CN108346153A - The machine learning of defects in timber and restorative procedure, device, system, electronic equipment - Google Patents
The machine learning of defects in timber and restorative procedure, device, system, electronic equipment Download PDFInfo
- Publication number
- CN108346153A CN108346153A CN201810241515.9A CN201810241515A CN108346153A CN 108346153 A CN108346153 A CN 108346153A CN 201810241515 A CN201810241515 A CN 201810241515A CN 108346153 A CN108346153 A CN 108346153A
- Authority
- CN
- China
- Prior art keywords
- timber
- defects
- repaired
- defect
- image
- 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
- 230000007547 defect Effects 0.000 title claims abstract description 243
- 238000000034 method Methods 0.000 title claims abstract description 89
- 238000010801 machine learning Methods 0.000 title claims abstract description 35
- 230000008439 repair process Effects 0.000 claims abstract description 50
- 238000011084 recovery Methods 0.000 claims abstract description 13
- 230000008569 process Effects 0.000 claims description 37
- 238000003860 storage Methods 0.000 claims description 12
- 238000003384 imaging method Methods 0.000 claims description 11
- 238000012549 training Methods 0.000 claims description 9
- 239000002023 wood Substances 0.000 abstract description 33
- 238000013461 design Methods 0.000 abstract description 3
- 238000010586 diagram Methods 0.000 description 12
- 238000012545 processing Methods 0.000 description 12
- 230000006870 function Effects 0.000 description 7
- 238000005286 illumination Methods 0.000 description 7
- 230000006854 communication Effects 0.000 description 6
- 239000000463 material Substances 0.000 description 6
- 238000004891 communication Methods 0.000 description 5
- 230000002950 deficient Effects 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 5
- 230000035699 permeability Effects 0.000 description 5
- 238000004590 computer program Methods 0.000 description 4
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000006399 behavior Effects 0.000 description 2
- 230000014759 maintenance of location Effects 0.000 description 2
- 206010039509 Scab Diseases 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 238000004026 adhesive bonding Methods 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 239000007795 chemical reaction product Substances 0.000 description 1
- 238000003776 cleavage reaction Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 238000003708 edge detection Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 239000003292 glue Substances 0.000 description 1
- 230000003450 growing effect Effects 0.000 description 1
- 229910052500 inorganic mineral Inorganic materials 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 239000011707 mineral Substances 0.000 description 1
- 210000005036 nerve Anatomy 0.000 description 1
- 239000000047 product Substances 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 230000007017 scission Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
- G06T7/001—Industrial image inspection using an image reference approach
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
- G06T7/0006—Industrial image inspection using a design-rule based approach
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30161—Wood; Lumber
Abstract
The embodiment of the present disclosure discloses machine learning and restorative procedure, device, system, the electronic equipment of a kind of defects in timber.Wherein, the machine learning method of defects in timber includes：Obtain the image and labeled data of timber；The labeled data includes the defect mark and defect cutting mode mark of the timber；According to the image and labeled data of the timber, defects in timber identification model is trained；The defects in timber identification model defects in timber and exports recovery scenario for identification.The embodiment of the present disclosure, image data by obtaining wood sample completes identification and the defect repair conceptual design of defect, realize the automation of defects in timber reparation, overcome the defect that the inefficiency brought by manual identified in prior art, high labor cost and quality can not ensure, efficiency is improved, cost is saved.
Description
Technical field
This disclosure relates to field of computer technology, and in particular to a kind of machine learning of defects in timber and restorative procedure, dress
It sets, system, electronic equipment.
Background technology
In wood processing field, wood skin is a kind of important rapidoprint, passes through the timber progress to some fast growings
Cutting, can obtain rapidly the arbitrary wood skin material of size, these wood skin materials are carried out following process, can produce gluing
The end products such as plate, Recombinant Wood.However, the limitation grown with timber due to wooden skin process technology itself so that the quality of wood skin
Show certain randomness.This makes wood skin, and as a kind of nonstandard product, there are certain mass changes.These mass changes are led
Cause has some wood skins to need to carry out certain reparation that following process could be used for.One of them is characterized in that the wood skin after processing
Became uneven is even, this is because the randomness that timber itself is grown causes the hardness of timber uneven, when passing through cutter,
The wood skin thickness that the higher part of hardness generates is larger, and quality is higher.However the wood skin thickness that generates of the lower part of hardness compared with
Thin, caliper defects make the wood skin need that by certain repair process following process could be used at this.Timber itself also may be used
Can occur some critical defects due to other factors, for example, small holes caused by worms, scab, mineral line, aberration the defects of, these defects be not by
Caused by cutting, but the directly defect present in log.Wood skin with these defects is also required to by repair process
It could be used for following process.
In traditional operation, the reparation of wood skin is repaired by artificial mode, it is however generally that passes through the side of manual identified
Formula finds defect and the corresponding restorative procedure of defect.Defect part is rejected using cutting tool in fault location, then cuts one
For the sample of repairing, and defect sample and repairing sample are binded by way of bonding, completes mending course.
Invention content
The embodiment of the present disclosure provides machine learning and restorative procedure, device, system, the electronic equipment of a kind of defects in timber.
In a first aspect, providing a kind of machine learning method of defects in timber in the embodiment of the present disclosure, which is characterized in that packet
It includes：
Obtain the image and labeled data of timber；The labeled data includes the defect mark and defect of the timber
Cutting mode marks；
According to the image and labeled data of the timber, defects in timber identification model is trained；The timber lacks
Sunken identification model defects in timber and exports recovery scenario for identification.
Optionally, the image and labeled data of timber are obtained, including：
Obtain the original image of the timber；
After the timber is cut, the intermediate process image of the timber is obtained；
After the timber is repaired, the reparation image of the timber is obtained；
The defect mark is determined according to the original image and the intermediate process image and/or the reparation image；
The defect cutting mode mark is determined according to the intermediate process image and the original image.
