CN109919243A - A kind of scrap iron and steel type automatic identifying method and device based on CNN - Google Patents
A kind of scrap iron and steel type automatic identifying method and device based on CNN Download PDFInfo
- Publication number
- CN109919243A CN109919243A CN201910201422.8A CN201910201422A CN109919243A CN 109919243 A CN109919243 A CN 109919243A CN 201910201422 A CN201910201422 A CN 201910201422A CN 109919243 A CN109919243 A CN 109919243A
- Authority
- CN
- China
- Prior art keywords
- scrap iron
- image
- steel
- neural networks
- convolutional neural
- 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
- XEEYBQQBJWHFJM-UHFFFAOYSA-N Iron Chemical compound [Fe] XEEYBQQBJWHFJM-UHFFFAOYSA-N 0.000 title claims abstract description 222
- 229910052742 iron Inorganic materials 0.000 title claims abstract description 111
- 238000000034 method Methods 0.000 title claims abstract description 24
- 239000002436 steel type Substances 0.000 title claims abstract description 22
- 229910000831 Steel Inorganic materials 0.000 claims abstract description 94
- 239000010959 steel Substances 0.000 claims abstract description 94
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 81
- 238000012360 testing method Methods 0.000 claims abstract description 22
- 238000012549 training Methods 0.000 claims abstract description 21
- 238000013145 classification model Methods 0.000 claims abstract description 14
- 238000010606 normalization Methods 0.000 claims abstract description 11
- 238000013135 deep learning Methods 0.000 claims abstract description 9
- 239000000284 extract Substances 0.000 claims abstract description 4
- 238000012545 processing Methods 0.000 claims description 10
- 238000013507 mapping Methods 0.000 claims description 7
- 230000002708 enhancing effect Effects 0.000 claims description 6
- 230000006835 compression Effects 0.000 claims description 5
- 238000007906 compression Methods 0.000 claims description 5
- 230000009467 reduction Effects 0.000 claims description 5
- 238000007781 pre-processing Methods 0.000 claims description 3
- 230000002787 reinforcement Effects 0.000 claims description 3
- 210000003771 C cell Anatomy 0.000 claims description 2
- 210000004027 cell Anatomy 0.000 claims description 2
- 238000005516 engineering process Methods 0.000 description 12
- 239000000463 material Substances 0.000 description 11
- 244000046052 Phaseolus vulgaris Species 0.000 description 8
- 235000010627 Phaseolus vulgaris Nutrition 0.000 description 8
- 238000004064 recycling Methods 0.000 description 4
- 235000008429 bread Nutrition 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 239000010819 recyclable waste Substances 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 239000002699 waste material Substances 0.000 description 2
- 241000272525 Anas platyrhynchos Species 0.000 description 1
- 241000208340 Araliaceae Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000004883 computer application Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000011946 reduction process Methods 0.000 description 1
- 238000007493 shaping process Methods 0.000 description 1
- 235000014347 soups Nutrition 0.000 description 1
- 238000013517 stratification Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- -1 waste steel bar Chemical compound 0.000 description 1
Abstract
The invention discloses a kind of scrap iron and steel type automatic identifying method and device based on CNN, is collected scrap iron and steel image, establishes scrap iron and steel sample database, and the image of collection is divided into training image collection and test chart image set;Scrap iron and steel image is pre-processed;Scrap iron and steel training image is inputted, machine deep learning is carried out by convolutional neural networks, extracts scrap iron and steel characteristics of image;According to training result, adjusting and optimizing parameter establishes the scrap iron and steel image classification model based on convolutional neural networks.The present invention realizes scrap iron and steel type automatic identification.Compared to manual identified classification method, the objectivity and normalization of scrap iron and steel identification are improved, and recognition speed is fast, accuracy rate is high.
Description
Technical field
The invention belongs to recycling of recyclable waste processing technique field, especially scrap iron and steels to recycle processing technique field, specifically
It is related to a kind of scrap iron and steel type automatic identifying method and device based on CNN.
