CN106079338B - Embedded mold protection device based on wavelet passivation and image processing method thereof - Google Patents
Embedded mold protection device based on wavelet passivation and image processing method thereof Download PDFInfo
- Publication number
- CN106079338B CN106079338B CN201610459520.8A CN201610459520A CN106079338B CN 106079338 B CN106079338 B CN 106079338B CN 201610459520 A CN201610459520 A CN 201610459520A CN 106079338 B CN106079338 B CN 106079338B
- Authority
- CN
- China
- Prior art keywords
- image
- protection controller
- pixel
- processing
- wavelet
- 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.)
- Active
Links
- 238000002161 passivation Methods 0.000 title claims abstract description 22
- 238000003672 processing method Methods 0.000 title claims abstract description 10
- 238000012545 processing Methods 0.000 claims abstract description 52
- 238000001914 filtration Methods 0.000 claims abstract description 11
- 238000012706 support-vector machine Methods 0.000 claims abstract description 8
- 238000001746 injection moulding Methods 0.000 claims description 32
- 238000000034 method Methods 0.000 claims description 20
- 238000000354 decomposition reaction Methods 0.000 claims description 14
- 230000008569 process Effects 0.000 claims description 12
- 230000001681 protective effect Effects 0.000 claims description 12
- 229920003023 plastic Polymers 0.000 claims description 10
- 239000004033 plastic Substances 0.000 claims description 10
- 238000006243 chemical reaction Methods 0.000 claims description 6
- 230000002708 enhancing effect Effects 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 238000004458 analytical method Methods 0.000 claims description 4
- 238000006073 displacement reaction Methods 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 abstract 1
- 230000006870 function Effects 0.000 description 6
- 238000004519 manufacturing process Methods 0.000 description 6
- 238000013461 design Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 238000004891 communication Methods 0.000 description 3
- 238000001514 detection method Methods 0.000 description 3
- 230000009471 action Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 238000005266 casting Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000000465 moulding Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001276 controlling effect Effects 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 208000002925 dental caries Diseases 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 239000012943 hotmelt Substances 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 238000002347 injection Methods 0.000 description 1
- 239000007924 injection Substances 0.000 description 1
- 239000007788 liquid Substances 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 239000000243 solution Substances 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B29—WORKING OF PLASTICS; WORKING OF SUBSTANCES IN A PLASTIC STATE IN GENERAL
- B29C—SHAPING OR JOINING OF PLASTICS; SHAPING OF MATERIAL IN A PLASTIC STATE, NOT OTHERWISE PROVIDED FOR; AFTER-TREATMENT OF THE SHAPED PRODUCTS, e.g. REPAIRING
- B29C45/00—Injection moulding, i.e. forcing the required volume of moulding material through a nozzle into a closed mould; Apparatus therefor
- B29C45/17—Component parts, details or accessories; Auxiliary operations
- B29C45/84—Safety devices
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B29—WORKING OF PLASTICS; WORKING OF SUBSTANCES IN A PLASTIC STATE IN GENERAL
- B29C—SHAPING OR JOINING OF PLASTICS; SHAPING OF MATERIAL IN A PLASTIC STATE, NOT OTHERWISE PROVIDED FOR; AFTER-TREATMENT OF THE SHAPED PRODUCTS, e.g. REPAIRING
- B29C45/00—Injection moulding, i.e. forcing the required volume of moulding material through a nozzle into a closed mould; Apparatus therefor
- B29C45/17—Component parts, details or accessories; Auxiliary operations
- B29C45/76—Measuring, controlling or regulating
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B29—WORKING OF PLASTICS; WORKING OF SUBSTANCES IN A PLASTIC STATE IN GENERAL
- B29C—SHAPING OR JOINING OF PLASTICS; SHAPING OF MATERIAL IN A PLASTIC STATE, NOT OTHERWISE PROVIDED FOR; AFTER-TREATMENT OF THE SHAPED PRODUCTS, e.g. REPAIRING
- B29C45/00—Injection moulding, i.e. forcing the required volume of moulding material through a nozzle into a closed mould; Apparatus therefor
- B29C45/17—Component parts, details or accessories; Auxiliary operations
- B29C45/84—Safety devices
- B29C45/844—Preventing damage caused by obstructions or foreign matter caught between mould halves during mould closing, e.g. moulded parts or runners
Landscapes
- Engineering & Computer Science (AREA)
- Manufacturing & Machinery (AREA)
- Mechanical Engineering (AREA)
- Image Processing (AREA)
- Injection Moulding Of Plastics Or The Like (AREA)
Abstract
the invention discloses an embedded mold protection device based on wavelet passivation and an image processing method thereof, wherein OMAP L138 is used as a hardware platform, an image processing algorithm uses graying and mean filtering for image preprocessing, the wavelet passivation highlights residues, and then a support vector machine is used for judging whether residues exist in the wavelet passivated and binarized images, so that the accuracy is greatly improved.
