CN109784202A - Recognition methods, device and the washing machine and computer readable storage medium of object are left in article to be washed - Google Patents
Recognition methods, device and the washing machine and computer readable storage medium of object are left in article to be washed Download PDFInfo
- Publication number
- CN109784202A CN109784202A CN201811584130.9A CN201811584130A CN109784202A CN 109784202 A CN109784202 A CN 109784202A CN 201811584130 A CN201811584130 A CN 201811584130A CN 109784202 A CN109784202 A CN 109784202A
- Authority
- CN
- China
- Prior art keywords
- article
- washed
- neural networks
- convolutional neural
- image information
- 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
Abstract
The application provides recognition methods, device and the washing machine and computer readable storage medium that object is left in a kind of article to be washed.The described method includes: obtaining the image information of article to be washed;The image information of the article to be washed is inputted into convolutional neural networks model, to obtain leaving object recognition result;Residue information alert is carried out according to the residue recognition result.By the scheme of the application, it can intelligently identify without human intervention and leave object, avoid and irremediable Important Economic is caused to lose and the loss of some important informations due to leaving object in laundry item.
Description
Technical field
This application involves recognition methods, dresses that object is left in washing machine technology field more particularly to a kind of article to be washed
It sets and washing machine and computer readable storage medium.
Background technique
With the rapid development of artificial intelligence and big data technology, intelligent washing machine is more single in conventional washer function
In the case of integrate under Scenario operating mode realize machine washing article polymorphic type, realize multifunction, reduce the movement of people.
However, machine washing of clothes brings the problem of headache, the accidentally mobile electronic device such as U that valuable object is such as not capable of washing
Disk, mobile phone etc. are retained in article-cleaning, easily cause the loss of Important Economic loss and some important informations, bring is not
Therefore repairable consequence needs a kind of washing machine for being able to detect that and leaving object in article to be washed.
In view of the above-mentioned problems, not yet proposing effective solution mode at present.
Summary of the invention
In view of this, the application proposes recognition methods, device and the washing machine and meter of leaving object in a kind of article to be washed
Calculation machine readable storage medium storing program for executing, the analysis of the image information of washing product is treated by convolutional neural networks model, obtains leaving object
Recognition result, can without human intervention and intelligently identify and leave object, avoid because leaving object in laundry item
In and cause irremediable Important Economic to lose and the loss of some important informations.
According to the one aspect of the application, provide the recognition methods that object is left in a kind of article to be washed, comprising: obtain to
The image information of washing product;The image information of the article to be washed is inputted into convolutional neural networks model, to obtain leaving object
Recognition result;Residue information alert is carried out according to the residue recognition result.
Further, the image information of article to be washed is obtained by X ray sensor;And/or the convolutional neural networks
Model is trained convolutional neural networks and is verified and obtains by great amount of samples picture;And/or described leave article identification knot
Fruit includes the classification for leaving object.
Further, the convolutional neural networks successively include input layer, convolutional layer, pond layer, full articulamentum and
Softmax layers, using the picture element matrix of the image information of the article to be washed as the input layer.
Further, the convolutional neural networks include more wheel convolutional layers and pond layer, by more wheel convolutional layers with
Pond layer, by described image information at the higher feature of information content;And/or the convolutional neural networks include at least
Distributed nature expression is mapped to sample labeling space by the full articulamentum by two layers of full articulamentum;And/or pass through institute
State classification of the softmax layer completion of convolutional neural networks to object is left.
According to the another aspect of the application, the identification device that object is left in a kind of article to be washed is provided, comprising: image
Acquiring unit, for obtaining the image information of article to be washed;Object recognition unit is left, for by the image of the article to be washed
Information input convolutional neural networks model, to obtain leaving object recognition result.
Further, the image of article to be washed is obtained by X ray sensor;And/or
The convolutional neural networks model is trained convolutional neural networks and is verified by great amount of samples picture to be obtained;
And/or the object recognition result of leaving includes the classification for leaving object.