Background technique
Regenerated resources are disaggregatedly recycled, are the premises that following process utilizes, turns waste into wealth.However, due to
Regenerated resources source complexity, huge number, material difference are big, are accurately identified to regenerated resources and not a duck soup of classifying.Mesh
Before, in recycling of recyclable waste manufacture field, mainly apply manual sort's method, especially the recycling of scrap iron and steel, examine goods, just plus
The links such as work, grass-roots work personnel can only often identify common scrap iron and steel type by rule of thumb, with very big subjectivity and not
Certainty, professional risk and moral hazard are relatively high, inefficiency.Therefore, it is necessary to explore to assist or replace artificial knowledge
The method of other regenerated resources type.Nearly ten years, image procossing and machine learning techniques development are very fast, using these technologies come real
Existing Computer Automatic Recognition classification is just possibly realized.The regenerated resources types such as scrap iron and steel are carried out using computer technology to know automatically
Not, speed is fast, and objectivity is strong, and accuracy is high, can judge problem by accident to avoid brought by subjective understanding when artificial identification.
Machine automatic identification technology is a popular computer application investigative technique, it belongs to the research of computer vision
One of hot spot, especially in recent years with deep learning flourish, deep learning achieves in Target detection and identification
Huge success.Convolutional neural networks CNN is that a kind of operated with trainable filter group and local neighborhood pondization is used alternatingly
In on original input picture, and obtain a kind of deep learning model of gradually complicated stratification characteristics of image.It is aided with appropriate
Regularization means, CNN can be obtained in visual object identification mission independent of any manual extraction feature
Ideal performance.Currently, CNN is in product defects detection, material analysis, text and digital identification, automatic Pilot, family
It is used widely in automatic garbage classification etc..
Scrap iron and steel image has that general categories are uncertain, increment classification occasionally has that generation, small sample classification be more, sample vision circle
The features such as cover half paste and similar categorization are difficult to differentiate between;How the automatic knowledge of scrap iron and steel image is realized using machine recognition technology
Not, the problem of becoming this field urgent need to resolve.
Summary of the invention
In view of the above-mentioned problems, the present invention provides a kind of scrap iron and steel type automatic identifying method and device based on CNN, it will
This deep learning nerual network technique of CNN is applied to the image automatic identification of scrap iron and steel, with the system and dress of the Technology design
The recycling that can be applied to scrap iron and steel and other regenerated resources, sorting, sorting, processing and other fields are set, can also be used as important composition
Part is used for regenerated resources investigation of materials.This technology can be reproduced the departments such as Resources Enterprises, research institution, inspection and quarantine mechanism
It is used, the means of automatic identification can also be provided not have the grass-roots work personnel of relevant professional knowledge.
In order to achieve the above objectives, the technical scheme of the present invention is realized as follows:
A kind of scrap iron and steel type automatic identifying method based on CNN, comprising:
S1, scrap iron and steel image is collected, establishes scrap iron and steel sample database, and the image of collection is divided into trained figure
Image set and test chart image set;
S2, scrap iron and steel image is pre-processed;
S3, the generation of normality data, the generation of small sample orientation and similar sample are carried out to pretreated scrap iron and steel image
Reinforcement generate, scrap iron and steel image is mapped as feature vector;
S4, building convolutional neural networks model and treated that training image collection is trained using step S3;
S5, according to training result, establish the difference feedback award strategy of true value label and test result, pass through iteration adjustment
Optimal Parameters establish the scrap iron and steel image classification model based on convolutional neural networks;
S6, step S3 treated test image will be used to input in trained convolutional neural networks model survey
Examination, judges whether to need return step S1 retraining;
S7, the automatic identification that scrap iron and steel image type is carried out using trained convolutional neural networks model.
Further, scrap iron and steel sample database described in step S1, be stored with the scrap iron and steel sample image of multiple types with
And corresponding scrap iron and steel classification grade.
Further, the pretreatment of scrap iron and steel image described in step S2 includes: to carry out size normalization processing, handles and is
256 × 256 pixels and at least one of: enhancing is restored, coding, compression, noise reduction.
Further, in step S4, convolutional neural networks are established using semi-supervised strategy, using hierarchical model sorting algorithm
It is controlled, i.e., the decision model of bottom is made of CNN classifier, and iteration dynamics is controlled using intensified learning in upper layer
System carries out machine deep learning, extracts scrap iron and steel characteristics of image.
Further, the scrap iron and steel image classification model described in step S5 based on convolutional neural networks includes:
A, by treated, image inputs first convolutional layer in convolutional neural networks, then result is input to pond
Layer;
B, upper one layer of output result is input to the sub-network in convolutional neural networks;
C, the output result of sub-network is input to alternatively distributed convolutional layer and pond layer in convolutional neural networks;
D, output result is input to the full articulamentum in convolutional neural networks;
E, upper one layer of output result is input to the last layer in convolutional neural networks, obtains image and belongs to each class
Other probability;
F, the probability for belonging to each classification according to image obtains scrap iron and steel image classification result.