Optionally, the labeled data further includes the reparation sample mark of the timber.
Optionally, further include：
The reparation is determined according to the reparation image of the timber and the intermediate process image and/or the original image
Sample marks.
Optionally, the defect cutting mode mark includes the cutting path of defect.
Second aspect, the embodiment of the present disclosure also disclose a kind of restorative procedure of defects in timber, including：
Obtain the original image of timber to be repaired；
The original image of the timber to be repaired is input in advance trained defects in timber identification model, institute is obtained
State the defect recognition result of timber to be repaired；Wherein, when identifying that the timber to be repaired has defect, the wood to be repaired
The defect recognition result of material further includes the defect cutting mode suitable for the defects in timber to be repaired.
Optionally, when identifying that the timber to be repaired has defect, the defect recognition result of the timber to be repaired
Further include the reparation sample identification and reparation parameter for being suitable for repairing the defects in timber to be repaired.
The third aspect, the embodiment of the present disclosure provide a kind of machine learning device of defects in timber, including：
First acquisition module is configured as obtaining the image and labeled data of timber；The labeled data includes described
The defect of timber marks and defect cutting mode mark；
Training module is configured as image and labeled data according to the timber, to defects in timber identification model into
Row training；The defects in timber identification model defects in timber and exports recovery scenario for identification.
Optionally, first acquisition module, including：
First acquisition submodule is configured as obtaining the original image of the timber；
Second acquisition submodule is configured as after the timber is cut, and obtains the intermediate process image of the timber；
Third acquisition submodule is configured as after the timber is repaired, and obtains the reparation image of the timber；
First determination sub-module is configured as according to the original image and the intermediate process image and/or described repaiies
Complex pattern determines the defect mark；
Second determination sub-module is configured as determining the defect according to the intermediate process image and the original image
Cutting mode marks.
Optionally, the labeled data further includes the reparation sample mark of the timber.
Optionally, described device further includes：
Determining module is configured as according to the reparation image of the timber and the intermediate process image and/or the original
Beginning image determines the reparation sample mark.
Optionally, the defect cutting mode mark includes the cutting path of defect.
Fourth aspect, the embodiment of the present disclosure additionally provide a kind of prosthetic device of defects in timber, including：
Second acquisition module is configured as obtaining the original image of timber to be repaired；
Third acquisition module is configured as the original image of the timber to be repaired being input to advance trained timber
In defect recognition model, the defect recognition result of the timber to be repaired is obtained；Wherein, the timber tool to be repaired is being identified
When defective, the defect recognition result of the timber to be repaired further includes the defect cutting side suitable for the defects in timber to be repaired
Formula.
Optionally, when identifying that the timber to be repaired has defect, the defect recognition result of the timber to be repaired
Further include the reparation sample identification and reparation parameter for being suitable for repairing the defects in timber to be repaired.
The function can also execute corresponding software realization by hardware realization by hardware.The hardware or
Software includes one or more modules corresponding with above-mentioned function.
In a possible design, in the structure of the machine learning device of defects in timber and the prosthetic device of defects in timber
Include memory and processor, the memory be used for store one or more support defects in timber machine learning device with
The prosthetic device of defects in timber executes the machine learning device and wood of defects in timber in above-mentioned first aspect and second aspect respectively
The computer instruction of the restorative procedure of material defect, the processor are configurable for executing the calculating stored in the memory
Machine instructs.The machine learning device of the defects in timber and the prosthetic device of defects in timber can also include communication interface, be used for
The machine learning device of defects in timber and the prosthetic device of defects in timber and other equipment or communication.
5th aspect, the embodiment of the present disclosure additionally provide a kind of repair system of defects in timber, including：
Imaging sensor, the original image for obtaining timber to be repaired；
Controller, for identifying whether the timber to be repaired has using trained defects in timber identification model in advance
Defect, and when the timber to be repaired has defect, defect on the timber to be repaired is exported to the reparation executive device
Cutting mode, the reparation sample identification suitable for repairing the defects in timber to be repaired and repair parameter；
Executive device is repaired, repairs sample for being obtained according to the reparation sample identification, and according to the reparation parameter
And it after the cutting mode cuts the defect on the timber to be repaired, is repaired using the reparation sample.
Optionally, the reparation executive device includes：
Sample placement unit is repaired, for capturing the reparation sample according to the reparation sample identification, lays equal stress on to stack and set
In the specified location of the timber to be repaired；The designated position is located at the fault location of the timber to be repaired, and according to institute
It states and repairs parameter determination；
Cutter unit, for being cut the reparation sample and the timber to be repaired according to the cutting mode
It cuts；
Repair unit, for by cutting after the reparation sample be pasted onto the timber to be repaired by cutting position
Place.
Optionally, the reparation parameter includes that the coordinate of the designated position and reparation sample overlapping are placed on institute
State the posture on timber to be repaired.
Optionally, the cutting mode includes two cutting line segment starting position coordinates, the ends to form symmetrical closed curve
Only position coordinates and curvature.
Optionally, the controller is additionally operable to the method according to first aspect and trains to obtain the defects in timber identification
Model.
Optionally, the controller is additionally operable to the defect that the method according to second aspect obtains the timber to be repaired
Recognition result.
Optionally, repair system further includes：
Memory, for storing existing reparation sample set, the existing reparation sample set includes repairing sample image and right
The reparation sample identification answered；
The controller matches the existing reparation sample set according to the defect recognition result, obtains described suitable for repairing
The reparation sample identification of the defects in timber to be repaired.
Optionally, repair system further includes：
Permeability illumination system, for the back throw light from the timber to be repaired so that the light of projection
Enter described image sensor across the timber to be repaired.
