CN109886302A - Caliber judgment method and terminal device based on machine learning - Google Patents
Caliber judgment method and terminal device based on machine learning Download PDFInfo
- Publication number
- CN109886302A CN109886302A CN201910053757.XA CN201910053757A CN109886302A CN 109886302 A CN109886302 A CN 109886302A CN 201910053757 A CN201910053757 A CN 201910053757A CN 109886302 A CN109886302 A CN 109886302A
- Authority
- CN
- China
- Prior art keywords
- pipeline
- model
- image
- discriminant function
- obtains
- 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 present invention provides a kind of caliber judgment method and terminal device based on machine learning, image corresponding to the pipeline of each model in the pipeline of preset m kind model is obtained respectively, obtain image set, wherein, for the pipeline of any model in the pipeline of preset m kind model, the pipeline of model corresponds to multiple images, and the pipe diameter size of the pipeline of any two kinds of models of the pipeline kind of m kind model is different；For any image in image set, label corresponding to image is obtained, label is used to indicate the model of pipeline in image, obtains training set；Default disaggregated model is trained by training set, obtains classifier；The image of pipeline to be sorted is inputted into classifier, obtains pipe diameter size corresponding to the model of pipeline in image.By the present invention in that being trained with the pipeline image of a large amount of different tube diameters sizes to disaggregated model, classifier is obtained, is classified by image of the classifier to the pipeline of pipe diameter size to be determined, obtains the pipe diameter size of pipeline in real time.
Description
Technical field
The invention belongs to field of computer technology more particularly to a kind of caliber judgment methods and terminal based on machine learning
Equipment.
Background technique
Ductile iron pipe according to the needs of practical application, is divided into different pipe diameters.As diameter be 300mm, 350mm,
400mm, 500mm etc. different standard specifications.Since tube surface needs to spray anti-decaying paint, different pipe diameter spray anti-decaying paints
Technological parameter is different, needs to judge online the diameter of pipe, to automatically adjust technological parameter, realizes automated production.But it is existing
There is technology not yet to provide a kind of feasible method, real-time judge can be carried out to pipe diameter size.
Summary of the invention
In view of this, the caliber judgment method that the embodiment of the invention provides a kind of based on machine learning, computer-readable
Storage medium and terminal device, to solve the problems, such as that the prior art can not determine pipe diameter size in real time.
The first aspect of the embodiment of the present invention provides a kind of caliber judgment method based on machine learning, comprising:
Image corresponding to the pipeline of each model in the pipeline of preset m kind model is obtained respectively, obtains image set,
Wherein, for the pipeline of any model in the pipeline of preset m kind model, the pipeline of the model corresponds to multiple images,
The pipe diameter size of the pipeline of any two kinds of models of the pipeline kind of the m kind model is different；
For any image that described image is concentrated, label corresponding to described image is obtained, the label is for indicating
The model of pipeline, obtains training set in described image；
Default disaggregated model is trained by the training set, obtains classifier；
The image of pipeline to be sorted is inputted into the classifier, obtains the model of pipeline in described image, and according to described
The model of pipeline obtains the pipe diameter size of the pipeline.
The second aspect of the embodiment of the present invention provides a kind of computer readable storage medium, the computer-readable storage
Media storage has computer-readable instruction, and the computer-readable instruction realizes following steps when being executed by processor:
Image corresponding to the pipeline of each model in the pipeline of preset m kind model is obtained respectively, obtains image set,
Wherein, for the pipeline of any model in the pipeline of preset m kind model, the pipeline of the model corresponds to multiple images,
The pipe diameter size of the pipeline of any two kinds of models of the pipeline kind of the m kind model is different；
For any image that described image is concentrated, label corresponding to described image is obtained, the label is for indicating
The model of pipeline, obtains training set in described image；
Default disaggregated model is trained by the training set, obtains classifier；
The image of pipeline to be sorted is inputted into the classifier, obtains the model of pipeline in described image, and according to described
The model of pipeline obtains the pipe diameter size of the pipeline.