Another aspect of the present invention additionally provides a kind of scrap iron and steel type automatic identification equipment based on CNN, comprising:
Collection module establishes scrap iron and steel sample database for being collected to scrap iron and steel image, and by the image of collection
It is divided into training image collection and test chart image set;
Preprocessing module, for being pre-processed to scrap iron and steel image;
Mapping block, for pretreated scrap iron and steel image carry out the generation of normality data, small sample orientation generate with
And the reinforcement of similar sample generates, and scrap iron and steel image is mapped as feature vector;
Model training module, for constructing convolutional neural networks model and using mapping block treated training image collection
It is trained;
Model building module, for establishing the difference feedback award plan of true value label and test result according to training result
Slightly, by iteration adjustment Optimal Parameters, the scrap iron and steel image classification model based on convolutional neural networks is established;
Test module, for that mapping block treated test image will be used to input trained convolutional neural networks
It is tested in model, judges whether to need retraining;
Identification module, for using trained convolutional neural networks model to carry out the automatic knowledge of scrap iron and steel image type
Not.
Further, scrap iron and steel sample database described in collection module is stored with the scrap iron and steel sample graph of multiple types
Picture and corresponding scrap iron and steel classification grade.
Further, include: in preprocessing module
Normalization unit carries out size normalization processing, handles as 256 × 256 pixels;
Selecting unit selects at least one of to execute: enhancing is restored, coding, compression, noise reduction.
Further, in model training module, comprising:
Unit is established, convolutional neural networks are established using semi-supervised strategy;
Control unit is controlled using hierarchical model sorting algorithm, i.e., the decision model of bottom is by CNN classifier structure
At upper layer controls iteration dynamics using intensified learning, carries out machine deep learning, extracts scrap iron and steel characteristics of image.
Further, the scrap iron and steel image classification model packet described in model building module based on convolutional neural networks
It includes:
A unit, by treated, image inputs first convolutional layer in convolutional neural networks, then result is input to pond
Change layer;
Upper one layer of output result is input to the sub-network in convolutional neural networks by unit B;
The output result of sub-network is input to alternatively distributed convolutional layer and pond layer in convolutional neural networks by C cell;
Output result is input to the full articulamentum in convolutional neural networks by D unit;
Upper one layer of output result is input to the last layer in convolutional neural networks by E unit, is obtained image and is belonged to respectively
The probability of a classification;
F cell, the probability for belonging to each classification according to image, obtain scrap iron and steel image classification result.
Compared with prior art, the present invention have it is following the utility model has the advantages that
The present invention carries out the feature extraction and deep learning of scrap iron and steel type by using convolutional neural networks technology, realizes
Scrap iron and steel type automatic identification.Compared to manual identified classification method, the objectivity and normalization of scrap iron and steel identification are improved,
And recognition speed is fast, and accuracy rate is high.
Detailed description of the invention
Fig. 1 is the technology path flow diagram of the embodiment of the present invention;
Fig. 2 is the scrap iron and steel image classification model schematic based on convolutional neural networks of the embodiment of the present invention;
Fig. 3 A is the image of input picture material beans of the embodiment of the present invention;
Fig. 3 B is the image of the material beans of output recognition result one of the embodiment of the present invention;
Fig. 3 C is the image of the stamping material clout of output recognition result two of the embodiment of the present invention;
Fig. 3 D is the image of the bread iron of output recognition result three of the embodiment of the present invention;
Fig. 4 A is the image of input picture train wheel of the embodiment of the present invention;
Fig. 4 B is the image of the material beans of output recognition result one of the embodiment of the present invention;
Fig. 4 C is the image of the steel strand rope of output recognition result two of the embodiment of the present invention;
Fig. 4 D is the image of the train wheel of output recognition result three of the embodiment of the present invention.
Specific embodiment
It should be noted that in the absence of conflict, the feature in embodiment and embodiment in the present invention can phase
Mutually combination.