Description
[technical field]
The present invention relates to a kind of embedded mould protective devices and its image processing method based on small echo passivation.
[background technology]
With the development of industry, a Main way of the production automation as industrial development, while needing to take into account production
Safety and high efficiency, but the processes of plastic castings is at present:Controller of plastic injection molding controls plastic mould molding, then toward mould
The plastic liquid of intracavitary injection hot melt shape, after a period of time, controller controls mold die sinking, and then thimble is by workpiece top
Go out, in order not to damage die cavity, situations such as casting will appear overlap after avoiding die cavity from damaging, deformation is generally all equipped with by injection molding machine
One worker, at this time worker will check whether die cavity has residue, judge after no residue operation controller of plastic injection molding into
Row secondary mould closing, the method cannot be guaranteed safety of workers, and not have industrial production automation.
Currently, general mold protecting device be all by one dedicated for operation image Processing Algorithm industrial personal computer, one
It is sleeved on the Image Acquisition of transmission data between industrial personal computer and controller of plastic injection molding to form with control device, industrial personal computer is adopted by image
Collection obtains image data with control device, then runs residue and judges algorithm, obtains result and issues image by communications protocol
Acquisition and control device, Image Acquisition are communicated with control device further according to analysis result and controller of plastic injection molding, transmission molding ginseng
Signal is examined, the method need to install industrial personal computer in production scene, of high cost, be not suitable for high-volume injection molding machine production scene, therefore deposit
In certain limitation.
Therefore, how to overcome above-mentioned defect, it has also become those skilled in the art's important topic urgently to be resolved hurrily.
[invention content]
The invention overcomes the shortcomings of the above-mentioned technology and provides a kind of embedded mould protective devices based on small echo passivation
And its image processing method, the image processing apparatus being made up of embedded system, it completes image procossing, carried out with other equipment
The control functions such as communication and alarm, substantially increase applicability and automation.
To achieve the above object, present invention employs following technical proposals:
A kind of image processing method of the embedded mould protective device based on small echo passivation, which includes shell 1,
Image procossing and protection controller 2, the institute of processing injection molding machine die cavity original image and output control signal are equipped in the shell 1
It states image procossing and controller 2 is protected to be connected with camera 3 for shooting injection molding machine die cavity original image, buzzer siren 4, use
In the communicator 5 for connecting and be transferred to warning message injection molding machine with injection molding machine, described image processing judges with protection controller 2
Image procossing and protection controller 2 treated camera 3 shoot die cavity image memory in residue, then image procossing is controlled with protection
Device 2 processed is alarmed by buzzer siren 4 and warning message is transferred to injection molding machine by communicator 5, the figure of the protective device
As processing step is as follows:
A, image procossing obtains the original image of injection molding machine die cavity with protection controller 2 by camera 3;
B, original image is carried out by gray processing processing with protection controller 2 by image procossing, analysis obtains original image
In each pixel R, G and B component, R, G of each pixel and B component are substituted into following formula operation:Temp=(r*
299+g*587+b*114+500)/1000, the temp values of each pixel are calculated, and the temp values of each pixel are covered into original
Beginning image slices vegetarian refreshments, obtains gray level image;
C, image procossing carries out median filter process with protection controller 2 to gray level image, with 3 pixel *, 3 pictures
It is multiple gray level image modules that vegetarian refreshments, which divides gray level image, to obtain the ash of 9 pixels in gray level image module
Each gray level image mould 9 pixel gray values in the block are carried out following formula operation by angle value respectively:Gx, y=med { f
(x-k, y-l), (k, l ∈ W) }, the med in formula is median operation, i.e., the pixel in W matrixes is first carried out size sequence, then
The intermediate value for finding out pixel in W matrixes is used in combination the Mesophyticum acquired to replace the pixel value of each gray level image module centers point, is obtained
To medium filtering image;
D, image procossing carries out 3 layers of wavelet transform process with protection controller 2 to medium filtering image, passes through wavelet basis letter
NumberWherein a is contraction-expansion factor, and τ is shift factor, and Ψ (x, y) is enabled to indicate that two dimension is substantially small
Wave, then two-dimentional continuous wavelet is defined as:
Wherein,The scale of expression Ψ (x, y) stretches and two-dimension displacement,In order to ensure small echo