The third aspect of the embodiment of the present invention provides a kind of terminal device, including memory, processor and is stored in
In the memory and the computer-readable instruction that can run on the processor, the processor executes the computer can
Following steps are realized when reading instruction:
Image corresponding to the pipeline of each model in the pipeline of preset m kind model is obtained respectively, obtains image set,
Wherein, for the pipeline of any model in the pipeline of preset m kind model, the pipeline of the model corresponds to multiple images,
The pipe diameter size of the pipeline of any two kinds of models of the pipeline kind of the m kind model is different；
For any image that described image is concentrated, label corresponding to described image is obtained, the label is for indicating
The model of pipeline, obtains training set in described image；
Default disaggregated model is trained by the training set, obtains classifier；
The image of pipeline to be sorted is inputted into the classifier, obtains the model of pipeline in described image, and according to described
The model of pipeline obtains the pipe diameter size of the pipeline.
The present invention provides a kind of caliber judgment method and terminal device based on machine learning, comprising: obtain respectively pre-
If m kind model pipeline in each model pipeline corresponding to image, obtain image set, wherein be directed to preset m kind
The pipeline of any model in the pipeline of model, the pipeline of model correspond to multiple images, and any two kinds of the pipeline kind of m kind model
The pipe diameter size of the pipeline of model is different；For any image in image set, label corresponding to image is obtained, label is used for
The model for indicating pipeline in image, obtains training set；Default disaggregated model is trained by training set, obtains classifier；
The image of pipeline to be sorted is inputted into classifier, obtains pipe diameter size corresponding to the model of pipeline in image.The present invention passes through
Disaggregated model is trained using the pipeline image of a large amount of different tube diameters sizes, obtains classifier, is treated really by classifier
The image of the pipeline of fixed-caliber size is classified, and obtains the pipe diameter size of pipeline in real time.
Detailed description of the invention
It to describe the technical solutions in the embodiments of the present invention more clearly, below will be to embodiment or description of the prior art
Needed in attached drawing be briefly described, it should be apparent that, the accompanying drawings in the following description is only of the invention some
Embodiment for those of ordinary skill in the art without any creative labor, can also be according to these
Attached drawing obtains other attached drawings.
Fig. 1 is a kind of flow diagram of the caliber judgment method based on machine learning provided in an embodiment of the present invention；
Fig. 2 is a kind of structural block diagram of the caliber judgment means based on machine learning provided in an embodiment of the present invention；
Fig. 3 is the schematic diagram that a kind of caliber based on machine learning provided in an embodiment of the present invention judges terminal device.
Specific embodiment
In being described below, for illustration and not for limitation, the tool of such as particular system structure, technology etc is proposed
Body details, to understand thoroughly the embodiment of the present invention.However, it will be clear to one skilled in the art that there is no these specific
The present invention also may be implemented in the other embodiments of details.In other situations, it omits to well-known system, device, electricity
The detailed description of road and method, in case unnecessary details interferes description of the invention.
In order to illustrate technical solutions according to the invention, the following is a description of specific embodiments.
The caliber judgment method based on machine learning that the embodiment of the invention provides a kind of.In conjunction with Fig. 1, this method comprises:
S101 obtains image corresponding to the pipeline of each model in the pipeline of preset m kind model respectively, obtains figure
Image set, wherein for the pipeline of any model in the pipeline of preset m kind model, the pipeline of the model corresponds to multiple
The pipe diameter size of image, the pipeline of any two kinds of models of the pipeline kind of the m kind model is different.
For example, the pipeline of the m kind model corresponds to the pipeline of 4 kinds of models, the respectively pipeline of diameter 300mm altogether,
Diameter is the pipeline of 350mm, the pipeline that the pipeline and diameter that diameter is 400mm are 500mm.
Diameter is the model model 1 of the pipeline of 300mm, and diameter is the model model 2 of the pipeline of 350mm, and diameter is
The model model 3 of the pipeline of 400mm, diameter are the model model 4 of the pipeline of 500mm.
For the pipeline of each model, Image Acquisition is carried out by pipeline of the image collecting device to the model, is obtained
Multiple images corresponding to the pipeline of the model.
Optionally, during Image Acquisition, image collecting device is identical at a distance from the pipeline of every kind of model, and every kind
The placement position of the pipeline of model is identical.
S102 obtains label corresponding to described image for any image that described image is concentrated, and the label is used for
The model for indicating pipeline in described image, obtains training set.
Every image is manually marked, is obtained corresponding to the image for indicating the model of pipeline in the image
Label.All acquired images collectively constitute training set after manually marking in step S101.
S103 is trained default disaggregated model by the training set, obtains classifier.
Default disaggregated model is trained by the sample in training set, optionally, which is to support
The disaggregated model of vector machine.