6th aspect, the embodiment of the present disclosure provide a kind of electronic equipment, including memory and processor；Wherein, described
Memory is for storing one or more computer instruction, wherein one or more computer instruction is by the processor
It executes to realize the method and step described in first aspect or second aspect.
7th aspect, the embodiment of the present disclosure provides a kind of computer readable storage medium, for storing defects in timber
Computer instruction used in the device of machine learning device and defects in timber, it includes for executing above-mentioned first aspect or second
Computer instruction in aspect involved by the machine learning method of defects in timber or the method for defects in timber.
The technical solution that the embodiment of the present disclosure provides can include the following benefits：
The embodiment of the present disclosure, with Machine self-learning method, utilizes after the image and labeled data for obtaining timber
The image and labeled data of the timber train the artificial intelligence model that can identify defects in timber, the artificial intelligence model
It is not only able to identify the defect of timber, additionally it is possible to export the recovery scenario of defect.The embodiment of the present disclosure, by obtaining timber sample
This image data completes identification and the defect repair conceptual design of defect, realizes the automation of defects in timber reparation, overcomes
The defect that inefficiency, high labor cost and the quality brought by manual identified in prior art can not ensure, improves
Efficiency saves cost.
It should be understood that above general description and following detailed description is only exemplary and explanatory, not
The disclosure can be limited.
Description of the drawings
In conjunction with attached drawing, by the detailed description of following non-limiting embodiment, the other feature of the disclosure, purpose and excellent
Point will be apparent.In the accompanying drawings：
Fig. 1 shows the flow chart of the machine learning method of the defects in timber according to one embodiment of the disclosure；
Fig. 2 shows the flow charts of the step S101 of embodiment according to Fig. 1；
Fig. 3 shows the flow chart of the restorative procedure of the defects in timber according to one embodiment of the disclosure；
Fig. 4 shows the structure diagram of the repair system of the defects in timber according to one embodiment of the disclosure；
Fig. 5 shows the structure diagram of the reparation executive device 403 of embodiment according to Fig.4,；
Fig. 6 shows the schematic diagram of the timber image according to one embodiment of the disclosure；
Fig. 7 is adapted for for realizing that the electronics of the machine learning method of the defects in timber according to one embodiment of the disclosure is set
Standby structural schematic diagram.
Specific implementation mode
Hereinafter, the illustrative embodiments of the disclosure will be described in detail with reference to the attached drawings, so that those skilled in the art can
Easily realize them.In addition, for the sake of clarity, the portion unrelated with description illustrative embodiments is omitted in the accompanying drawings
Point.
In the disclosure, it should be appreciated that the term of " comprising " or " having " etc. is intended to refer to disclosed in this specification
Feature, number, step, behavior, the presence of component, part or combinations thereof, and be not intended to exclude other one or more features,
Number, step, behavior, component, part or combinations thereof there is a possibility that or be added.
It also should be noted that in the absence of conflict, the feature in embodiment and embodiment in the disclosure
It can be combined with each other.The disclosure is described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.
Fig. 1 shows the flow chart of the machine learning method of the defects in timber according to one embodiment of the disclosure.Such as Fig. 1 institutes
Show, the machine learning method of the defects in timber includes the following steps S101-S102：
In step S101, the image and labeled data of timber are obtained；The labeled data includes lacking for the timber
Fall into mark and defect cutting mode mark；
In step s 102, according to the image and labeled data of the timber, defects in timber identification model is instructed
Practice；The defects in timber identification model defects in timber and exports recovery scenario for identification.
In the present embodiment, the image of timber can be obtained by imaging sensor, can obtain timber by imaging sensor
Original image, timber the intermediate process image after defect part is removed when carrying out defect processing and/or after being repaired
Repair image etc..Timber can be the wood skin either plank with complex surface to be repaired.Labeled data may include lacking for timber
Fall into mark and defect cutting mode mark；The defect of timber is noted for whether mark timber has defect, defect cutting side
Formula is noted for, when timber has defect, marking position and cutting path of defect etc..The labeled data of timber can be
Directly show according to scheduled format, can also be by by the original image of same wood sample, cut defect
Reparation image after intermediate process image and timber afterwards is repaired is associated to imply expression.
Fig. 6 shows the image schematic diagram according to the wood sample of the embodiment of the present disclosure.As shown in fig. 6, (a) shows tool
The original image of defective wood sample (b) is shown the intermediate manuscript after the defect part excision in wood sample
Picture (c) shows the reparation sample suitable for repairing wood sample.
In the present embodiment, by the timber image and labeled data of acquisition, a machine learning model can be trained i.e. wooden
Material defect recognition model, the model can identify the defect on timber, additionally it is possible to export the scheme of defect repair, and then can be by
Timber is repaired according to recovery scenario.
The present embodiment can be based on different machine learning models, such as neural network, convolutional neural networks, depth nerve
It is one or more in network, support vector machines, K-means, K-neighbors, decision tree, random forest, Bayesian network
Combination.
In an optional realization method of the present embodiment, as shown in Fig. 2, the step S101, that is, obtain the figure of timber
The step of picture and labeled data, further comprise the steps S201-S202：
In step s 201, the original image of the timber is obtained；
In step S202, after the timber is cut, the intermediate process image of the timber is obtained；
In step S203, after the timber is repaired, the reparation image of the timber is obtained；
In step S204, determined according to the original image and the intermediate process image and/or the reparation image
The defect mark；
In step S205, the defect cutting mode mark is determined according to the intermediate process image and the original image
Note.