The present invention provides a kind of scrap iron and steel type automatic identifying methods.This method is as shown in Figure 1, 2, comprising:
Establish scrap iron and steel sample database, be stored with multiple scrap iron and steel sample image in the sample database, and with corresponding steel scrap
The corresponding classification grade of iron sample image, wherein scrap iron and steel sample image includes sample graph corresponding to different classes of scrap iron and steel
Picture.Scrap iron and steel sample image is collected by extensive collection mode and channel, it can be by the side that manually shoots on the spot
Formula is also possible to the mode of network downloading, multiple channel and mode can be combined, to obtain enough sample images,
Constitute sample database.Scrap iron and steel sample database should include sample image corresponding to different classes of scrap iron and steel, such as waste steel bar, steel
Ironworking leftover pieces, write-off equipment etc., wherein different classes of scrap iron and steel again be divided into according to thickness, length, volume it is different
Specification.
Scrap iron and steel sample image is pre-processed, including carries out size normalization processing, is processed into 256 × 256
Pixel and at least one of: enhancing is restored, coding, compression, noise reduction.
The generation of normality data, the generation of small sample orientation and similar sample is carried out to pretreated scrap iron and steel image to add
Johnson & Johnson is at being mapped as feature vector for scrap iron and steel image.
Scrap iron and steel image classification model is constructed based on convolutional neural networks technology.The bottom decision of model is by CNN classifier
It constitutes, upper layer controls iteration dynamics using intensified learning, and dynamic adjusts the classified weight between a steel scrap type
Confidence level is controlled by the reward value in intensified learning strategy.
By the training image typing model after mapping, scrap iron and steel image classification model is trained.
According to training result, parameter adjusting and optimizing is carried out, scrap iron and steel image classification model is ultimately formed.
The scrap iron and steel image classification model based on CNN includes:
A. by treated, image inputs first convolutional layer in convolutional neural networks, then result is input to pond
Layer;
B. upper one layer of output result is input to the sub-network in convolutional neural networks;
C. upper one layer of output result is input to alternatively distributed convolutional layer and pond layer in convolutional neural networks;
D. the full articulamentum upper one layer of output result being input in convolutional neural networks;
E. upper one layer of output result is input to the last layer in convolutional neural networks, obtains image and belongs to each class
Other probability;
F. the probability for belonging to each classification according to image obtains scrap iron and steel image classification result.
It in one embodiment of this technology, is tested using test set image, selects the image (such as Fig. 3 A) of material beans,
Image is subjected to size normalization processing, is handled as 256 × 256 pixels.Material beans sample database belongs to large sample, and sample image reaches
96, therefore noise reduction process only is carried out to image, i.e. progress normality data generation.During model running, by output parameter
3 are set as, output recognition result is respectively to expect beans (Fig. 3 B), and matching degree reaches 83.69%;Stamping material clout (Fig. 3 C), matching degree
It is 14.85%;Bread iron (Fig. 3 D), matching degree 0.27%.Front three discrimination is 100%.From this test case as it can be seen that
The technology is preferable for the recognition result of relatively conventional big kind steel scrap type.When test or photo to be identified are handed over close to true
In easily, when provided steel scrap kinds of goods photo, which has preferable performance, and accuracy is higher.
In another embodiment of this technology, the image (such as Fig. 3 A) of the Train wheel of material beans is selected, image is carried out
Size normalization processing, is handled as 256 × 256 pixels.Train wheel sample database belongs to small sample, and sample database includes 31 samples
This image.The problems such as there are background accountings in test image larger, display portion tester, blur margin are clear, fuzzy, and
Sample database quantity is smaller.Therefore layered shaping is carried out to image, preceding scenery and background is labeled, while being restored, being dropped
It makes an uproar and enhancing is handled.Weight adjustment algorithm is used later, changes the weight ratio into ginseng, while carrying out linear fit, to offset sample
The problem of this imbalance is brought.During model running, output parameter is set as 3, output recognition result is respectively to expect beans (figure
4B), matching degree 32.16%;Steel strand rope (Fig. 4 C), matching degree 21.47%;Train wheel (Fig. 4 D), matching degree is
16.38%.Front three discrimination is 100%, but accurately identifies result and appear in third position.Thus test case is as it can be seen that the skill
Art can pass through differentiation feature generating algorithm and mould in the case where identified object defect or background account for relatively high or confusion
Type optimizes to improve discrimination.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all in essence of the invention
Within mind and principle, any modification, equivalent replacement, improvement and so on be should all be included in the protection scope of the present invention.