Further, the convolutional neural networks successively include input layer, convolutional layer, pond layer, full articulamentum and
Softmax layers, using the picture element matrix of the image information of the article to be washed as the input layer.
Further, the convolutional neural networks include more wheel convolutional layers and pond layer, by more wheel convolutional layers with
Pond layer, by described image information at the higher feature of information content;And/or the convolutional neural networks include at least
Distributed nature expression is mapped to sample labeling space by the full articulamentum by two layers of full articulamentum;And/or pass through institute
State classification of the softmax layer completion of convolutional neural networks to object is left.
According to the another aspect of the application, a kind of washing machine, including identification device recited above and prompt dress are provided
It sets, the suggestion device is used to carry out residue information alert according to residue recognition result.
According to the another aspect of the application, a kind of computer readable storage medium is stored thereon with computer program,
It is characterized in that, as above any method is realized when which is executed by processor.
Recognition methods, device and the washing machine and computer of object are left in a kind of article to be washed proposed according to the application
Readable storage medium storing program for executing treats the analysis of the image information of washing product by convolutional neural networks model, obtains the knowledge for leaving object
Not as a result, it is possible to intelligently identify without human intervention and leave object, avoid due to leaving object in laundry item
Cause the loss of irremediable Important Economic loss and some important informations.
Above description is only the general introduction of technical scheme, in order to better understand the technological means of the application,
And can be implemented in accordance with the contents of the specification, with the preferred embodiment of the application and cooperate attached drawing below detailed description is as follows.
Detailed description of the invention
The attached drawing for constituting a part of the invention is used to provide further understanding of the present invention, schematic reality of the invention
It applies example and its explanation is used to explain the present invention, do not constitute improper limitations of the present invention.
In the accompanying drawings:
Fig. 1 shows a kind of schematic diagram of an embodiment of washing machine of the application;
Fig. 2 shows the schematic diagrames of an embodiment of the recognition methods that object is left in a kind of article to be washed of the application;
Fig. 3 shows a kind of schematic diagram of an embodiment of convolutional neural networks model of the application.
Specific embodiment
To keep the purposes, technical schemes and advantages of the application clearer, below in conjunction with the application specific embodiment and
Technical scheme is clearly and completely described in corresponding attached drawing.Obviously, described embodiment is only the application one
Section Example, instead of all the embodiments.Based on the embodiment in the application, those of ordinary skill in the art are not doing
Every other embodiment obtained under the premise of creative work out, shall fall in the protection scope of this application.
Fig. 1 shows a kind of schematic diagram of an embodiment of washing machine of the application.
As shown, washing machine 1 includes leaving object identification device 11 and suggestion device 12, object identification device 11 is left
Including image acquisition unit 111 and leave object recognition unit 112.
It should be understood that leaving object identification device 11 can be used as an independent product, it is not necessary to as laundry
A part of machine can also be not limited to using on washing machine.
Fig. 2 shows the schematic diagrames of an embodiment of the recognition methods that object is left in a kind of article to be washed of the application.
As shown, method includes the following steps:
S1 obtains the image information of article to be washed.
Specifically, image acquisition unit 111 obtains the image information of article to be washed.Image acquisition unit 111 can be each
The sensor of the available image information of kind, such as various cameras, but be preferably able to penetrate the inductor of cloth, such as X-ray
Sensor etc. treats washing product and is scanned to obtain the image information of article to be washed as luggage crosses subway safety check, the image
Information contains the image information for leaving object being mixed into article to be washed.It in one implementation, can be by article to be washed
After being put into washing machine inner tub such as clothing, washing product are treated by the sensor installed on washing machine inner tub and are shot to obtain it
Image information.Optionally, in order to more accurately identify, it usually needs multiple image informations of multi-angle, by being set in interior bucket
Multiple sensors are set, and/or control interior bucket rotation shakes article to be washed to different angle, alternately to clap laundry item
According to until acquisition image information reaches predetermined amount.Wherein, predetermined amount can be demarcated according to the actual situation, here with no restrictions.
The image information of article to be washed is inputted convolutional neural networks model, to obtain leaving object recognition result by S2.