The normalization factor that its front and back energy of transformation is constant and introduces, obtains 3 layers of wavelet image;
E, image procossing carries out small echo Passivation Treatment with protection controller 2 to 3 layers of wavelet image, i.e., respectively to 3 layers
The low frequency decomposition coefficient of wavelet image carries out enhancing processing, high-frequency decomposition coefficient carries out attenuation processing, as decomposition coefficient c
(i)>When 405, then low frequency decomposition coefficient enhancing processing is carried out, 4 times is increased to its weight, i.e. c (i)=4*c (i);When c (i)≤
When 405, then high frequency coefficient attenuation processing is carried out, original 0.1 is decayed to its weight, i.e. c (i)=0.1*c (i) is obtained small
Wave is passivated image;
F, image procossing carries out binary conversion treatment with protection controller 2 to small echo passivation image, by wavelet decomposition and again
Image after structure is denoted as f (x, y), then the image after binaryzation is
Wherein threshold is the threshold value that binaryzation is chosen, and obtains binary image;
G, image procossing carries out binary image with protection controller 2 the number statistics of pixel 0 and 1, and this is counted
Data are stored as an array;
H, multiple original images are shot to same injection molding machine die cavity, and each original image are subjected to step a-g image procossings,
The array per pictures is obtained, by each array input linear kernel function:K(xi,xj)=xi·xj, two class sample x will be obtainedi,yi,
Wherein i=1,2 ... n, n are the sample number obtained, work as xiFor the sample class w containing residue1When, yi=1, work as xiTo be free of
The sample class w of residue2When, yi=-1, the i.e. array property containing residue are labeled as 1, and the array property of noresidue marks
It is -1, and Training Support Vector Machines, that is, acquires disaggregated model:In parameter a*, b*;
I, when image procossing and protection controller 2 judge that array property labeled as 1, i.e., when containing residue in die cavity, is schemed
It as processing and protects controller 2 that will be fed back in controller of plastic injection molding containing residue information by communicator 5, and passes through buzzing
Alarm 4 is alarmed, and when controller 2 judges that array property is labeled as 1, i.e., when residue being not present in die cavity, controller 2 is not
Action.
Described image processing is connected with LED alarms 6 with protection controller 2.
Described image processing is connected with the image buffer 7 for image after caching process with protection controller 2.
Described image processing connect VGA displays 8 with protection controller 2.
The beneficial effects of the invention are as follows:
1, the image processing apparatus that the present invention is only made of embedded system, can complete image procossing and other equipment
The control functions such as communicated, alarmed, than generally by carrying out the industrial personal computer of image procossing as host computer and as slave computer
The image processing apparatus of control circuit composition is compacter, and required production scene space is small, can preferably adapt to different work
Environment than traditional mould protecting installs industrial personal computer and has saved many costs;
2, it is passivated using small echo and the image after gray scale is passivated processing, highlight the profile of die cavity, meanwhile, it avoids
The image processing algorithms such as the difference shadow method and matching method that generally use at present need to carry out the rigors of image calibration in advance, improve
The real-time and precision of algorithm, being conducive to the later stage judges that mold cavity whether there is residue;
3, it carries out judging die cavity whether containing residual using the method for 0 and 1 pixel number of image after statistics binaryzation
Object, the judgment mode than generally being compared point by point using each pixel of collected picture and samples pictures is rapider, simultaneously
It is lower to the storage capacity requirement of image processing system, meet the characteristic of embedded system;
4, in training mode use support vector machines, train whether two disaggregated models containing residue, using
Judge whether contain residue in collected die cavity image using this two disaggregated model when pattern.
5, realize it is non-contact fully automatically realize mold defencive function, improve the safety of injection molding machine production and complete autonomous
Property.
[description of the drawings]
Fig. 1 is the structural diagram of the present invention;
Fig. 2 is one of image procossing of the present invention and protection controller circuitry figure;
Fig. 3 is the two of image procossing of the present invention and protection controller circuitry figure;
Fig. 4 is the three of image procossing of the present invention and protection controller circuitry figure;
Fig. 5 is the four of image procossing of the present invention and protection controller circuitry figure;
Fig. 6 is communicator circuit diagram of the present invention;
Fig. 7 is one of image buffer circuit diagram of the present invention;
Fig. 8 is the two of image buffer circuit diagram of the present invention;
Fig. 9 is VGA display circuits of the present invention;
Figure 10 is buzzer siren of the present invention and LED alarm circuit figures;
Figure 11 is mold cavity original image;
Figure 12 is the die cavity image after mold gray processing and medium filtering;
Figure 13 is small echo passivation and the die cavity image after two-value.