Optionally, the sample in training set is divided into multiple batches, the default disaggregated model is trained in batches,
After all batches all complete once training to the disaggregated model, then an iteration is completed, the precision of model reaches default essence
After spending or reaching preset times to the number of iterations of the training of the model, deconditioning obtains classifier.
The image of pipeline to be sorted is inputted the classifier by S104, the model of pipeline in acquisition described image, and according to
The model of the pipeline obtains the pipe diameter size of the pipeline.
Acquire the image of pipeline to be sorted in real time, optionally, when acquiring the image of pipeline to be sorted, pipeline to be sorted with
In the placement position of the distance of image collecting device and pipeline to be sorted, with step S101 obtain training sample at a distance from and
Placement position is identical.
The image of pipeline to be sorted is inputted into classifier, is classified by image of the classifier to pipeline to be sorted, is classified
As a result it is one of described m kind model, obtains the type of pipeline to be sorted, can be obtained the caliber of pipeline to be sorted.
It optionally, is the precision for improving classification, before the image of pipeline to be sorted is inputted the classifier, the party
Method further include: the image of the pipeline to be sorted is subjected to greyscale transformation, median filtering, enhanced fuzzy and thresholding and is handled.
Optionally, the disaggregated model that disaggregated model is support vector machines is preset, since the disaggregated model of support vector machines is
Two-value classification problem, by constructing m (m-1)/2 discriminant function, obtains the default classification when pipeline model is greater than 2 kinds
Model, wherein for the pipeline of any model in the pipeline of preset m kind model, the pipeline of the model is corresponding (m-1)
A discriminant function, for any discriminant function, the discriminant function is used to differentiate two kinds of models in the m kind model, arbitrarily
Two kinds of models corresponding to two discriminant functions are not exactly the same.
For example, as described in rapid S101, to model 4, corresponding pipe diameter size is followed successively by the model model 1 of pipeline
300mm, 350mm, 400mm and 500mm construct 6 discriminant functions, first discriminant function is for sentencing then according to the above method
The image of other Class1 and the pipeline of type 2, second discriminant function are used to differentiate the image of the pipeline of Class1 and type 3, the
Three discriminant functions are used to differentiate the image of the pipeline of Class1 and type 4, and the 4th discriminant function is for differentiating type 2 and class
The image of the pipeline of type 3, the 5th discriminant function are used to differentiate the image of the pipeline of type 2 and type 4, the 6th discriminant function
For differentiating the image of the pipeline of type 3 and type 4.
Further, carrying out classification by image of the disaggregated model as described above to pipeline to be sorted includes:
Each discriminant function in respectively described m (m-1)/2 discriminant function is arranged priority, and according to priority by
Small sequence is arrived greatly to be ranked up the m (m-1)/2 discriminant function, obtains discriminant function queue；
Differentiated by image of the discriminant function in the discriminant function queue to the pipeline to be sorted, obtains institute
State the model of pipeline in image.
Specifically, successively calling the differentiation letter in queue according to the sequence of the discriminant function in the discriminant function queue
Several images to the pipeline to be sorted differentiate, if first in queue discriminant function can not be to the pipeline to be sorted
Image differentiated, then call second in queue discriminant function to differentiate the image of the pipeline to be sorted, directly
The first discriminant function of differentiation result is obtained to determining；
If first discriminant function to the differentiation result of the image of the pipeline to be sorted be in the m kind model the
One model, it is determined that the second model corresponding to first discriminant function, and the institute in the discriminant function queue is useful
It is deleted in the discriminant function for differentiating second model, carries out the by image of the remaining discriminant function to the pipeline to be sorted
Two wheels differentiate, until obtaining the type of the pipeline in the image of the pipeline to be sorted.
For example, priority, the priority > second of first discriminant function is arranged in respectively above-mentioned 6 discriminant functions
Priority > the 5th of the 4th discriminant function of priority > of the priority > third discriminant function of a discriminant function sentences
The priority of the 6th discriminant function of priority > of other function.
Discriminant function queue is then generated, according to the sequence of priority from high to low, the discriminant function in queue is followed successively by
One discriminant function, second discriminant function, third discriminant function, the 4th discriminant function, the 5th discriminant function and
Six discriminant functions.