Specifically, it leaves object recognition unit 112 and the image information of article to be washed is inputted into convolutional neural networks model, with
It obtains leaving object recognition result.
Convolutional neural networks (CNN) are one of the network structures of great representative in depth learning technology, are led in image procossing
Domain achieves very big success.The application identified using convolutional neural networks to leaving object, will be with outstanding work
Make efficiency.In this application, convolutional neural networks model is trained and is verified to convolutional neural networks by great amount of samples picture
And it obtains.Initialization, which is carried out, using great amount of samples of the convolutional neural networks to picture database is converted to the data that can be handled by it
Collection realizes the training of convolutional neural networks model and tests verifying and builds the seed superiority of model to obtain final mask.Sample graph
Piece source can be sample clothing being put into the image information that washing machine inner tub is shot, and be also possible to obtain by network
Existing sample image information.And the image information of article to be washed, it will also become new samples pictures and be used for training optimization convolution
Neural network model.
Optionally, convolutional neural networks successively include input layer, convolutional layer, pond layer, full articulamentum and softmax layers.
Input layer generally represents the picture element matrix of picture, and since input layer, convolutional layer analyzes each piece in neural network
The higher feature of degree is taken out, the depth of node matrix is increased.Pond layer acts on convolutional layer treated n dimension matrix,
In the case where not changing matrix depth, the number of last full articulamentum interior joint is further reduced, to reach the entire mind of reduction
Purpose through parameter in network reduces matrix size, convenient for processing.
In one implementation, convolutional neural networks include more wheel convolutional layers and pond layer in the application, by taking turns more
Image information is abstracted into the higher feature of information content by convolutional layer and pond layer.Convolutional layer and pond layer are being taken turns certainly through excessive
After after the integrated disposal processing that dynamic character extracts, it is believed that image information is abstracted into the higher feature of information content.
In one implementation, convolutional neural networks include at least two layers full articulamentum (for example, by using 2 layers in the application
Full articulamentum), distributed nature expression is mapped to sample labeling space so as to preferably mention by least 2 layers full articulamentum
Pure other features of reduction reduce interference characteristic.The more Floating-point Computations of the number of plies of full articulamentum are bigger, calculate slower, it is possible to understand that
Ground, those skilled in the art can use one layer of full articulamentum according to practical operation demand, can also use at least two layers, it is preferable that
It is advisable using 2 layers.
Optionally, leaving object recognition result includes the classification for leaving object, passes through the softmax layer of convolutional neural networks
Realize the classification to object is left.The classification is for example including without residue, residue is electronic equipment, file etc., electronics
Equipment can also be classified as USB flash disk, mobile phone, wrist-watch, other electronic equipments etc., without being limited thereto.
Fig. 3 shows a kind of schematic diagram of an embodiment of convolutional neural networks of the application.As shown,
Washing machine inputs convolutional layer using the original image information picture element matrix of article to be washed as input layer 201 first
202;
Right side unit-node is calculated by left side minor matrix interior joint by the way of propagated forward, is weighted in conjunction with node
With and parameter sharing mechanism complete convolutional layer propagated forward process;
Pond layer 203 acts on the output of convolutional layer using propagated forward and maximum value or average operations, reduces matrix
Length and width, reduce matrix size;
Taken turns more convolutional layer and pond layer effect (although being illustrated as 2 layers, it is to be understood that, art technology
Personnel can be adjusted to the number of plies of any needs according to actual needs), it is higher that the information in image is abstracted into information content
Feature;
Full articulamentum 205 (although being illustrated as 2 layers, it is to be understood that, those skilled in the art can be according to reality
Demand is adjusted to the number of plies of any needs) by distributed nature expression be mapped to sample labeling space, thus by eigenmatrix into
Distributed nature expression is mapped to sample labeling space by highly purified other features of reduction of row;
Finally classification results are completed by the effect of softmax layer 206.
S3 carries out residue information alert according to residue recognition result.
Specifically, suggestion device 12 carries out residue information alert according to residue recognition result.