[specific implementation mode]
Feature of present invention and other correlated characteristics are described in further detail by embodiment below in conjunction with attached drawing, so as to
In the understanding of technical staff of the same trade:
As shown in Figure 1, a kind of embedded mould protective device based on small echo passivation, includes shell 1, the shell 1
It is interior to be equipped with processing injection molding machine die cavity original image and export the image procossing and protection controller 2 for controlling signal, at described image
Reason is connected with camera 3 for shooting injection molding machine die cavity original image, buzzer siren 4 with protection controller 2, is used for and injection molding
Machine connects and is transferred to warning message the communicator 5 of injection molding machine, and described image processing judges image procossing with protection controller 2
Die cavity image memory is shot in residue with protection controller 2 treated camera 3, then image procossing and protection controller 2 are logical
It crosses the alarm of buzzer siren 4 and warning message is transferred to by injection molding machine by communicator 5.
Wherein, described image processing is connected with LED alarms 6 with protection controller 2, and described image processing is controlled with protection
Device 2 is connected with the image buffer 7 for image after caching process, and described image processing connect VGA with protection controller 2 and shows
Device 8.
As shown in Fig. 2, Fig. 3, Fig. 4 and Fig. 5, image procossing is OMAPL138 main control chips with protection controller 2.
As shown in fig. 6, communicator 5 is communicated using asynchronous serial bus communication modes with injection molding machine, by handling result
Feed back to injection molding machine.
As shown in Figure 7 and Figure 8, image buffer 8 includes two parts, and one is that algorithm is carried out in system operation
The DDR2 used is needed when processing, stores the image data in each calculating process, the other is FLASH, after storing operation
Every pictures pixel 0 and 1 number, ensure system power failure after, the algorithm of this system and the training sample of support vector machines
It does not lose in this library.
As shown in figure 9, VGA displays 9 mainly show the result of image procossing after each algorithm operation.
As shown in Figure 10, alarm module is made of buzzer siren 4 and LED alarms 6, when algorithm judges mould at this time
There is intracavitary residue buzzer circuit just to be driven to carry out alarm and LED flickers.
When work, include the following steps:
Step A:When external equipment inputs a high level by Transistor-Transistor Logic level to the D4 pins in Fig. 2, OMAPL138 will
By from pin W19VP_DIN [0], W18VP_DIN [1], W17VP_DIN [2], V7VP_DIN [3], the W16VP_DIN in Fig. 5
[4]、R14VP_DIN[5]、V16VP_DIN[6]、U18VP_DIN[7]、P17VP_DIN[8]、R15VP_DIN[9]、R19VP_
DIN[10]、R18VP_DIN[11]、T16VP_DIN[12]、U19VP_DIN[13]、V19VP_DIN[14]、V18VP_DIN[15]
Obtain image data;
Step B:By pin U15DDR_D [0] in Fig. 5, U1DDR_D [1], V14DDR_D [2], U13DDR_D [3],
V13DDR_D[4]、V12DDR_D[5]、W12DDR_D[6]、W11DDR_D[7]、T13DDR_D[8]、T11DDR_D[9]、
T10DDR_D[10]、T12DDR_D[11]、U10DDR_D[12]、V10DDR_D[13]、U11DDR_D[14]、W10DDR_D[15]
It caches in one group of instant image data to DDR2;
Step C:Image processing algorithm is called, image data is handled, in the result cache to DDR2 of processing, is counted in algorithm
After the number of complete image 0 and 1 pixel, result is saved in FLASH by pin G17SPI_SIMO and H17SPI_SOMI in Fig. 4
In;
Step D:Judge that die cavity whether after the result containing residue, passes through drawing in Fig. 4 by image processing algorithm
Foot F17UART_RXD and F16UART_TXD send result to external controller of plastic injection molding;
Step E:If it is judged that die cavity, which contains residue, to drive warning circuit by the pin B3GPIO0 [5] in Fig. 2
It alarms and passes through pin C3GPIO0 [1] and drive the warning of LED flashing circuits.
A kind of image processing method of the embedded mould protective device based on small echo passivation, its step are as follows:
A, image procossing obtains the original image of injection molding machine die cavity with protection controller 2 by camera 3, with 7 sets of 800*400
The mold cavity image of pixel is example, and such as the original image that Figure 11 is the 5th mold chamber, the left sides Figure 11 are the mould for having residue
Chamber image placed one piece of plastic product of this mold in the plastics forming die cavity of die cavity middle, be noresidue on the right of Figure 11
Object die cavity image.