When the image to pipeline to be sorted differentiates, first discriminant function is called first, due to first differentiation
Function is used to differentiate the pipeline of Class1 and type 2, therefore possible there are three types of differentiation results, i.e., the type of pipeline to be sorted is class
Type 1, the type of pipeline to be sorted are type 2 or differentiation failure.If differentiating failure, first discriminant function can not be treated
The image of classification pipeline is differentiated, next discriminant function in team is called to differentiate the image of pipeline to be sorted,
Until obtaining can be derived that the discriminant function of conduit types in image.
In embodiments of the present invention, it is with the differentiation result of the image of first discriminant function you can get it pipeline to be sorted
Example is illustrated.
Assuming that first discriminant function is Class1 to the differentiation result of the image of pipeline to be sorted, then will own in queue
For differentiate type 2 discriminant function delete, remaining discriminant function include second discriminant function, for differentiate Class1 and
Type 3；Third discriminant function, for differentiating Class1 and type 4；6th discriminant function, for differentiating type 3 and type
4；
The second wheel is carried out later to differentiate, is differentiated first by second discriminant function, if differentiation result is Class1,
The discriminant function for differentiating type 3 is deleted, that is, deletes the 6th discriminant function, is finally differentiated by second discriminant function wait divide
The image of pipeline obtains differentiating result.
The caliber judgment method based on machine learning that the embodiment of the invention provides a kind of, this method comprises: obtaining respectively
Image corresponding to the pipeline of each model, obtains image set in the pipeline of preset m kind model, wherein is directed to preset m
The pipeline of any model in the pipeline of kind model, the pipeline of model correspond to multiple images, the pipeline kind any two of m kind model
The pipe diameter size of the pipeline of kind model is different；For any image in image set, label corresponding to image is obtained, label is used
In the model for indicating pipeline in image, training set is obtained；Default disaggregated model is trained by training set, is classified
Device；The image of pipeline to be sorted is inputted into classifier, obtains pipe diameter size corresponding to the model of pipeline in image.The present invention is logical
It crosses and disaggregated model is trained using the pipeline image of a large amount of different tube diameters sizes, obtain classifier, treated by classifier
It determines that the image of the pipeline of pipe diameter size is classified, obtains the pipe diameter size of pipeline in real time.
Fig. 2 is a kind of caliber judgment means schematic diagram based on machine learning provided in an embodiment of the present invention, in conjunction with Fig. 2,
The device includes: image acquisition unit 21, training set acquiring unit 22, training unit 23 and taxon 24；
Image acquisition unit 21, the pipeline institute of each model is right in the pipeline for obtaining preset m kind model respectively
The image answered, obtains image set, wherein for the pipeline of any model in the pipeline of preset m kind model, the model
Pipeline correspond to multiple images, the pipe diameter size of the pipeline of any two kinds of models of the pipeline kind of the m kind model is different；
Training set acquiring unit 22, any image for concentrating for described image, obtains corresponding to described image
Label, the label are used to indicate the model of pipeline in described image, obtain training set；
Training unit 23 obtains classifier for being trained by the training set to default disaggregated model；
Taxon 24 obtains pipeline in described image for the image of pipeline to be sorted to be inputted the classifier
Model, and the pipe diameter size of the pipeline is obtained according to the model of the pipeline.
Further, the taxon 24 is also used to:
The image of the pipeline to be sorted is subjected to greyscale transformation, median filtering, enhanced fuzzy and thresholding processing.
Further, the training unit 23 is also used to:
M (m-1)/2 discriminant function is constructed, obtains the default disaggregated model, wherein for preset m kind model
The pipeline of any model in pipeline, corresponding (m-1) a discriminant function of the pipeline of the model, for any discriminant function,
The discriminant function is used to differentiate two kinds of models in the m kind model, two kinds of models corresponding to any two discriminant function
It is not exactly the same.
Further, the taxon 24 is also used to:
Each discriminant function in respectively described m (m-1)/2 discriminant function is arranged priority, and according to priority by
Small sequence is arrived greatly to be ranked up the m (m-1)/2 discriminant function, obtains discriminant function queue；
Differentiated by image of the discriminant function in the discriminant function queue to the pipeline to be sorted, obtains institute
State the model of pipeline in image.