Optionally, it after obtaining residue recognition result, is prompted by vision and/or audible means to user.
In one implementation, display screen and/or audio player are installed on intelligent washing machine, it can be by aobvious
Show screen display and/or is prompted by audio player plays voice.
In one implementation, the electronic equipments such as user terminal such as mobile phone, plate, wrist-watch by telephone number, under
It carries at least one modes such as corresponding application program app to be associated with washing machine, washing machine suggestion device 12 is by residue recognition result
User terminal is sent to by network data and/or short message mode to complete in user terminal by vision and/or audible means
Prompt.
By prompt, user can take out residue in time, prevent loss of the residue in cleaning.
The application also proposed a kind of computer readable storage medium, can be read-only memory, and disk or CD etc. are each
Kind can store the medium of program code, be stored with computer program, wherein when the program is executed by processor, it can be achieved that originally
Apply for the recognition methods that object is left in article to be washed proposed above.
It to recognition methods, device and the washing machine of leaving object in the article to be washed of the application and computer-readable deposits above
Storage media treats the analysis of the image information of washing product by convolutional neural networks model, obtains the recognition result for leaving object,
It can intelligently identify without human intervention and leave object, avoid that cause due to leaving object in laundry item can not
The loss of the Important Economic loss and some important informations retrieved.
It should be noted that in the absence of conflict, the feature in embodiment and embodiment in the present invention can phase
Mutually combination.
Those of ordinary skill in the art will appreciate that realizing that all or part of the steps of above-described embodiment can pass through hardware
It completes, relevant hardware can also be instructed to complete by program.
The above described is only a preferred embodiment of the present invention, be not intended to limit the present invention in any form, according to
According to technical spirit any simple modification, equivalent change and modification to the above embodiments of the invention, this hair is still fallen within
In the range of bright technical solution.
Claims (10)
1. leaving the recognition methods of object in a kind of article to be washed characterized by comprising
Obtain the image information of article to be washed;
The image information of the article to be washed is inputted into convolutional neural networks model, to obtain leaving object recognition result;
Residue information alert is carried out according to the residue recognition result.
2. the method as described in claim 1, it is characterised in that:
The image information of article to be washed is obtained by X ray sensor;
And/or
The convolutional neural networks model is trained convolutional neural networks and is verified by great amount of samples picture to be obtained;
And/or
The object recognition result of leaving includes the classification for leaving object.
3. method according to claim 2, it is characterised in that:
The convolutional neural networks successively include input layer, convolutional layer, pond layer, full articulamentum and softmax layers,
Using the picture element matrix of the image information of the article to be washed as the input layer.
4. method as claimed in claim 3, it is characterised in that:
The convolutional neural networks include more wheel convolutional layers and pond layer, by more wheel convolutional layers and pond layer, by image
Information is at the higher feature of information content;
And/or
The convolutional neural networks include at least two layers full articulamentum, indicate to map by distributed nature by the full articulamentum
To sample labeling space;
And/or
The classification to object is left is completed by the softmax layer of the convolutional neural networks.
5. leaving the identification device of object in a kind of article to be washed characterized by comprising
Image acquisition unit, for obtaining the image information of article to be washed;
Object recognition unit is left, for the image information of the article to be washed to be inputted convolutional neural networks model, to obtain
Leave object recognition result.
6. device as claimed in claim 5, it is characterised in that:
The image information of article to be washed is obtained by X ray sensor;
And/or
The convolutional neural networks model is trained convolutional neural networks and is verified by great amount of samples picture to be obtained;
And/or
The object recognition result of leaving includes the classification for leaving object.
7. such as device described in claim 5 or 6, it is characterised in that:
The convolutional neural networks successively include input layer, convolutional layer, pond layer, full articulamentum and softmax layers,
Using the picture element matrix of the image information of the article to be washed as the input layer.