B, original image is carried out by gray processing processing with protection controller 2 by image procossing, analysis obtains original image
In each pixel R, G and B component, R, G of each pixel and B component are substituted into following formula operation:Temp=(r*
299+g*587+b*114+500)/1000, the temp values of each pixel are calculated, and the temp values of each pixel are covered into original
Beginning image slices vegetarian refreshments, obtains gray level image;
C, image procossing carries out median filter process with protection controller 2 to gray level image, with 3 pixel *, 3 pictures
It is multiple gray level image modules that vegetarian refreshments, which divides gray level image, to obtain the ash of 9 pixels in gray level image module
Each gray level image mould 9 pixel gray values in the block are carried out following formula operation by angle value respectively:Gx, y=med { f
(x-k, y-l), (k, l ∈ W) }, the med in formula is median operation, i.e., the pixel in W matrixes is first carried out size sequence, then
The intermediate value for finding out pixel in W matrixes is used in combination the Mesophyticum acquired to replace the pixel value of each gray level image module centers point, is obtained
To medium filtering image, Figure 12 is the design sketch after the 5th mold gray processing and medium filtering, and wherein the left sides Figure 12 are to have residual
Design sketch after the mold gray processing and medium filtering of object, the right Figure 12 are the mold gray processing and medium filtering aftereffect of noresidue
Fruit is schemed;
D, image procossing carries out 3 layers of wavelet transform process with protection controller 2 to medium filtering image, passes through wavelet basis functionWherein a is contraction-expansion factor, and τ is shift factor, and Ψ (x, y) is enabled to indicate that two dimension is substantially small
Wave, then two-dimentional continuous wavelet is defined as:
Wherein,The scale of expression Ψ (x, y) stretches and two-dimension displacement,It is small in order to ensure
The normalization factor that its front and back energy of wave conversion is constant and introduces, obtains 3 layers of wavelet image;
E, image procossing carries out small echo Passivation Treatment with protection controller 2 to 3 layers of wavelet image, i.e., respectively to 3 layers
The low frequency decomposition coefficient of wavelet image carries out enhancing processing, high-frequency decomposition coefficient carries out attenuation processing, as decomposition coefficient c
(i)>When 405, then low frequency decomposition coefficient enhancing processing is carried out, 4 times is increased to its weight, i.e. c (i)=4*c (i);When c (i)≤
When 405, then high frequency coefficient attenuation processing is carried out, original 0.1 is decayed to its weight, i.e. c (i)=0.1*c (i) is obtained small
Wave is passivated image;
F, image procossing carries out binary conversion treatment with protection controller 2 to small echo passivation image, by wavelet decomposition and again
Image after structure is denoted as f (x, y), then the image after binaryzation is
Wherein threshold is the threshold value that binaryzation is chosen, and obtains binary image;
G, image procossing carries out binary image with protection controller 2 the number statistics of pixel 0 and 1, and this is counted
Data are stored as an array, the small echo passivation and the design sketch after binaryzation that Figure 13 is the 5th mold, wherein Figure 13
The left side is the small echo passivation for having residue and the design sketch after binaryzation, the profile of residue can be good at it is discernable, by
In the rough physical characteristic of die cavity so that there is the profile of some die cavitys in image, but to the shadow of residue differentiation
It rings less, is the small echo passivation for not having residue and the design sketch after binaryzation on the right of Figure 13, counts and obtain the 0 and 1 of residue
The value of pixel is respectively 4523 and 315477;The value of 0 and 1 pixel of noresidue is respectively 2023 and 317177;
H, multiple original images are shot to same injection molding machine die cavity, and each original image are subjected to step a-g image procossings,
The array per pictures is obtained, by each array input linear kernel function:K(xi,xj)=xi·xj, two class sample x will be obtainedi,yi,
Wherein i=1,2 ... n, n are the sample number obtained, work as xiFor the sample class w containing residue1When, yi=1, work as xiTo be free of
The sample class w of residue2When, yi=-1, the i.e. array property containing residue are labeled as 1, and the array property of noresidue marks
It is -1, and Training Support Vector Machines, that is, acquires disaggregated model:In parameter a*, b*,
After the pixels statistics value for obtaining multiple images of this die cavity, using it to being instructed based on the support vector machines in linear kernel function
Practice, it is 91.66% to obtain accuracy rate;
I, when image procossing and protection controller 2 judge that array property labeled as 1, i.e., when containing residue in die cavity, is schemed
It as processing and protects controller 2 that will be fed back in controller of plastic injection molding containing residue information by communicator 5, and passes through buzzing
Alarm 4 is alarmed, and when controller 2 judges that array property is labeled as 1, i.e., when residue being not present in die cavity, controller 2 is not
Action.
