rubbish sorting device based on machine vision |
| CN105730930A (en) * | 2016-02-04 | 2016-07-06 | 杭州睿生环保科技有限公司 | Intelligent environment-friendly garbage classifying and recycling system and method for recycling and classifying garbage by using same |
| CN106275945A (en) * | 2016-08-17 | 2017-01-04 | 浪潮电子信息产业股份有限公司 | A kind of garbage recovery device, garbage classification system and using method |
| CN106241106A (en) * | 2016-08-31 | 2016-12-21 | 合肥工业大学 | A kind of refuse receptacle |
| CN106395186A (en) * | 2016-12-05 | 2017-02-15 | 上海商溱环保节能科技有限公司 | Intelligent urban waste bin |
Cited By (25)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107909072A (en) * | 2017-09-29 | 2018-04-13 | 广东数相智能科技有限公司 | A kind of vegetation type recognition methods, electronic equipment, storage medium and device |
| CN108171255A (en) * | 2017-11-22 | 2018-06-15 | 广东数相智能科技有限公司 | Picture association intensity ratings method and device based on image identification |
| CN109898265A (en) * | 2017-12-08 | 2019-06-18 | 青岛海尔洗衣机有限公司 | A kind of detergent additive put-on method and washing machine |
| CN108910340A (en) * | 2018-05-25 | 2018-11-30 | 四川斐讯信息技术有限公司 | A kind of control method and system of intelligent garbage bin |
| CN111217055A (en) * | 2018-11-27 | 2020-06-02 | 深圳利万联科技有限公司 | Garbage putting supervision method, device, server and system |
| CN110428065A (en) * | 2019-07-25 | 2019-11-08 | 北京思迪环保科技服务有限公司 | A kind of fast receipts cloud management plateform system |
| CN110569874A (en) * | 2019-08-05 | 2019-12-13 | 深圳大学 | Garbage classification method and device, intelligent terminal and storage medium |
| CN110525821A (en) * | 2019-08-20 | 2019-12-03 | 北京精英系统科技有限公司 | A kind of system promoting separate waste collection site operation efficiency |
| CN110647833A (en) * | 2019-09-18 | 2020-01-03 | 上海合合信息科技发展有限公司 | Garbage classification judgment device and method |
| CN110626662A (en) * | 2019-10-12 | 2019-12-31 | 张颢宸 | Image recognition-based garbage self-classification method and device |
| CN111086796A (en) * | 2019-11-05 | 2020-05-01 | 广东拜登网络技术有限公司 | Perfecting method for garbage type identification and background server |
| CN111086797A (en) * | 2019-11-25 | 2020-05-01 | 徐州立方机电设备制造有限公司 | Dustbin with recognition function |
| CN111237765A (en) * | 2019-11-28 | 2020-06-05 | 武汉九象云大数据科技有限公司 | Garbage pretreatment system and method for garbage proportioning based on information technology |
| CN111241967B (en) * | 2020-01-06 | 2023-08-15 | 珠海格力电器股份有限公司 | Garbage classification method, garbage classification device, terminal equipment and storage medium |
| CN111241967A (en) * | 2020-01-06 | 2020-06-05 | 珠海格力电器股份有限公司 | Garbage classification method, system, terminal equipment and storage medium |
| CN111178329A (en) * | 2020-01-17 | 2020-05-19 | 小圾(上海)环保科技有限公司 | Big data collection method for garbage image |
| CN111178329B (en) * | 2020-01-17 | 2023-05-26 | 小圾(上海)环保科技有限公司 | Big data collection method of garbage image |
| CN111242236A (en) * | 2020-01-19 | 2020-06-05 | 北京声智科技有限公司 | Rubbish type identification method, server, client, system and medium |
| CN111400583A (en) * | 2020-03-11 | 2020-07-10 | 北京字节跳动网络技术有限公司 | Information searching method and device, electronic equipment and storage medium |
| CN111559586A (en) * | 2020-04-29 | 2020-08-21 | 南京信息职业技术学院 | Household intelligent garbage classification and identification system and method |
| CN112364758A (en) * | 2020-11-10 | 2021-02-12 | 湖北惠立网络科技有限公司 | Garbage classification recovery method and system based on multi-target image recognition |
| CN112837491A (en) * | 2020-12-31 | 2021-05-25 | 苏州农业职业技术学院 | Residential quarter is with intelligent collection system of rubbish from cooking |
| CN112767196A (en) * | 2021-01-15 | 2021-05-07 | 森罗云科技(广州)有限公司 | Intelligent property management information sharing platform for smart community |
| CN112866647A (en) * | 2021-01-15 | 2021-05-28 | 森罗云科技(广州)有限公司 | Intelligent property management system based on smart community |
| CN113936251A (en) * | 2021-11-01 | 2022-01-14 | 深圳绿达环境工程技术有限公司 | Image recognition's categorised intelligent supervisory equipment of rubbish |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107054937A (en) | A kind of refuse classification suggestion device and system based on image recognition | |
| CN107054936A (en) | A kind of refuse classification prompting dustbin and system based on image recognition | |
| CN107092914A (en) | Refuse classification method, device and system based on image recognition | |
| CN105151594B (en) | A kind of rubbish recycling management system based on internet, management method and dustbin | |
| CN110155572B (en) | Intelligent community garbage classification system and method | |
| CN207671068U (en) | Intelligent garbage classification retracting device and system based on custom sex education | |
| CN105173477B (en) | A kind of dustbin and the separate waste collection system and method based on the network platform | |
| CN106295965B (en) | Internet-based garbage classification recycling method and system and garbage classification terminal | |
| CN109684979A (en) | A kind of refuse classification method based on image recognition technology, device and electronic equipment | |
| CN110697273A (en) | Intelligent household garbage identification and automatic classification system and method based on iterative learning control | |
| CN206546593U (en) | House refuse intelligent classification reclaims cloud identifying system | |
| CN203186930U (en) | Intelligentized rubbish classification recycling system | |
| CN104443921A (en) | Intelligent garbage classification throwing, collecting and utilizing system | |
| CN104386388A (en) | Intelligent waste classification and recovery system and recovery method thereof | |
| CN107792561A (en) | A kind of sorting rubbish smartphone system and method | |
| CN106056226A (en) | Cloud recognition system and cloud recognition method for intelligent classification and recycling of domestic garbage | |
| CN111559588A (en) | Intelligent garbage can for classified garbage throwing and classified garbage throwing method | |
| CN103213785A (en) | Multifunctional renewable resource recovery system based on solar energy and internet of things | |
| CN106516490A (en) | Novel intelligent garbage collection station | |
| CN111583508A (en) | Garbage classification management system | |
| CN103387104A (en) | Garbage recycling and recharging system | |
| CN109775180A (en) | Intelligent classification dustbin and its control method | |
| CN105534082A (en) | Self-service type express cabinet | |
| CN106219104A (en) | A kind of beverage bottle recovering device | |
| CN204399863U (en) | Refuse classification intelligence is thrown, is received and utilize system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170818 |
|
| RJ01 | Rejection of invention patent application after publication |