In the optional realization method, engineering can be completed by obtaining the timber image during manually repairing timber
Practise acquisition and the mark of model sample data.For example, first, the wood sample before a processing is obtained by imaging sensor
Image, the image is as original sample for obtaining markup information.Later, in operator by traditional approach by original sample
The defects of after excision, obtain one and intermediate process sample.It is obtained in an intermediate processing sample by imaging sensor
Between process image, and by the centre processing sample be associated with original sample.Since wood skin is a kind of thin sheet timber, it is processed
Timber be usually all than relatively thin wood skin or plank, therefore the excision of defect can completely to be passed through among entire sample
It wears, therefore can identify the part of cutting from the image of intermediate processing sample by a simple limb recognition.It will be former
Beginning sample image, intermediate sample image and the cut portion that identifies are associated, so that it may with obtain include defect mark and
The labeled data of cutting mode.Finally, the intermediate processing sample after cutting is repaired using repairing sample in operator
Afterwards, the sample after being repaired can obtain the reparation image of sample after repairing by image sensing device.By repairing for the sample
The reparation image with the timber image of original sample, intermediate process image, repairing sample is associated again, so that it may to be used for
Repair the recovery scenario of the sample.So far, the image data and mending option for being labelled with cutting method have been automatically obtained.Pass through
The image and labeled data of both data, that is, timber train one or two neural network, will obtain for defect recognition and
The machine learning model of automatic cutting and reparation.
In one embodiment, by the way that the original image of same wood sample and processing intermediate image to be associated, machine
Learning model can be obtained labeled data, if the two does not have difference, then it is assumed that defect is not present in the sample.If in processing
Between image and original image there is larger difference, then defect mark will be set as at difference.It in another embodiment, can also be single
Solely defect can also be obtained using intermediate process image to mark, because fault location has obviously after being removed in intermediate process image
Feature, can just can determine that by simple edge detection or pattern match.It is understood that machine learning model can be with
It is automatically learned the defects of labeled data mark based on wood sample, intermediate process image and original image can also be passed through
The cutting mode mark being automatically learned in labeled data.
In one embodiment, labeled data can also include repairing sample mark.It repairs sample and is noted for mark currently
The defect of wood sample is repaired using which kind of reparation sample, can be by being established between wood sample and reparation sample
Matching relationship mark.For example, the matching relationship is the optimal repairing sample picked out according to original sample by artificial mode
This.Incidence relation between texture between the two, color can be used for the machine that training one can carry out repairing screening sample
Device learning model.
In an optional realization method of the present embodiment, the method further includes：
The reparation is determined according to the reparation image of the timber and the intermediate process image and/or the original image
Sample marks.
In the optional realization method, it can directly represent to correspond to have by preset format to lack to repair sample mark
The image of the reparation sample of sunken wood sample can obtain the mark etc. for repairing sample image, can also be to pass through timber
The incidence relation of repairing image and intermediate process image determine, or by the original image of timber, repair image and
What the incidence relation between intermediate process image determined.
In an optional realization method of the present embodiment, the defect cutting mode mark includes the cutting road of defect
Diameter.
In the optional realization method, cutting path can be by individual intermediate process image or original image in
Between incidence relation between process image determine, and indicated by the coordinate at cut portion edge.It is understood that cutting
Path can also be indicated by the way of the specific cutting part that other can be identified for that defect, such as cutting path can be
The elliptical path being made of two sections of cutting curves can pass through starting point, end point and curvature { (a1, b1), (a2, b2), k }
It indicates, mean curvature can also indicate with radius of curvature d.
Fig. 3 shows the flow chart of the restorative procedure of the defects in timber according to one embodiment of the disclosure.As shown in figure 3, institute
The restorative procedure for stating defects in timber includes the following steps S301-S302：
In step S301, the original image of timber to be repaired is obtained；
In step s 302, the original image of the timber to be repaired trained defects in timber in advance are input to identify
In model, the defect recognition result of the timber to be repaired is obtained；Wherein, identifying the timber to be repaired with defect
When, the defect recognition result of the timber to be repaired further includes the defect cutting mode suitable for the defects in timber to be repaired.
In the present embodiment, by training defects in timber identification model in advance, and the original image of timber to be repaired is defeated
Enter into defects in timber identification model, defects in timber identification model exports the defect recognition knot of timber to be repaired by analysis and identification
Fruit.Defect recognition result may include whether timber to be repaired has defective mark, can also include defect cutting mode；It lacks
Sunken cutting mode may include position and cutting path of defect etc..
The present embodiment automatically identifies on timber whether have defect by the machine learning model Jing Guo pre-training, simultaneously
Cutting mode when timber has defect when the output reparation defect.The present embodiment automatic identification defects in timber and can be given
Go out recovery scenario, overcoming needs manually to carry out defect recognition to timber in prior art and formulate the defect of recovery scenario, improves
The efficiency and accuracy rate of timber reparation, saves a large amount of cost of labor.
In one embodiment, when identifying that the timber to be repaired has defect, the defect of the timber to be repaired is known
Other result further includes being suitable for repairing the reparation sample identification and reparation parameter of the defects in timber to be repaired.
In the optional realization method, defects in timber identification model can also be exported to be lacked suitable for repairing the timber to be repaired
Sunken reparation sample, such as repair the mark of sample and repair parameter.For example, can safeguard a database in systems, it is used for
By associated storages such as the existing image for repairing sample and marks, defects in timber identification model is on identifying timber to be repaired
Defect when, be also based on color, texture of wood surface etc. determine be suitable for repair the timber to be repaired reparation sample, and
The existing reparation sample to match is obtained from database by way of images match, and exports reparation sample identification.Meanwhile
Reparation parameter can also be exported.It may include opposite when reparation sample is placed on timber to be repaired at rejected region to repair parameter
Posture and relative coordinate.For example, { sn, θ, x, y } indicates the repairing sample number for current timber to be repaired, opposite appearance
State and relative coordinate；{ (a1, b1), (a2, b2), k } parameter indicates the cutting route for current timber to be repaired.Repairing
Executive device can will repair sample and be placed in the position repaired parameter and indicated, two samples are overlapped and are reinforced.For example, machine
Tool arm first fixes timber to be repaired, and by repair sample rotate to θ angles, further will repair sample be moved to x,
Y } at position, places and repair sample and simultaneously fix two samples.Hereafter, repairing two samples of executive device pair according to cutting path into
Row cutting, and then obtain a repairing sample consistent with cutting profile.