Further, the taxon 24 is also used to:
According to the sequence of the discriminant function in the discriminant function queue, successively call the discriminant function in queue to described
The image of pipeline to be sorted differentiated, if first in queue discriminant function can not image to the pipeline to be sorted into
Row differentiates, then second in queue discriminant function is called to differentiate the image of the pipeline to be sorted, until determining
To the first discriminant function for differentiating result；
If first discriminant function to the differentiation result of the image of the pipeline to be sorted be in the m kind model the
One model, it is determined that the second model corresponding to first discriminant function, and the institute in the discriminant function queue is useful
It is deleted in the discriminant function for differentiating second model, carries out the by image of the remaining discriminant function to the pipeline to be sorted
Two wheels differentiate, until obtaining the type of the pipeline in the image of the pipeline to be sorted.
The caliber judgment means based on machine learning that the embodiment of the invention provides a kind of, the device are pre- by obtaining respectively
If m kind model pipeline in each model pipeline corresponding to image, obtain image set, wherein be directed to preset m kind
The pipeline of any model in the pipeline of model, the pipeline of model correspond to multiple images, and any two kinds of the pipeline kind of m kind model
The pipe diameter size of the pipeline of model is different；For any image in image set, label corresponding to image is obtained, label is used for
The model for indicating pipeline in image, obtains training set；Default disaggregated model is trained by training set, obtains classifier；
The image of pipeline to be sorted is inputted into classifier, obtains pipe diameter size corresponding to the model of pipeline in image.The device passes through
Disaggregated model is trained using the pipeline image of a large amount of different tube diameters sizes, obtains classifier, is treated really by classifier
The image of the pipeline of fixed-caliber size is classified, and obtains the pipe diameter size of pipeline in real time.
Fig. 3 is a kind of schematic diagram of terminal device provided in an embodiment of the present invention.As shown in figure 3, the terminal of the embodiment
Equipment 3 includes: processor 30, memory 31 and is stored in the memory 31 and can run on the processor 30
Computer program 32, such as the caliber determining program based on machine learning.The processor 30 executes the computer program 32
Step in the above-mentioned each caliber judgment method embodiment based on machine learning of Shi Shixian, for example, step 101 shown in FIG. 1 to
104.Alternatively, the processor 30 realizes each module/unit in above-mentioned each Installation practice when executing the computer program 32
Function, such as the function of module 21 to 24 shown in Fig. 2.
Illustratively, the computer program 32 can be divided into one or more module/units, it is one or
Multiple module/units are stored in the memory 31, and are executed by the processor 30, to complete the present invention.Described one
A or multiple module/units can be the series of computation machine program instruction section that can complete specific function, which is used for
Implementation procedure of the computer program 32 in the terminal device 3 is described.
The terminal device 3 can be the calculating such as desktop PC, notebook, palm PC and cloud server and set
It is standby.The terminal device may include, but be not limited only to, processor 30, memory 31.It will be understood by those skilled in the art that Fig. 3
The only example of terminal device 3 does not constitute the restriction to terminal device 3, may include than illustrating more or fewer portions
Part perhaps combines certain components or different components, such as the terminal device can also include input-output equipment, net
Network access device, bus etc..
The processor 30 can be central processing unit (Central Processing Unit, CPU), can also be
Other general processors, digital signal processor (Digital Signal Processor, DSP), specific integrated circuit
(Application Specific Integrated Circuit, ASIC), field programmable gate array (Field-
Programmable Gate Array, FPGA) either other programmable logic device, discrete gate or transistor logic,
Discrete hardware components etc..General processor can be microprocessor or the processor is also possible to any conventional processor
Deng.
The memory 31 can be the internal storage unit of the terminal device 3, such as the hard disk or interior of terminal device 3
It deposits.The memory 31 is also possible to the External memory equipment of the terminal device 3, such as be equipped on the terminal device 3
Plug-in type hard disk, intelligent memory card (Smart Media Card, SMC), secure digital (Secure Digital, SD) card dodge
Deposit card (Flash Card) etc..Further, the memory 31 can also both include the storage inside list of the terminal device 3
Member also includes External memory equipment.The memory 31 is for storing needed for the computer program and the terminal device
Other programs and data.The memory 31 can be also used for temporarily storing the data that has exported or will export.
The embodiment of the present invention also provides a kind of computer readable storage medium, and the computer-readable recording medium storage has
Computer program, the computer program realize the pipe described in any of the above-described embodiment based on machine learning when being executed by processor
The step of diameter judgment method.
It, can also be in addition, the functional units in various embodiments of the present invention may be integrated into one processing unit
It is that each unit physically exists alone, can also be integrated in one unit with two or more units.Above-mentioned integrated list
Member both can take the form of hardware realization, can also realize in the form of software functional units.