For verification algorithm performance, verified using the mold cavity image of 7 sets of 800*400 pixels.Have 10 in per mold
Opening has the die cavity image of residue, the image of 10 noresidues, wherein 4 figures for having residue and 4 noresidues of extraction
As the template as Training Support Vector Machines, remaining 12 are proved sample as algorithm.Identical experiment condition is used simultaneously, with
Gray level co-occurrence matrixes matching algorithm, difference shadow method obtain the accuracy rate of three kinds of algorithm detection residues such as performance comparison algorithm
Shown in table 1, each algorithm operation takes as shown in table 2.
1 die cavity residue Detection accuracy of table
The detection of 2 die cavity residue of table takes
When this programme algorithm applies to the 1st, 4,5,6 and 7 mold, accuracy rate all maintains 90% or more, the 2nd set and
3 sets of accuracys rate are relatively low, by that this two sets of image analyses, can must analyze its reason, mainly illumination is strong in image acquisition process
It spends different, causes to have residue and noresidue die cavity image to carry out wavelet transformation compared with after binaryzation 0 and 1 number of pixels
Close, next step algorithm optimization only need to carry out luminosity compensation to such case.
According to Tables 1 and 2 experimental data, this programme algorithm has higher than gray level co-occurrence matrixes matching algorithm and difference shadow method
Accuracy rate and it is lower take.
As described above, this case protection is a kind of embedded mould protective device being passivated based on small echo and its image procossing
Method, all technical solutions same or similar with this case should all be shown as falling into the protection domain of this case.
Claims (4)
1. a kind of image processing method of the embedded mould protective device based on small echo passivation, it is characterised in that:The device packet
Included shell (1), be equipped in the shell (1) processing injection molding machine die cavity original image and output control signal image procossing and
Controller (2), described image processing is protected to be connected with the phase for shooting injection molding machine die cavity original image with protection controller (2)
Machine (3), buzzer siren (4), the communicator (5) for connecting and being transferred to warning message injection molding machine with injection molding machine, it is described
Image procossing and protection controller (2) judge image procossing and protection controller (2) treated that camera (3) shoots die cavity image
Inside there is residue, then image procossing is alarmed by buzzer siren (4) with protection controller (2) and passes through communicator (5)
Warning message is transferred to injection molding machine, the image processing step of the protective device is as follows:
A, image procossing obtains the original image of injection molding machine die cavity with protection controller (2) by camera (3);
B, original image is carried out by gray processing processing with protection controller (2) by image procossing, analysis obtains in original image
R, G of each pixel and B component are substituted into following formula operation by R, G and B component of each pixel:Temp=(r*
299+g*587+b*114+500)/1000, the temp values of each pixel are calculated, and the temp values of each pixel are covered into original
Beginning image slices vegetarian refreshments, obtains gray level image;
C, image procossing carries out median filter process with protection controller (2) to gray level image, with 3 pixel *, 3 pixels
It is multiple gray level image modules that point, which divides gray level image, to obtain the gray scale of 9 pixels in gray level image module
Value, following formula operation is carried out by each gray level image mould 9 pixel gray values in the block respectively:G (x, y)=med { f
(x-k, y-l), (k, l ∈ W) }, the med in formula is median operation, i.e., the pixel in W matrixes is first carried out size sequence, then
The intermediate value for finding out pixel in W matrixes is used in combination the Mesophyticum acquired to replace the pixel value of each gray level image module centers point, is obtained
To medium filtering image;
D, image procossing carries out 3 layers of wavelet transform process with protection controller (2) to medium filtering image, passes through wavelet basis functionWherein a is contraction-expansion factor, and τ is shift factor, and Ψ (x, y) is enabled to indicate that two dimension is substantially small
Wave, then two-dimentional continuous wavelet is defined as:
Wherein,The scale of expression Ψ (x, y) stretches and two-dimension displacement,It is small in order to ensure
The normalization factor that its front and back energy of wave conversion is constant and introduces, obtains 3 layers of wavelet image;