Fig. 4 shows the structure diagram of the repair system 400 of the defects in timber according to one embodiment of the disclosure.Such as Fig. 4 institutes
Show, the repair system 400 of the defects in timber includes：
Imaging sensor 401, the original image for obtaining timber to be repaired；
Controller 402, for whether identifying the timber to be repaired using trained defects in timber identification model in advance
With defect, and when the timber to be repaired has defect, exported on the timber to be repaired to the reparation executive device
The cutting mode of defect, the reparation sample identification suitable for repairing the defects in timber to be repaired and reparation parameter；
Executive device 403 is repaired, repairs sample for being obtained according to the reparation sample identification, and join according to the reparation
The several and described cutting mode is repaired after the defect cutting on the timber to be repaired using the reparation sample.
In the present embodiment, in timber repair process, imaging sensor 401 is by the original image of the timber to be repaired of acquisition
It sends controller 402 to, trained defects in timber identification model can be previously stored in controller 402, and utilize the mould
Original image is identified in type.Controller 402 can be located locally together with imaging sensor, can also be located at high in the clouds.Control
Device 402 processed is waited for when having defect on identifying timber to be repaired by the cutting mode of defect on timber to be repaired, suitable for repairing
It repairs the reparation sample identification of timber and reparation parameter exports and gives reparation executive device 403.
Repair executive device 403 based on the cutting mode of defect on the timber to be repaired obtained, be suitable for repairing it is to be repaired
The reparation sample identification and reparation parameter of timber repair timber to be repaired.
The detail of the training method of defects in timber identification model may refer to embodiment illustrated in fig. 1 and related content
Description, details are not described herein.
Controller can be found in attached using the details that defects in timber identification model identifies defects in timber and provides recovery scenario
The description of embodiment illustrated in fig. 3 and related content, details are not described herein.
In an optional realization method of the present embodiment, as shown in figure 5, reparation executive device 403 includes：
Sample placement unit 501 is repaired, for capturing the reparation sample according to the reparation sample identification, lays equal stress on and stacks
Set the specified location in the timber to be repaired；The designated position is located at the fault location of the timber to be repaired, and according to
The reparation parameter determines；
Cutter unit 502, for being carried out the reparation sample and the timber to be repaired according to the cutting mode
Cutting；
Repair unit 503, for by cutting after the reparation sample be pasted onto the timber to be repaired by cleavage
Set place.
In the optional realization method, repairs executive device 403 and obtain reparation sample identification from controller 402 and repair
After multiple parameter, can timber to be repaired be first fixed on pre-position, control again repair 501 crawl of sample placement unit later
The reparation sample of respective identification, places it at the rejected region of timber to be repaired, and rejected region is provided by reparation parameter.Then
Control cutter unit cuts according to cutting path and repairs sample and timber to be repaired simultaneously, and then has been cut off defect part
The reparation sample of timber and well cutting to be repaired.Finally, the step of unit 503 is by automatic glue application and stickup is repaired, will be cut
The reparation sample cut is pasted onto rejected region on timber to be repaired.
In an optional realization method of the present embodiment, it is described repair parameter include the designated position coordinate and
The posture repaired sample overlapping and be placed on the timber to be repaired.
In the optional realization method, designated position can be the wood to be repaired that defects in timber identification model is identified
The relative coordinate of defective locations on material, posture can repair relative attitude residing when sample is placed on timber to be repaired.
For example, { sn, θ, x, y } indicates repairing sample number, relative attitude and the relative coordinate for current timber to be repaired；
{ (a1, b1), (a2, b2), k } parameter indicates the cutting route for current timber to be repaired.Repairing executive device can incite somebody to action
Repairing sample is placed in the position for repairing parameter instruction, and two samples are overlapped and are reinforced.For example, mechanical arm will wait for first
It repairs timber to fix, and θ angles are rotated to by sample is repaired, will further repair sample and be moved at the position { x, y }, placement is repaiied
Duplicate sample sheet simultaneously fixes two samples.
In an optional realization method of the present embodiment, the cutting mode includes to form symmetrical closed curve two
Cut line segment starting position coordinates, final position coordinate and curvature.
Cutting mode includes cutting path in the optional realization method, and cutting path is a closed curve, such as can be with
The curve being symmetrically closed by two sections is constituted, can any one section of starting position coordinates, final position coordinate in two sections of curves
And curvature indicates.Such as the elliptical path that cutting path can be made of two sections of cutting curves, starting can be passed through
Point, end point and curvature { (a1, b1), (a2, b2), k } indicate that mean curvature can also be indicated with radius of curvature d.
In an optional realization method of the present embodiment, controller can be by executing embodiment illustrated in fig. 1 and correlation
The machine learning method of defects in timber described in content trains to obtain defects in timber identification model, and detail can be found in
The description to Fig. 1 and related content is stated, details are not described herein.
In an optional realization method of the present embodiment, controller can also be by executing embodiment illustrated in fig. 3 and phase
The restorative procedure for holding described defects in timber inside the Pass obtains the defect recognition of the timber to be repaired as a result, detail can join
See the above-mentioned description to Fig. 3 and related content, details are not described herein.