If the integrated unit is realized in the form of SFU software functional unit and sells or use as independent product
When, it can store in a computer readable storage medium.Based on this understanding, technical solution of the present invention is substantially
The all or part of the part that contributes to existing technology or the technical solution can be in the form of software products in other words
It embodies, which is stored in a storage medium, including some instructions are used so that a computer
Equipment (can be personal computer, server or the network equipment etc.) executes the complete of each embodiment the method for the present invention
Portion or part steps.And storage medium above-mentioned includes: USB flash disk, mobile hard disk, read-only memory (ROM, Read-Only
Memory), random access memory (RAM, Random Access Memory), magnetic or disk etc. are various can store journey
The medium of sequence code.
Embodiment described above is merely illustrative of the technical solution of the present invention, rather than its limitations；Although referring to aforementioned reality
Applying example, invention is explained in detail, those skilled in the art should understand that: it still can be to aforementioned each
Technical solution documented by embodiment is modified or equivalent replacement of some of the technical features；And these are modified
Or replacement, the essence of corresponding technical solution is departed from the spirit and scope of the technical scheme of various embodiments of the present invention, it should all
It is included within protection scope of the present invention.
Claims (10)
1. a kind of caliber judgment method based on machine learning, which is characterized in that this method comprises:
Image corresponding to the pipeline of each model in the pipeline of preset m kind model is obtained respectively, obtains image set,
In, for the pipeline of any model in the pipeline of preset m kind model, the pipeline of the model corresponds to multiple images, institute
The pipe diameter size for stating the pipeline of any two kinds of models of pipeline kind of m kind model is different；
For any image that described image is concentrated, label corresponding to described image is obtained, the label is for indicating described
The model of pipeline, obtains training set in image；
Default disaggregated model is trained by the training set, obtains classifier；
The image of pipeline to be sorted is inputted into the classifier, obtains the model of pipeline in described image, and according to the pipeline
Model obtain the pipe diameter size of the pipeline.
2. the caliber judgment method according to claim 1 based on machine learning, which is characterized in that described by pipe to be sorted
The image in road inputs before the classifier, this method further include:
The image of the pipeline to be sorted is subjected to greyscale transformation, median filtering, enhanced fuzzy and thresholding processing.
3. the caliber judgment method according to claim 1 or 2 based on machine learning, which is characterized in that this method is also wrapped
It includes:
M (m-1)/2 discriminant function is constructed, obtains the default disaggregated model, wherein for the pipeline of preset m kind model
In any model pipeline, corresponding (m-1) a discriminant function of the pipeline of the model is described for any discriminant function
Discriminant function is used to differentiate that two kinds of models in the m kind model, two kinds of models corresponding to any two discriminant function to be endless
It is exactly the same.
4. the caliber judgment method according to claim 3 based on machine learning, which is characterized in that described by pipe to be sorted
The image in road inputs the classifier, and the model for obtaining pipeline in described image includes:
Each discriminant function in respectively described m (m-1)/2 discriminant function is arranged priority, and according to priority by greatly to
Small sequence is ranked up the m (m-1)/2 discriminant function, obtains discriminant function queue；
Differentiated by image of the discriminant function in the discriminant function queue to the pipeline to be sorted, obtains the figure
The model of pipeline as in.
5. the caliber judgment method according to claim 4 based on machine learning, which is characterized in that described to be sentenced by described
Discriminant function in other function queue differentiates the image of the pipeline to be sorted, obtains the model of pipeline in described image
Include:
According to the sequence of the discriminant function in the discriminant function queue, successively call the discriminant function in queue to described wait divide
The image of class pipeline is differentiated, if first in queue discriminant function can not sentence the image of the pipeline to be sorted
Not, then second discriminant function in queue is called to differentiate the image of the pipeline to be sorted, until determination is sentenced
First discriminant function of other result；
If first discriminant function is to the first type that the differentiation result of the image of the pipeline to be sorted is in the m kind model
Number, it is determined that the second model corresponding to first discriminant function, and all in the discriminant function queue are used to sentence
The discriminant function of not described second model is deleted, and carries out the second wheel by image of the remaining discriminant function to the pipeline to be sorted
Differentiate, until obtaining the type of the pipeline in the image of the pipeline to be sorted.
6. a kind of computer readable storage medium, the computer-readable recording medium storage has computer program, and feature exists
In when the computer program is executed by processor the step of any one of such as claim 1 to 5 of realization the method.