Claims (10)
1. a kind of scrap iron and steel type automatic identifying method based on CNN characterized by comprising
S1, scrap iron and steel image is collected, establishes scrap iron and steel sample database, and the image of collection is divided into training image collection
With test chart image set;
S2, scrap iron and steel image is pre-processed;
S3, pretreated scrap iron and steel image progress normality data generation, the generation of small sample orientation and similar sample are added
Johnson & Johnson is at being mapped as feature vector for scrap iron and steel image;
S4, building convolutional neural networks model and treated that training image collection is trained using step S3;
S5, according to training result, establish the difference feedback award strategy of true value label and test result, optimized by iteration adjustment
Parameter establishes the scrap iron and steel image classification model based on convolutional neural networks;
S6, step S3 treated test image will be used to input in trained convolutional neural networks model test, sentence
It is disconnected whether to need return step S1 retraining;
S7, the automatic identification that scrap iron and steel image type is carried out using trained convolutional neural networks model.
2. a kind of scrap iron and steel type automatic identifying method based on CNN according to claim 1, which is characterized in that step
Scrap iron and steel sample database described in S1 is stored with scrap iron and steel sample image and the classification of corresponding scrap iron and steel of multiple types etc.
Grade.
3. a kind of scrap iron and steel type automatic identifying method based on CNN according to claim 1, which is characterized in that step
The pretreatment of scrap iron and steel image described in S2 includes: to carry out size normalization processing, is handled as 256 × 256 pixels, and with down toward
One of few: enhancing is restored, coding, compression, noise reduction.
4. a kind of scrap iron and steel type automatic identifying method based on CNN according to claim 1, which is characterized in that step
In S4, convolutional neural networks are established using semi-supervised strategy, are controlled using hierarchical model sorting algorithm, is i.e. the decision of bottom
Model is made of CNN classifier, and upper layer controls iteration dynamics using intensified learning, is carried out machine deep learning, is mentioned
Take scrap iron and steel characteristics of image.
5. a kind of scrap iron and steel type automatic identifying method based on CNN according to claim 1, which is characterized in that step
Scrap iron and steel image classification model described in S5 based on convolutional neural networks includes:
A, by treated, image inputs first convolutional layer in convolutional neural networks, then result is input to pond layer;
B, upper one layer of output result is input to the sub-network in convolutional neural networks;
C, the output result of sub-network is input to alternatively distributed convolutional layer and pond layer in convolutional neural networks;
D, output result is input to the full articulamentum in convolutional neural networks;
E, upper one layer of output result is input to the last layer in convolutional neural networks, obtains image and belongs to each classification
Probability;
F, the probability for belonging to each classification according to image obtains scrap iron and steel image classification result.
6. a kind of scrap iron and steel type automatic identification equipment based on CNN characterized by comprising
Collection module establishes scrap iron and steel sample database, and the image of collection is divided into for being collected to scrap iron and steel image
Training image collection and test chart image set;
Preprocessing module, for being pre-processed to scrap iron and steel image;
Mapping block, for carrying out to pretreated scrap iron and steel image, the generation of normality data, small sample orientation generates and phase
It is generated like the reinforcement of sample, scrap iron and steel image is mapped as feature vector;
Model training module, for constructing convolutional neural networks model and being carried out using mapping block treated training image collection
Training;
Model building module, for establishing the difference feedback award strategy of true value label and test result, leading to according to training result
Iteration adjustment Optimal Parameters are crossed, the scrap iron and steel image classification model based on convolutional neural networks is established;
Test module, for that mapping block treated test image will be used to input trained convolutional neural networks model
It tests in the middle, judges whether to need retraining;
Identification module, for using trained convolutional neural networks model to carry out the automatic identification of scrap iron and steel image type.
7. a kind of scrap iron and steel type automatic identification equipment based on CNN according to claim 6, which is characterized in that collect
Scrap iron and steel sample database described in module is stored with scrap iron and steel sample image and the classification of corresponding scrap iron and steel of multiple types
Grade.
8. a kind of scrap iron and steel type automatic identification equipment based on CNN according to claim 6, which is characterized in that pre- place
Include: in reason module
Normalization unit carries out size normalization processing, handles as 256 × 256 pixels;
Selecting unit selects at least one of to execute: enhancing is restored, coding, compression, noise reduction.
9. a kind of scrap iron and steel type automatic identification equipment based on CNN according to claim 6, which is characterized in that model
In training module, comprising:
Unit is established, convolutional neural networks are established using semi-supervised strategy;
Control unit is controlled using hierarchical model sorting algorithm, i.e., the decision model of bottom is made of CNN classifier, on
Layer controls iteration dynamics using intensified learning, carries out machine deep learning, extracts scrap iron and steel characteristics of image.