In an optional realization method of the present embodiment, the repair system 400 of the defects in timber further includes：
Memory, for storing existing reparation sample set, the existing reparation sample set includes repairing sample image and right
The reparation sample identification answered；
The controller 402 matches the existing reparation sample set according to the defect recognition result, is suitable for described in acquisition
Repair the reparation sample identification of the defects in timber to be repaired.
In the optional realization method, repair system can also store the local existing figure for repairing sample by memory
As and mark etc., when defect to identify timber to be repaired in controller, from memory by way of images match
Middle matching obtains the reparation sample for being suitable for repairing defects in timber to be repaired, and by reparation executive device 403 according to the reparation of acquisition
Sample identification captures corresponding reparation sample, is repaired to timber to be repaired.
In an optional realization method of the present embodiment, the repair system 400 of the defects in timber further includes：
Permeability illumination system, for the back throw light from the timber to be repaired so that the light of projection
Enter described image sensor across the timber to be repaired.
In the optional realization method, imaging sensor 401 can be used when obtaining the image of timber to be repaired and be penetrated
Property illumination system.Permeability illumination system can control light by light from relatively thin timber back projection and regulating illumination intensity
According to intensity, permeability illumination system can make the light of projection enter imaging sensor 401 across timber.This passability light
Lighting system can make machine learning model capture the thickness information of timber, and then more accurately identify due in uneven thickness
Caused by defects in timber.When lacking this permeability illumination system, the thickness information of timber may be easily detected by exterior view
It is obtained as the extraction of feature.
Following is embodiment of the present disclosure, can be used for executing embodiments of the present disclosure.
The embodiment of the present disclosure additionally provides a kind of machine learning device of defects in timber, including：
First acquisition module is configured as obtaining the image and labeled data of timber；The labeled data includes described
The defect of timber marks and defect cutting mode mark；
Training module is configured as image and labeled data according to the timber, to defects in timber identification model into
Row training；The defects in timber identification model defects in timber and exports recovery scenario for identification.
In an optional realization method of the present embodiment, first acquisition module, including：
First acquisition submodule is configured as obtaining the original image of the timber；
Second acquisition submodule is configured as after the timber is cut, and obtains the intermediate process image of the timber；
Third acquisition submodule is configured as after the timber is repaired, and obtains the reparation image of the timber；
First determination sub-module is configured as according to the original image and the intermediate process image and/or described repaiies
Complex pattern determines the defect mark；
Second determination sub-module is configured as determining the defect according to the intermediate process image and the original image
Cutting mode marks.
In an optional realization method of the present embodiment, the labeled data further includes the reparation sample mark of the timber
Note.
In an optional realization method of the present embodiment, the machine learning device of defects in timber further includes：
Determining module is configured as according to the reparation image of the timber and the intermediate process image and/or the original
Beginning image determines the reparation sample mark.
In an optional realization method of the present embodiment, the defect cutting mode mark includes the cutting road of defect
Diameter.
The embodiment of the present disclosure additionally provides a kind of prosthetic device of defects in timber, including：
Second acquisition module is configured as obtaining the original image of timber to be repaired；
Third acquisition module is configured as the original image of the timber to be repaired being input to advance trained timber
In defect recognition model, the defect recognition result of the timber to be repaired is obtained；Wherein, the timber tool to be repaired is being identified
When defective, the defect recognition result of the timber to be repaired further includes the defect cutting side suitable for the defects in timber to be repaired
Formula.
It is described when identifying that the timber to be repaired has defect in an optional realization method of the present embodiment
The defect recognition result of timber to be repaired further includes being suitable for repairing the reparation sample identification of the defects in timber to be repaired and repairing
Multiple parameter.
Fig. 7 is adapted for the electronic equipment of the machine learning method for realizing the defects in timber according to disclosure embodiment
Structural schematic diagram.
As shown in fig. 7, electronic equipment 700 includes central processing unit (CPU) 701, it can be according to being stored in read-only deposit
Program in reservoir (ROM) 702 is held from the program that storage section 708 is loaded into random access storage device (RAM) 703
Various processing in the above-mentioned embodiment shown in FIG. 1 of row.In RAM703, be also stored with electronic equipment 700 operate it is required
Various programs and data.CPU701, ROM702 and RAM703 are connected with each other by bus 704.Input/output (I/O) interface
705 are also connected to bus 704.
It is connected to I/O interfaces 705 with lower component：Importation 706 including keyboard, mouse etc.；It is penetrated including such as cathode
The output par, c 707 of spool (CRT), liquid crystal display (LCD) etc. and loud speaker etc.；Storage section 708 including hard disk etc.；
And the communications portion 709 of the network interface card including LAN card, modem etc..Communications portion 709 via such as because
The network of spy's net executes communication process.Driver 710 is also according to needing to be connected to I/O interfaces 705.Detachable media 711, such as
Disk, CD, magneto-optic disk, semiconductor memory etc. are mounted on driver 710, as needed in order to be read from thereon
Computer program be mounted into storage section 708 as needed.
Particularly, according to embodiment of the present disclosure, it is soft to may be implemented as computer above with reference to Fig. 1 methods described
Part program.For example, embodiment of the present disclosure includes a kind of computer program product comprising be tangibly embodied in and its readable
Computer program on medium, the computer program include the program code of the method for executing Fig. 1.In such implementation
In mode, which can be downloaded and installed by communications portion 709 from network, and/or from detachable media
711 are mounted.
Electronic equipment shown in Fig. 7 can equally be well applied to realize the reparation side of the defects in timber according to disclosure embodiment
Method.