7. a kind of terminal device, which is characterized in that the terminal device includes memory, processor, is stored on the memory
There is the computer program that can be run on the processor, the processor realizes following step when executing the computer program
It is rapid:
Image corresponding to the pipeline of each model in the pipeline of preset m kind model is obtained respectively, obtains image set,
In, for the pipeline of any model in the pipeline of preset m kind model, the pipeline of the model corresponds to multiple images, institute
The pipe diameter size for stating the pipeline of any two kinds of models of pipeline kind of m kind model is different；
For any image that described image is concentrated, label corresponding to described image is obtained, the label is for indicating described
The model of pipeline, obtains training set in image；
Default disaggregated model is trained by the training set, obtains classifier；
The image of pipeline to be sorted is inputted into the classifier, obtains the model of pipeline in described image, and according to the pipeline
Model obtain the pipe diameter size of the pipeline.
8. terminal device according to claim 7, which is characterized in that the terminal device is also used to:
The image of the pipeline to be sorted is subjected to greyscale transformation, median filtering, enhanced fuzzy and thresholding processing.
9. terminal device according to claim 7 or 8, which is characterized in that the terminal device is also used to:
M (m-1)/2 discriminant function is constructed, obtains the default disaggregated model, wherein for the pipeline of preset m kind model
In any model pipeline, corresponding (m-1) a discriminant function of the pipeline of the model is described for any discriminant function
Discriminant function is used to differentiate two kinds of models in the m kind model.
10. terminal device according to claim 9, which is characterized in that the terminal device is used for:
Each discriminant function in respectively described m (m-1)/2 discriminant function is arranged priority, and according to priority by greatly to
Small sequence is ranked up the m (m-1)/2 discriminant function, obtains discriminant function queue；
Differentiated by image of the discriminant function in the discriminant function queue to the pipeline to be sorted, obtains the figure
The model of pipeline as in.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910053757.XA CN109886302A (en) | 2019-01-21 | 2019-01-21 | Caliber judgment method and terminal device based on machine learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910053757.XA CN109886302A (en) | 2019-01-21 | 2019-01-21 | Caliber judgment method and terminal device based on machine learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109886302A true CN109886302A (en) | 2019-06-14 |
Family
ID=66926337
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910053757.XA Pending CN109886302A (en) | 2019-01-21 | 2019-01-21 | Caliber judgment method and terminal device based on machine learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109886302A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114585443A (en) * | 2019-10-31 | 2022-06-03 | 美国西门子医学诊断股份有限公司 | Apparatus and method for training a diagnostic analyzer model |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106022518A (en) * | 2016-05-17 | 2016-10-12 | 清华大学 | Pipe damage probability prediction method based on BP neural network |
| CN107464245A (en) * | 2017-06-29 | 2017-12-12 | 北京捷通华声科技股份有限公司 | A kind of localization method and device at picture structure edge |
| CN107886500A (en) * | 2017-10-13 | 2018-04-06 | 北京邮电大学 | A kind of production monitoring method and system based on machine vision and machine learning |
| CN107886133A (en) * | 2017-11-29 | 2018-04-06 | 南京市测绘勘察研究院股份有限公司 | A kind of underground piping defect inspection method based on deep learning |
| CN108009175A (en) * | 2016-10-28 | 2018-05-08 | 中国石油天然气股份有限公司 | A kind of detection method and detection device of oil-gas pipeline pit |
| CN108334842A (en) * | 2018-02-02 | 2018-07-27 | 成都国铁电气设备有限公司 | A method of identification pantograph-catenary current collection arcing size |
| WO2018138579A1 (en) * | 2017-01-25 | 2018-08-02 | Ucb Biopharma Sprl | Method and system for predicting optimal epilepsy treatment regimes |
| CN108375353A (en) * | 2017-12-14 | 2018-08-07 | 北京木业邦科技有限公司 | A kind of log detection method, apparatus and equipment |
| CN108982522A (en) * | 2018-08-09 | 2018-12-11 | 北京百度网讯科技有限公司 | Method and apparatus for detecting defect of pipeline |
-
2019
- 2019-01-21 CN CN201910053757.XA patent/CN109886302A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106022518A (en) * | 2016-05-17 | 2016-10-12 | 清华大学 | Pipe damage probability prediction method based on BP neural network |