10. a kind of scrap iron and steel type automatic identification equipment based on CNN according to claim 6, which is characterized in that model
Establishing the scrap iron and steel image classification model described in module based on convolutional neural networks includes:
A unit, by treated, image inputs first convolutional layer in convolutional neural networks, then result is input to pond
Layer;
Upper one layer of output result is input to the sub-network in convolutional neural networks by unit B;
The output result of sub-network is input to alternatively distributed convolutional layer and pond layer in convolutional neural networks by C cell;
Output result is input to the full articulamentum in convolutional neural networks by D unit;
Upper one layer of output result is input to the last layer in convolutional neural networks by E unit, is obtained image and is belonged to each class
Other probability;
F cell, the probability for belonging to each classification according to image, obtain scrap iron and steel image classification result.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910201422.8A CN109919243A (en) | 2019-03-15 | 2019-03-15 | A kind of scrap iron and steel type automatic identifying method and device based on CNN |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910201422.8A CN109919243A (en) | 2019-03-15 | 2019-03-15 | A kind of scrap iron and steel type automatic identifying method and device based on CNN |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109919243A true CN109919243A (en) | 2019-06-21 |
Family
ID=66965259
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910201422.8A Pending CN109919243A (en) | 2019-03-15 | 2019-03-15 | A kind of scrap iron and steel type automatic identifying method and device based on CNN |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109919243A (en) |
Cited By (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110427896A (en) * | 2019-08-07 | 2019-11-08 | 成都理工大学 | A kind of garbage classification intelligence system based on convolutional neural networks |
CN110472692A (en) * | 2019-08-21 | 2019-11-19 | 天津拾起卖科技有限公司 | A kind of portable, hand-held device scrap iron and steel type automatic identification equipment |
CN110619355A (en) * | 2019-08-28 | 2019-12-27 | 武汉科技大学 | Automatic steel material microstructure identification method based on deep learning |
CN110717455A (en) * | 2019-10-10 | 2020-01-21 | 北京同创信通科技有限公司 | Method for classifying and detecting grades of scrap steel in storage |
CN112348791A (en) * | 2020-11-04 | 2021-02-09 | 中冶赛迪重庆信息技术有限公司 | Intelligent scrap steel detecting and judging method, system, medium and terminal based on machine vision |
CN112541427A (en) * | 2020-12-10 | 2021-03-23 | 欧冶链金再生资源有限公司 | Identification and material quantity evaluation method for high-quality heavy steel scrap |
CN112744439A (en) * | 2021-01-15 | 2021-05-04 | 湖南镭目科技有限公司 | Remote scrap steel monitoring system based on deep learning technology |
CN112801391A (en) * | 2021-02-04 | 2021-05-14 | 科大智能物联技术有限公司 | Artificial intelligent scrap steel impurity deduction rating method and system |
CN112881412A (en) * | 2021-02-01 | 2021-06-01 | 南京耘瞳科技有限公司 | Method for detecting non-metal foreign bodies in scrap steel products |
CN113191176A (en) * | 2020-06-08 | 2021-07-30 | 阿里巴巴集团控股有限公司 | Data processing method and device |
CN113627830A (en) * | 2021-09-08 | 2021-11-09 | 西安智者云集云计算有限公司 | Scrap steel grading system and method |
CN114078126A (en) * | 2022-01-19 | 2022-02-22 | 江苏金恒信息科技股份有限公司 | Scrap steel grading method and device based on machine learning |
CN114414580A (en) * | 2022-03-30 | 2022-04-29 | 新极技术(北京)有限公司 | Method for identifying and deducting impurities on scrap steel |
CN117132828A (en) * | 2023-08-30 | 2023-11-28 | 常州润来科技有限公司 | Automatic classification method and system for solid waste in copper pipe machining process |
Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101853400A (en) * | 2010-05-20 | 2010-10-06 | 武汉大学 | Multiclass image classification method based on active learning and semi-supervised learning |
CN105488515A (en) * | 2014-09-17 | 2016-04-13 | 富士通株式会社 | Method for training convolutional neural network classifier and image processing device |