Flow chart in attached drawing and block diagram, it is illustrated that according to the system, method and computer of the various embodiments of the disclosure
The architecture, function and operation in the cards of program product.In this regard, each box in course diagram or block diagram can be with
A part for a module, section or code is represented, a part for the module, section or code includes one or more
Executable instruction for implementing the specified logical function.It should also be noted that in some implementations as replacements, institute in box
The function of mark can also occur in a different order than that indicated in the drawings.For example, two boxes succeedingly indicated are practical
On can be basically executed in parallel, they can also be executed in the opposite order sometimes, this is depended on the functions involved.Also it wants
It is noted that the combination of each box in block diagram and or flow chart and the box in block diagram and or flow chart, Ke Yiyong
The dedicated hardware based system of defined functions or operations is executed to realize, or can be referred to specialized hardware and computer
The combination of order is realized.
Being described in unit or module involved in disclosure embodiment can be realized by way of software, also may be used
It is realized in a manner of by hardware.Described unit or module can also be arranged in the processor, these units or module
Title do not constitute the restriction to the unit or module itself under certain conditions.
As on the other hand, the disclosure additionally provides a kind of computer readable storage medium, the computer-readable storage medium
Matter can be computer readable storage medium included in device described in the above embodiment；Can also be individualism,
Without the computer readable storage medium in supplying equipment.There are one computer-readable recording medium storages or more than one journey
Sequence, described program is used for executing by one or more than one processor is described in disclosed method.
Above description is only the preferred embodiment of the disclosure and the explanation to institute's application technology principle.People in the art
Member should be appreciated that invention scope involved in the disclosure, however it is not limited to technology made of the specific combination of above-mentioned technical characteristic
Scheme, while should also cover in the case where not departing from the inventive concept, it is carried out by above-mentioned technical characteristic or its equivalent feature
Other technical solutions of arbitrary combination and formation.Such as features described above has similar work(with (but not limited to) disclosed in the disclosure
Can technical characteristic replaced mutually and the technical solution that is formed.
Claims (10)
1. a kind of machine learning method of defects in timber, which is characterized in that including：
Obtain the image and labeled data of timber；The labeled data includes defect mark and the defect cutting of the timber
Mode marks；
According to the image and labeled data of the timber, defects in timber identification model is trained；The defects in timber are known
Other model defects in timber and exports recovery scenario for identification.
2. machine learning method according to claim 1, which is characterized in that the image and labeled data of timber are obtained,
Including：
Obtain the original image of the timber；
After the timber is cut, the intermediate process image of the timber is obtained；
After the timber is repaired, the reparation image of the timber is obtained；
The defect mark is determined according to the original image and the intermediate process image and/or the reparation image；
The defect cutting mode mark is determined according to the intermediate process image and the original image.
3. machine learning method according to claim 2, which is characterized in that the labeled data further includes the timber
Repair sample mark.
4. machine learning method according to claim 3, which is characterized in that further include：
The reparation sample is determined according to the reparation image of the timber and the intermediate process image and/or the original image
Mark.
5. a kind of restorative procedure of defects in timber, which is characterized in that including：
Obtain the original image of timber to be repaired；
The original image of the timber to be repaired is input in advance trained defects in timber identification model, is waited for described in acquisition
Repair the defect recognition result of timber；Wherein, when identifying that the timber to be repaired has defect, the timber to be repaired
Defect recognition result further includes the defect cutting mode suitable for the defects in timber to be repaired.
6. a kind of machine learning device of defects in timber, which is characterized in that including：
First acquisition module is configured as obtaining the image and labeled data of timber；The labeled data includes the timber
Defect mark and defect cutting mode mark；
Training module is configured as image and labeled data according to the timber, is instructed to defects in timber identification model
Practice；The defects in timber identification model defects in timber and exports recovery scenario for identification.
7. a kind of prosthetic device of defects in timber, which is characterized in that including：
Second acquisition module is configured as obtaining the original image of timber to be repaired；
Third acquisition module is configured as the original image of the timber to be repaired being input to advance trained defects in timber
In identification model, the defect recognition result of the timber to be repaired is obtained；Wherein, identifying that it is scarce that the timber to be repaired has
When falling into, the defect recognition result of the timber to be repaired further includes the defect cutting mode suitable for the defects in timber to be repaired.
Suitable for repairing the reparation sample identification of the defects in timber to be repaired and repairing parameter.
8. a kind of repair system of defects in timber, which is characterized in that including：
Imaging sensor, the original image for obtaining timber to be repaired；
Controller is lacked for identifying whether the timber to be repaired has using trained defects in timber identification model in advance
It falls into, and when the timber to be repaired has defect, defect on the timber to be repaired is exported to the reparation executive device
Cutting mode, the reparation sample identification suitable for repairing the defects in timber to be repaired and reparation parameter；
Repair executive device, for obtaining reparation sample according to the reparations sample identification, and according to the reparation parameter and
The cutting mode is repaired after the defect cutting on the timber to be repaired using the reparation sample.
9. a kind of electronic equipment, which is characterized in that including memory and processor；Wherein,
The memory is for storing one or more computer instruction, wherein one or more computer instruction is by institute
Processor is stated to execute to realize claim 1-5 any one of them method and steps.