| CN108009175A (en) * | 2016-10-28 | 2018-05-08 | 中国石油天然气股份有限公司 | A kind of detection method and detection device of oil-gas pipeline pit |
| WO2018138579A1 (en) * | 2017-01-25 | 2018-08-02 | Ucb Biopharma Sprl | Method and system for predicting optimal epilepsy treatment regimes |
| CN107464245A (en) * | 2017-06-29 | 2017-12-12 | 北京捷通华声科技股份有限公司 | A kind of localization method and device at picture structure edge |
| CN107886500A (en) * | 2017-10-13 | 2018-04-06 | 北京邮电大学 | A kind of production monitoring method and system based on machine vision and machine learning |
| CN107886133A (en) * | 2017-11-29 | 2018-04-06 | 南京市测绘勘察研究院股份有限公司 | A kind of underground piping defect inspection method based on deep learning |
| CN108375353A (en) * | 2017-12-14 | 2018-08-07 | 北京木业邦科技有限公司 | A kind of log detection method, apparatus and equipment |
| CN108334842A (en) * | 2018-02-02 | 2018-07-27 | 成都国铁电气设备有限公司 | A method of identification pantograph-catenary current collection arcing size |
| CN108982522A (en) * | 2018-08-09 | 2018-12-11 | 北京百度网讯科技有限公司 | Method and apparatus for detecting defect of pipeline |
Non-Patent Citations (3)
| Title |
|---|
| TERESA VIDAL-CALLEJA等: "Automatic detection and verification of pipeline construction features with multi-modal data", 《2014 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS》 * |
| 周伟峰等: "SVM与组合矩在工件识别中的应用研究", 《重庆工商大学学报（自然科学版）》 * |
| 韩小明等: "基于大数据和神经网络的管道完整性预测方法", 《油气储运》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114585443A (en) * | 2019-10-31 | 2022-06-03 | 美国西门子医学诊断股份有限公司 | Apparatus and method for training a diagnostic analyzer model |
| CN114585443B (en) * | 2019-10-31 | 2023-11-03 | 美国西门子医学诊断股份有限公司 | Apparatus and method for training diagnostic analyzer model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Vivekanandam | Design an adaptive hybrid approach for genetic algorithm to detect effective malware detection in android division | |
| CN108229341A (en) | Sorting technique and device, electronic equipment, computer storage media, program | |
| CN110968701A (en) | Relationship map establishing method, device and equipment for graph neural network | |
| CN111126481A (en) | Training method and device of neural network model | |
| Wang et al. | A computer vision based machine learning approach for fatigue crack initiation sites recognition | |
| CN114821217B (en) | Image recognition method and device based on quantum classical hybrid neural network | |
| CN109753498A (en) | data cleaning method and terminal device based on machine learning | |
| CN111914939A (en) | Method, device and equipment for identifying blurred image and computer readable storage medium | |
| CN110992307A (en) | Insulator positioning and identifying method and device based on YOLO | |
| Li et al. | Detail preservation and feature refinement for object detection | |
| CN109583367A (en) | Image text row detection method and device, storage medium and electronic equipment | |
| CN115859302A (en) | Source code vulnerability detection method, device, equipment and storage medium | |
| CN109886302A (en) | Caliber judgment method and terminal device based on machine learning | |
| Jin et al. | End Image Defect Detection of Float Glass Based on Faster Region-Based Convolutional Neural Network. | |
| Fu et al. | Region‐based fully convolutional networks with deformable convolution and attention fusion for steel surface defect detection in industrial Internet of Things | |
| CN108509876A (en) | For the object detecting method of video, device, equipment, storage medium and program | |
| CN107909414A (en) | The anti-cheat method and device of application program | |
| CN111353577B (en) | Multi-task-based cascade combination model optimization method and device and terminal equipment | |
| CN114359300B (en) | Optimization method, device and system of image segmentation model and storage medium | |
| CN109446229B (en) | Big data and deep learning based identification method and robot system | |
| CN110472246A (en) | Work order classification method, device and storage medium | |
| Eldosoky et al. | Deep transfer learning for wall bulge endpoints regression for autonomous decoration robots | |
| CN113705567B (en) | Ship crack detection method, system, equipment and computer readable storage medium | |
| CN110390096A (en) | A kind of park evaluation method and device | |
| Andriosopoulou et al. | Defect Recognition in High-Pressure Die-Casting Parts Using Neural Networks and Transfer Learning |
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: 20190614 |
|
| RJ01 | Rejection of invention patent application after publication |