CN106842925A (en) * | 2017-01-20 | 2017-06-13 | 清华大学 | A kind of locomotive smart steering method and system based on deeply study |
CN107679491A (en) * | 2017-09-29 | 2018-02-09 | 华中师范大学 | A kind of 3D convolutional neural networks sign Language Recognition Methods for merging multi-modal data |
CN107798381A (en) * | 2017-11-13 | 2018-03-13 | 河海大学 | A kind of image-recognizing method based on convolutional neural networks |
CN107863147A (en) * | 2017-10-24 | 2018-03-30 | 清华大学 | The method of medical diagnosis based on depth convolutional neural networks |
CN108288077A (en) * | 2018-04-17 | 2018-07-17 | 天津和或节能科技有限公司 | Grading of old paper device establishes device and method, grading of old paper system and method |
CN108596203A (en) * | 2018-03-13 | 2018-09-28 | 北京交通大学 | Optimization method of the pond layer in parallel to pantograph carbon slide surface abrasion detection model |
CN108764006A (en) * | 2018-02-05 | 2018-11-06 | 北京航空航天大学 | A kind of SAR image object detection method based on deeply study |
CN109034205A (en) * | 2018-06-29 | 2018-12-18 | 西安交通大学 | Image classification method based on the semi-supervised deep learning of direct-push |
-
2019
- 2019-03-15 CN CN201910201422.8A patent/CN109919243A/en active Pending
Patent Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101853400A (en) * | 2010-05-20 | 2010-10-06 | 武汉大学 | Multiclass image classification method based on active learning and semi-supervised learning |
CN105488515A (en) * | 2014-09-17 | 2016-04-13 | 富士通株式会社 | Method for training convolutional neural network classifier and image processing device |
CN106842925A (en) * | 2017-01-20 | 2017-06-13 | 清华大学 | A kind of locomotive smart steering method and system based on deeply study |
CN107679491A (en) * | 2017-09-29 | 2018-02-09 | 华中师范大学 | A kind of 3D convolutional neural networks sign Language Recognition Methods for merging multi-modal data |
CN107863147A (en) * | 2017-10-24 | 2018-03-30 | 清华大学 | The method of medical diagnosis based on depth convolutional neural networks |
CN107798381A (en) * | 2017-11-13 | 2018-03-13 | 河海大学 | A kind of image-recognizing method based on convolutional neural networks |
CN108764006A (en) * | 2018-02-05 | 2018-11-06 | 北京航空航天大学 | A kind of SAR image object detection method based on deeply study |
CN108596203A (en) * | 2018-03-13 | 2018-09-28 | 北京交通大学 | Optimization method of the pond layer in parallel to pantograph carbon slide surface abrasion detection model |
CN108288077A (en) * | 2018-04-17 | 2018-07-17 | 天津和或节能科技有限公司 | Grading of old paper device establishes device and method, grading of old paper system and method |
CN109034205A (en) * | 2018-06-29 | 2018-12-18 | 西安交通大学 | Image classification method based on the semi-supervised deep learning of direct-push |
Cited By (21)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110427896A (en) * | 2019-08-07 | 2019-11-08 | 成都理工大学 | A kind of garbage classification intelligence system based on convolutional neural networks |
CN110472692A (en) * | 2019-08-21 | 2019-11-19 | 天津拾起卖科技有限公司 | A kind of portable, hand-held device scrap iron and steel type automatic identification equipment |
CN110619355A (en) * | 2019-08-28 | 2019-12-27 | 武汉科技大学 | Automatic steel material microstructure identification method based on deep learning |
CN110619355B (en) * | 2019-08-28 | 2023-01-06 | 武汉科技大学 | Automatic steel material microstructure identification method based on deep learning |
CN110717455A (en) * | 2019-10-10 | 2020-01-21 | 北京同创信通科技有限公司 | Method for classifying and detecting grades of scrap steel in storage |
CN110717455B (en) * | 2019-10-10 | 2021-05-18 | 北京同创信通科技有限公司 | Method for classifying and detecting grades of scrap steel in storage |
CN113191176A (en) * | 2020-06-08 | 2021-07-30 | 阿里巴巴集团控股有限公司 | Data processing method and device |
CN112348791A (en) * | 2020-11-04 | 2021-02-09 | 中冶赛迪重庆信息技术有限公司 | Intelligent scrap steel detecting and judging method, system, medium and terminal based on machine vision |
CN112348791B (en) * | 2020-11-04 | 2023-03-14 | 中冶赛迪信息技术(重庆)有限公司 | Intelligent scrap steel detecting and judging method, system, medium and terminal based on machine vision |
CN112541427A (en) * | 2020-12-10 | 2021-03-23 | 欧冶链金再生资源有限公司 | Identification and material quantity evaluation method for high-quality heavy steel scrap |