8. device as claimed in claim 7, it is characterised in that:
The convolutional neural networks include more wheel convolutional layers and pond layer, by more wheel convolutional layers and pond layer, by image
Information is at the higher feature of information content;
And/or
The convolutional neural networks include at least two layers full articulamentum, indicate to map by distributed nature by the full articulamentum
To sample labeling space;
And/or
The classification to object is left is completed by the softmax layer of the convolutional neural networks.
9. a kind of washing machine, which is characterized in that including identification device as claimed in claim 5 and suggestion device, the prompt
Device is used to carry out residue information alert according to residue recognition result.
10. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the program is by processor
The method as described in any in claim 1-4 is realized when execution.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811584130.9A CN109784202A (en) | 2018-12-24 | 2018-12-24 | Recognition methods, device and the washing machine and computer readable storage medium of object are left in article to be washed |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811584130.9A CN109784202A (en) | 2018-12-24 | 2018-12-24 | Recognition methods, device and the washing machine and computer readable storage medium of object are left in article to be washed |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109784202A true CN109784202A (en) | 2019-05-21 |
Family
ID=66498144
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811584130.9A Pending CN109784202A (en) | 2018-12-24 | 2018-12-24 | Recognition methods, device and the washing machine and computer readable storage medium of object are left in article to be washed |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109784202A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112686183A (en) * | 2021-01-04 | 2021-04-20 | 大陆投资(中国)有限公司 | Remnant detection device, system, method and electronic equipment |
Citations (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1389724A (en) * | 2001-06-04 | 2003-01-08 | 安立股份有限公司 | Apparatus for detecting X-ray impurities to simultaneously detect multiple X-rays with different energy |
CN103487451A (en) * | 2013-10-15 | 2014-01-01 | 无锡艾科瑞思产品设计与研究有限公司 | Multi-angle dynamic detection device and method for flour foods |
US20160148079A1 (en) * | 2014-11-21 | 2016-05-26 | Adobe Systems Incorporated | Object detection using cascaded convolutional neural networks |
US9424494B1 (en) * | 2016-01-28 | 2016-08-23 | International Business Machines Corporation | Pure convolutional neural network localization |
CN105931217A (en) * | 2016-04-05 | 2016-09-07 | 李红伟 | Image processing technology-based airport pavement FOD (foreign object debris) detection method |
WO2017151673A1 (en) * | 2016-02-29 | 2017-09-08 | Digitalglobe, Inc. | Synthesizing training data for broad area geospatial object detection |
CN107301406A (en) * | 2017-07-13 | 2017-10-27 | 珠海多智科技有限公司 | Fast face angle recognition method based on deep learning |
CN107392313A (en) * | 2017-06-12 | 2017-11-24 | 五邑大学 | A kind of rail recognition methods based on deep learning |
CN107447442A (en) * | 2017-08-31 | 2017-12-08 | 深圳市盛路物联通讯技术有限公司 | A kind of alarm method and smart machine based on image recognition technology |
CN107653625A (en) * | 2017-08-30 | 2018-02-02 | 深圳市盛路物联通讯技术有限公司 | A kind of intelligent detecting method and smart machine |
CN108734143A (en) * | 2018-05-28 | 2018-11-02 | 江苏迪伦智能科技有限公司 | A kind of transmission line of electricity online test method based on binocular vision of crusing robot |
CN108796988A (en) * | 2018-05-30 | 2018-11-13 | 苏州树云网络科技有限公司 | A kind of foreign matter detection method |
CN108823898A (en) * | 2018-05-30 | 2018-11-16 | 苏州树云网络科技有限公司 | A kind of foreign matter detection system for washing machine |
-
2018
- 2018-12-24 CN CN201811584130.9A patent/CN109784202A/en active Pending
Patent Citations (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1389724A (en) * | 2001-06-04 | 2003-01-08 | 安立股份有限公司 | Apparatus for detecting X-ray impurities to simultaneously detect multiple X-rays with different energy |
CN103487451A (en) * | 2013-10-15 | 2014-01-01 | 无锡艾科瑞思产品设计与研究有限公司 | Multi-angle dynamic detection device and method for flour foods |