E, image procossing carries out small echo Passivation Treatment with protection controller (2) to 3 layers of wavelet image, i.e., small to 3 layers respectively
The low frequency decomposition coefficient of wave conversion image carries out enhancing processing, high-frequency decomposition coefficient carries out attenuation processing, as decomposition coefficient c (i)>
When 405, then low frequency decomposition coefficient enhancing processing is carried out, 4 times is increased to its weight, i.e. c (i)=4*c (i);When c (i)≤405
When, then high frequency coefficient attenuation processing is carried out, original 0.1 is decayed to its weight, i.e. it is blunt to obtain small echo by c (i)=0.1*c (i)
Change image;
F, image procossing carries out binary conversion treatment with protection controller (2) to small echo passivation image, by wavelet function feedback
Image afterwards is denoted as f (x, y), then the image after binaryzation is
Wherein threshold is the threshold value that binaryzation is chosen, and obtains binary image;
G, image procossing carries out pixel 0 and 1 number statistics with protection controller (2) to binary image, and by this statistical number
It is stored according to as an array;
H, multiple original images are shot to same injection molding machine die cavity, and each original image is subjected to step a-g image procossings, obtained
Array per pictures, by each array input linear kernel function:K(xi,xj)=xi·xj, two class sample (x will be obtainedi,yi),
Middle i=1,2 ... n, n are the sample number obtained, work as xiFor the sample class w containing residue1When, yi=1, work as xiIt is residual to be free of
Stay the sample class w of object2When, yi=-1, the i.e. array property containing residue are labeled as 1, and the array property of noresidue marks
It is -1, and Training Support Vector Machines, that is, acquires disaggregated model:In parameter a*, b*;
I, when image procossing judges that array property is labeled as 1 with protection controller (2), i.e., when in die cavity containing residue, image
Processing will be fed back to containing residue information in controller of plastic injection molding with protection controller (2) by communicator (5), and pass through bee
Ring alarm (4) is alarmed, and when controller (2) judges that array property is labeled as 1, i.e. die cavity is interior there is no when residue, controls
Device (2) processed is failure to actuate.
2. a kind of image processing method of embedded mould protective device based on small echo passivation according to claim 1,
It is characterized in that:Described image processing is connected with LED alarms (6) with protection controller (2).
3. a kind of image processing method of embedded mould protective device based on small echo passivation according to claim 1,
It is characterized in that:Described image processing is connected with the image buffer for image after caching process with protection controller (2)
(7)。
4. a kind of image processing method of embedded mould protective device based on small echo passivation according to claim 1,
It is characterized in that:Described image processing connect VGA displays (8) with protection controller (2).
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610459520.8A CN106079338B (en) | 2016-06-21 | 2016-06-21 | Embedded mold protection device based on wavelet passivation and image processing method thereof |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610459520.8A CN106079338B (en) | 2016-06-21 | 2016-06-21 | Embedded mold protection device based on wavelet passivation and image processing method thereof |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106079338A CN106079338A (en) | 2016-11-09 |
CN106079338B true CN106079338B (en) | 2018-07-31 |
Family
ID=57252066
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610459520.8A Active CN106079338B (en) | 2016-06-21 | 2016-06-21 | Embedded mold protection device based on wavelet passivation and image processing method thereof |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106079338B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109857367B (en) * | 2018-12-13 | 2023-06-30 | 上海集成电路研发中心有限公司 | Wavelet decomposition accelerating circuit for embedded image processing |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN201296028Y (en) * | 2007-07-02 | 2009-08-26 | 优志旺电机株式会社 | Die set monitoring device |
CN102363354A (en) * | 2011-11-02 | 2012-02-29 | 宁波安信数控技术有限公司 | Mechanical-vision-based injection molding machine mold protecting system and method |
CN102529020A (en) * | 2012-01-31 | 2012-07-04 | 温州大学 | Visual-motor integrated device for mould detection and protection as well as part detection and picking |
CN202572915U (en) * | 2012-03-16 | 2012-12-05 | 温州大学 | Control device of injection molding machine with mould protection |
CN203418734U (en) * | 2013-06-19 | 2014-02-05 | 东莞市旭通达模具塑胶有限公司 | Machine vision based multifunctional injection mold control device |