CN112541427B (en) * | 2020-12-10 | 2022-11-01 | 欧冶链金再生资源有限公司 | Identification and material quantity evaluation method for high-quality heavy steel scrap |
CN112744439A (en) * | 2021-01-15 | 2021-05-04 | 湖南镭目科技有限公司 | Remote scrap steel monitoring system based on deep learning technology |
CN112881412A (en) * | 2021-02-01 | 2021-06-01 | 南京耘瞳科技有限公司 | Method for detecting non-metal foreign bodies in scrap steel products |
CN112881412B (en) * | 2021-02-01 | 2023-03-10 | 南京耘瞳科技有限公司 | Method for detecting non-metal foreign matters in scrap steel products |
CN112801391B (en) * | 2021-02-04 | 2021-11-19 | 科大智能物联技术股份有限公司 | Artificial intelligent scrap steel impurity deduction rating method and system |
CN112801391A (en) * | 2021-02-04 | 2021-05-14 | 科大智能物联技术有限公司 | Artificial intelligent scrap steel impurity deduction rating method and system |
CN113627830A (en) * | 2021-09-08 | 2021-11-09 | 西安智者云集云计算有限公司 | Scrap steel grading system and method |
CN114078126A (en) * | 2022-01-19 | 2022-02-22 | 江苏金恒信息科技股份有限公司 | Scrap steel grading method and device based on machine learning |
CN114414580A (en) * | 2022-03-30 | 2022-04-29 | 新极技术(北京)有限公司 | Method for identifying and deducting impurities on scrap steel |
CN117132828A (en) * | 2023-08-30 | 2023-11-28 | 常州润来科技有限公司 | Automatic classification method and system for solid waste in copper pipe machining process |
CN117132828B (en) * | 2023-08-30 | 2024-03-19 | 常州润来科技有限公司 | Automatic classification method and system for solid waste in copper pipe machining process |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109919243A (en) | A kind of scrap iron and steel type automatic identifying method and device based on CNN | |
CN108154134B (en) | Pornographic image detection method is broadcast live in internet based on depth convolutional neural networks | |
CN111144496B (en) | Garbage classification method based on hybrid convolutional neural network | |
CN110779937A (en) | Casting product internal defect intelligent detection device | |
CN105303193B (en) | A kind of passenger number statistical system based on single-frame images processing | |
CN112083017A (en) | Intelligent detection device, method and medium for internal defects of welding seam | |
WO2021217775A1 (en) | Steel scrap rating method and apparatus based on fusion of neural network models | |
CN111186656A (en) | Target garbage classification method and intelligent garbage can | |
CN104766097B (en) | Surface of aluminum plate defect classification method based on BP neural network and SVMs | |
CN105678332A (en) | Converter steel-making endpoint determination method and system based on flame image CNN recognizing and modeling process | |
CN111401418A (en) | Employee dressing specification detection method based on improved Faster r-cnn | |
CN110827260B (en) | Cloth defect classification method based on LBP characteristics and convolutional neural network | |
CN109308479B (en) | Deep learning-based campus garbage automatic classification and recovery method | |
CN112541427B (en) | Identification and material quantity evaluation method for high-quality heavy steel scrap | |
CN104751186A (en) | Iris image quality classification method based on BP (back propagation) network and wavelet transformation | |
CN113362299B (en) | X-ray security inspection image detection method based on improved YOLOv4 | |
CN107644218A (en) | The method of work of crowded region behavioural analysis judgement is realized based on image collecting function | |
Zheng et al. | Direct neural network application for automated cell recognition | |
KR20220156603A (en) | Method and electronic device for automated waste management | |
CN107145778A (en) | A kind of intrusion detection method and device | |
CN109523514A (en) | To the batch imaging quality assessment method of Inverse Synthetic Aperture Radar ISAR | |
Mookkaiah et al. | Design and development of smart Internet of Things–based solid waste management system using computer vision | |
CN114515705A (en) | Coal gangue sorting system based on multi-mode imaging analysis | |
Sree et al. | Waste Management Detection Using Deep Learning | |
CN117088015A (en) | Intelligent garbage disposal method, system and storage medium |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190621 |