10. a kind of computer readable storage medium, is stored thereon with computer instruction, which is characterized in that the computer instruction quilt
Claim 1-5 any one of them method and steps are realized when processor executes.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810241515.9A CN108346153A (en) | 2018-03-22 | 2018-03-22 | The machine learning of defects in timber and restorative procedure, device, system, electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810241515.9A CN108346153A (en) | 2018-03-22 | 2018-03-22 | The machine learning of defects in timber and restorative procedure, device, system, electronic equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108346153A true CN108346153A (en) | 2018-07-31 |
Family
ID=62957823
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810241515.9A Pending CN108346153A (en) | 2018-03-22 | 2018-03-22 | The machine learning of defects in timber and restorative procedure, device, system, electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108346153A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109345505A (en) * | 2018-08-09 | 2019-02-15 | 北京木业邦科技有限公司 | A kind of material checks equipment |
| CN109544514A (en) * | 2018-11-05 | 2019-03-29 | 华侨大学 | A kind of sawn timber identity identification method, device and equipment merging performance characteristic |
| CN110363759A (en) * | 2019-07-22 | 2019-10-22 | 国家超级计算天津中心 | Three-dimensional mould tuning parameter determines method and device |
| CN111862028A (en) * | 2020-07-14 | 2020-10-30 | 南京林业大学 | Wood defect detecting and sorting device and method based on depth camera and depth learning |
| CN112330606A (en) * | 2020-10-20 | 2021-02-05 | 西安工程大学 | Defect detection method based on machine learning |
| CN114734513A (en) * | 2022-05-10 | 2022-07-12 | 绍兴昊华木业有限公司 | Device and method for repairing surface defects of wood board |
| CN116559119A (en) * | 2023-05-11 | 2023-08-08 | 东北林业大学 | Deep learning-based wood dyeing color difference detection method, system and medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN205538740U (en) * | 2016-01-29 | 2016-08-31 | 广州番禺职业技术学院 | Intelligence timber surface defect detection system |
| CN107392896A (en) * | 2017-07-14 | 2017-11-24 | 佛山市南海区广工大数控装备协同创新研究院 | A kind of Wood Defects Testing method and system based on deep learning |
-
2018
- 2018-03-22 CN CN201810241515.9A patent/CN108346153A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN205538740U (en) * | 2016-01-29 | 2016-08-31 | 广州番禺职业技术学院 | Intelligence timber surface defect detection system |
| CN107392896A (en) * | 2017-07-14 | 2017-11-24 | 佛山市南海区广工大数控装备协同创新研究院 | A kind of Wood Defects Testing method and system based on deep learning |
Non-Patent Citations (1)
| Title |
|---|
| 戚大伟等: "人工神经网络在木材缺陷检测中的应用", 《森林工程》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109345505A (en) * | 2018-08-09 | 2019-02-15 | 北京木业邦科技有限公司 | A kind of material checks equipment |
| CN109544514A (en) * | 2018-11-05 | 2019-03-29 | 华侨大学 | A kind of sawn timber identity identification method, device and equipment merging performance characteristic |
| CN109544514B (en) * | 2018-11-05 | 2020-07-24 | 华侨大学 | Sawn timber identity identification method, device and equipment integrating apparent features |
| CN110363759A (en) * | 2019-07-22 | 2019-10-22 | 国家超级计算天津中心 | Three-dimensional mould tuning parameter determines method and device |
| CN111862028A (en) * | 2020-07-14 | 2020-10-30 | 南京林业大学 | Wood defect detecting and sorting device and method based on depth camera and depth learning |
| CN112330606A (en) * | 2020-10-20 | 2021-02-05 | 西安工程大学 | Defect detection method based on machine learning |
| CN114734513A (en) * | 2022-05-10 | 2022-07-12 | 绍兴昊华木业有限公司 | Device and method for repairing surface defects of wood board |
| CN114734513B (en) * | 2022-05-10 | 2022-10-04 | 绍兴昊华木业有限公司 | Device and method for repairing surface defects of wood board |
| CN116559119A (en) * | 2023-05-11 | 2023-08-08 | 东北林业大学 | Deep learning-based wood dyeing color difference detection method, system and medium |
| CN116559119B (en) * | 2023-05-11 | 2024-01-26 | 东北林业大学 | Deep learning-based wood dyeing color difference detection method, system and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108346153A (en) | The machine learning of defects in timber and restorative procedure, device, system, electronic equipment | |
| CN109993734A (en) | Method and apparatus for output information | |
| CN109118482A (en) | A kind of panel defect analysis method, device and storage medium | |
| CN108247764B (en) | Wood board cutting method and device based on machine learning, electronic equipment and medium | |
| CN110310259A (en) | It is a kind of that flaw detection method is tied based on the wood for improving YOLOv3 algorithm | |
| JP2019076819A (en) | Sorting system | |
| CN108262809B (en) | Wood board processing method and device based on artificial intelligence, electronic equipment and medium | |
| CN107832780A (en) | Low confidence sample processing method and system are sorted based on artificial intelligence plank | |
| CN110712202B (en) | Special-shaped component grabbing method, device and system, control device and storage medium | |
| CN110646425B (en) | Tobacco leaf online auxiliary grading method and system | |
| WO2020052352A1 (en) | Method and device for damage segmentation of vehicle damage image | |
| CN113083804A (en) | Laser intelligent derusting method and system and readable medium | |
| CN109946304A (en) | Surface defects of parts on-line detecting system and detection method based on characteristic matching | |
| CN107024478A (en) | A kind of part processes automatic defect identifying system | |
| CN108375353A (en) | A kind of log detection method, apparatus and equipment | |
| WO2023052537A1 (en) | Method for detecting at least one defect on a support, device and computer program associated | |
| CN109685756A (en) | Image feature automatic identifier, system and method | |
| US20220036371A1 (en) | Identifying and grading system and related methods for collectable items | |
| CN117022971B (en) | Intelligent logistics stacking robot control system | |
| CN114092935A (en) | Textile fiber identification method based on convolutional neural network | |
| CN113628159A (en) | Full-automatic training method and device based on deep learning network and storage medium | |
| CN114820618A (en) | Defect detection model training method, device, equipment and storage medium | |
| CN105855186A (en) | Automatic sorting device and method for textile materials | |
| CN112581495A (en) | Image processing method, device, equipment and storage medium | |
| JP2004527048A (en) | Identification method of recorded information |
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: 20180731 |
|
| RJ01 | Rejection of invention patent application after publication |