CN104626492A (en) * | 2015-01-08 | 2015-05-20 | 中国计量学院 | Injection molding monitoring detection system based on machine vision and operation method |
CN205929382U (en) * | 2016-06-21 | 2017-02-08 | 电子科技大学中山学院 | Embedded mould protection device based on wavelet passivation |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS5734914A (en) * | 1980-08-12 | 1982-02-25 | Fukai Kogyo Kk | Detecting method for residue in resin molding machine |
WO2011137264A1 (en) * | 2010-04-28 | 2011-11-03 | Mettler-Toledo, Inc. | Thermal imaging of molded objects |
-
2016
- 2016-06-21 CN CN201610459520.8A patent/CN106079338B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN201296028Y (en) * | 2007-07-02 | 2009-08-26 | 优志旺电机株式会社 | Die set monitoring device |
CN102363354A (en) * | 2011-11-02 | 2012-02-29 | 宁波安信数控技术有限公司 | Mechanical-vision-based injection molding machine mold protecting system and method |
CN102529020A (en) * | 2012-01-31 | 2012-07-04 | 温州大学 | Visual-motor integrated device for mould detection and protection as well as part detection and picking |
CN202572915U (en) * | 2012-03-16 | 2012-12-05 | 温州大学 | Control device of injection molding machine with mould protection |
CN203418734U (en) * | 2013-06-19 | 2014-02-05 | 东莞市旭通达模具塑胶有限公司 | Machine vision based multifunctional injection mold control device |
CN104626492A (en) * | 2015-01-08 | 2015-05-20 | 中国计量学院 | Injection molding monitoring detection system based on machine vision and operation method |
CN205929382U (en) * | 2016-06-21 | 2017-02-08 | 电子科技大学中山学院 | Embedded mould protection device based on wavelet passivation |
Non-Patent Citations (1)
Title |
---|
基于机器视觉的注塑模具型腔异常检测;杨丁等;《工程塑料应用》;20140430;第74-77页 * |
Also Published As
Publication number | Publication date |
---|---|
CN106079338A (en) | 2016-11-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109508688B (en) | Skeleton-based behavior detection method, terminal equipment and computer storage medium | |
US20200210702A1 (en) | Apparatus and method for image processing to calculate likelihood of image of target object detected from input image | |
CN111046948B (en) | Point cloud simulation and deep learning workpiece pose identification and robot feeding method | |
CN111507317B (en) | Vision-based rotary equipment operation glove wearing detection method and system | |
CN109376589B (en) | ROV deformation small target identification method based on convolution kernel screening SSD network | |
CN110428464B (en) | Multi-class out-of-order workpiece robot grabbing pose estimation method based on deep learning | |
CN112069887B (en) | Face recognition method, device, terminal equipment and storage medium | |
CN104613869B (en) | Method and system for detecting elliptical hole group based on maximum inscribed circle | |
CN111259934A (en) | Stacked object 6D pose estimation method and device based on deep learning | |
Yusoff et al. | Real-time hevea leaves diseases identification using Sobel edge algorithm on FPGA: A preliminary study | |
CN106079338B (en) | Embedded mold protection device based on wavelet passivation and image processing method thereof | |
WO2021040958A1 (en) | System and method for early event detection using generative and discriminative machine learning models | |
CN108205432A (en) | A kind of real-time eliminating method of observation experiment data outliers | |
Zhang et al. | Prioritizing robotic grasping of stacked fruit clusters based on stalk location in RGB-D images | |
CN115641517A (en) | Machine vision defect identification method and system, edge side device and storage medium | |
CN205929382U (en) | Embedded mould protection device based on wavelet passivation | |
Chen et al. | Classification and positioning of circuit board components based on improved YOLOv5 | |
CN112381867B (en) | Automatic filling method for large-area depth image cavity of industrial sorting assembly line | |
Ren et al. | Steel Surface Defect Detection Using Improved Deep Learning Algorithm: ECA-SimSPPF-SIoU-Yolov5 | |
CN113487538A (en) | Multi-target segmentation defect detection method and device and computer storage medium thereof | |
CN117381793A (en) | Material intelligent detection visual system based on deep learning | |
Daogang et al. | Anomaly identification of critical power plant facilities based on YOLOX-CBAM | |
CN111639538B (en) | Casting positioning method based on vision | |
CN111709404B (en) | Machine room legacy identification method, system and equipment | |
Yang et al. | Cherry recognition based on color channel transform |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
TR01 | Transfer of patent right | ||
TR01 | Transfer of patent right |
Effective date of registration: 20190605 Address after: 528400 No. 56 Dongfu North Road, Nantou Town, Zhongshan City, Guangdong Province Patentee after: Zhongshan Langdi Electric Appliances Co., Ltd. Address before: 528400 No. 1 College Road, Shiqi District, Zhongshan City, Guangdong Province Patentee before: Zhongshan College, Electronic Science &. Technology Univ. |