US20160148079A1 (en) * | 2014-11-21 | 2016-05-26 | Adobe Systems Incorporated | Object detection using cascaded convolutional neural networks |
US9424494B1 (en) * | 2016-01-28 | 2016-08-23 | International Business Machines Corporation | Pure convolutional neural network localization |
WO2017151673A1 (en) * | 2016-02-29 | 2017-09-08 | Digitalglobe, Inc. | Synthesizing training data for broad area geospatial object detection |
CN105931217A (en) * | 2016-04-05 | 2016-09-07 | 李红伟 | Image processing technology-based airport pavement FOD (foreign object debris) detection method |
CN107392313A (en) * | 2017-06-12 | 2017-11-24 | 五邑大学 | A kind of rail recognition methods based on deep learning |
CN107301406A (en) * | 2017-07-13 | 2017-10-27 | 珠海多智科技有限公司 | Fast face angle recognition method based on deep learning |
CN107653625A (en) * | 2017-08-30 | 2018-02-02 | 深圳市盛路物联通讯技术有限公司 | A kind of intelligent detecting method and smart machine |
CN107447442A (en) * | 2017-08-31 | 2017-12-08 | 深圳市盛路物联通讯技术有限公司 | A kind of alarm method and smart machine based on image recognition technology |
CN108734143A (en) * | 2018-05-28 | 2018-11-02 | 江苏迪伦智能科技有限公司 | A kind of transmission line of electricity online test method based on binocular vision of crusing robot |
CN108796988A (en) * | 2018-05-30 | 2018-11-13 | 苏州树云网络科技有限公司 | A kind of foreign matter detection method |
CN108823898A (en) * | 2018-05-30 | 2018-11-16 | 苏州树云网络科技有限公司 | A kind of foreign matter detection system for washing machine |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112686183A (en) * | 2021-01-04 | 2021-04-20 | 大陆投资(中国)有限公司 | Remnant detection device, system, method and electronic equipment |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
TWI746674B (en) | Type prediction method, device and electronic equipment for identifying objects in images | |
CN107392091B (en) | Agricultural artificial intelligence crop detection method, mobile terminal and computer readable medium | |
CN109816009A (en) | Multi-tag image classification method, device and equipment based on picture scroll product | |
CN109658455A (en) | Image processing method and processing equipment | |
CN106868792A (en) | The clothes processing method and device of image content-based search engine | |
CN108108767A (en) | A kind of cereal recognition methods, device and computer storage media | |
CN107851195A (en) | Target detection is carried out using neutral net | |
CN107742107A (en) | Facial image sorting technique, device and server | |
CN108960207A (en) | A kind of method of image recognition, system and associated component | |
CN104200249B (en) | A kind of method of clothing automatic collocation, apparatus and system | |
CN107808126A (en) | Vehicle retrieval method and device | |
CN109145766A (en) | Model training method, device, recognition methods, electronic equipment and storage medium | |
CN109359515A (en) | A kind of method and device that the attributive character for target object is identified | |
CN110516671A (en) | Training method, image detecting method and the device of neural network model | |
US11475500B2 (en) | Device and method for item recommendation based on visual elements | |
CN106548179A (en) | The detection method of object and dress ornament key point, device and electronic equipment | |
Klein et al. | Salient pattern detection using W 2 on multivariate normal distributions | |
CN108230232A (en) | The method and relevant apparatus of a kind of image procossing | |
Bedeli et al. | Clothing identification via deep learning: forensic applications | |
CN108229413A (en) | Pest and disease damage kind identification method and device | |
Dai et al. | Tan: Temporal aggregation network for dense multi-label action recognition | |
AU2017208235A1 (en) | Relative position encoding based networks for action recognition | |
Baslamisli et al. | Shadingnet: Image intrinsics by fine-grained shading decomposition | |
Zhang et al. | Study of visual saliency detection via nonlocal anisotropic diffusion equation | |
CN107391540A (en) | A kind of small routine methods of exhibiting, device and grader |
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: 20190521 |
|
RJ01 | Rejection of invention patent application after publication |