CN109752391A - A kind of carrot Surface Defect Recognition quantization method based on machine vision - Google Patents
A kind of carrot Surface Defect Recognition quantization method based on machine vision Download PDFInfo
- Publication number
- CN109752391A CN109752391A CN201910162568.6A CN201910162568A CN109752391A CN 109752391 A CN109752391 A CN 109752391A CN 201910162568 A CN201910162568 A CN 201910162568A CN 109752391 A CN109752391 A CN 109752391A
- Authority
- CN
- China
- Prior art keywords
- carrot
- image
- quantization method
- matrix
- gray
- 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.)
- Granted
Links
- 244000000626 Daucus carota Species 0.000 title claims abstract description 117
- 235000002767 Daucus carota Nutrition 0.000 title claims abstract description 117
- 238000013139 quantization Methods 0.000 title claims abstract description 49
- 230000007547 defect Effects 0.000 title claims abstract description 42
- 238000000034 method Methods 0.000 title claims abstract description 42
- 238000005336 cracking Methods 0.000 claims abstract description 22
- 238000005452 bending Methods 0.000 claims abstract description 15
- 239000011159 matrix material Substances 0.000 claims description 32
- 210000000746 body region Anatomy 0.000 claims description 6
- 230000003628 erosive effect Effects 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 239000011800 void material Substances 0.000 claims description 3
- 238000013316 zoning Methods 0.000 claims description 2
- 239000000284 extract Substances 0.000 claims 1
- 238000001514 detection method Methods 0.000 abstract description 11
- 238000004519 manufacturing process Methods 0.000 abstract description 9
- 238000012545 processing Methods 0.000 abstract description 4
- 230000008569 process Effects 0.000 abstract description 3
- 238000012271 agricultural production Methods 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 9
- 235000012055 fruits and vegetables Nutrition 0.000 description 4
- 230000008901 benefit Effects 0.000 description 2
- 230000002950 deficient Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 238000000926 separation method Methods 0.000 description 2
- 244000144730 Amygdalus persica Species 0.000 description 1
- 241000207199 Citrus Species 0.000 description 1
- 244000141359 Malus pumila Species 0.000 description 1
- 235000011430 Malus pumila Nutrition 0.000 description 1
- 235000015103 Malus silvestris Nutrition 0.000 description 1
- 235000006040 Prunus persica var persica Nutrition 0.000 description 1
- 241000220324 Pyrus Species 0.000 description 1
- 244000061456 Solanum tuberosum Species 0.000 description 1
- 235000002595 Solanum tuberosum Nutrition 0.000 description 1
- 241000607479 Yersinia pestis Species 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 235000020971 citrus fruits Nutrition 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 235000021017 pears Nutrition 0.000 description 1
- 230000000630 rising effect Effects 0.000 description 1
- 230000007306 turnover Effects 0.000 description 1
Abstract
The present invention relates to a kind of carrot Surface Defect Recognition quantization method based on machine vision, comprising: the pretreatment of carrot image；Carrot is broken defect recognition quantization method, carrot bending defect identification quantization method, carrot cracking defect identification quantization method, carrot fibrous root defect recognition quantization method and carrot small holes caused by worms defect recognition quantization method；The defects of carrot disclosed in this invention fracture, bending, cracking, lateral root and small holes caused by worms, identification quantified detection method using CCD camera obtained carrot realtime graphic, defect recognition quantization is carried out based on realtime graphic of the image processing techniques to acquisition, the subjectivity for overcoming artificial detection improves the objectivity and accuracy of defect recognition quantization.It processes and is classified applied to agricultural production, can be improved production efficiency, reduce production cost.
Description
Technical field
The present invention relates to technical field of agricultural product process, and in particular to a kind of carrot surface defect based on machine vision
Identify quantization method.
Background technique
The selling by grade of carrot helps to promote the market competitiveness and selling price of carrot, improves carrot production
Enterprise profit.Carrot sorting at present, which relies primarily on, to be accomplished manually, and this sorting mode low efficiency, subjectivity is strong, standard is not tight
Lattice, the inherent shortcoming with artificial separation.And with the rising of employment cost, carrot processing enterprise is further had compressed
Profit.Artificial separation is no longer satisfied the production requirement of carrot processing enterprise.
Machine vision provides efficient one kind, low cost and real-time fruits and vegetables on-line checking, stage division, is regarded using machine
Feel is classified carrot online, can save labour, improves identification accuracy, guarantees the consistency of carrot classification,
Production cost is reduced for carrot manufacturing enterprise, improves industrial competition.Have at present largely with machine vision to fruits and vegetables point
The research of grade, but it is primarily adapted for use in spherical or spherical fruits and vegetables, such as apple, peach, pears, citrus and potato.These technologies
Application on elongated fruits and vegetables has limitation, it is therefore desirable to research and develop the machine vision method for being suitble to carrot sorting.
The carrot method for separating based on machine vision of relative maturity not yet domestic at present, according to Chinese people's republicanism
State's domestic trade professional standard (SB/T10450-2007), carrot purchase and sale class requirement (Department of Commerce, the People's Republic of China (PRC)
2007.12.28 issue), the key index for influencing carrot purchase and sale grade has twisted, bending, cracking, blue or green head and disease pest wound etc.,
Its measurement standard is relatively fuzzyyer, brings difficulty for carrot defects detection, needs to quantify the defect index of carrot.
Summary of the invention
Existing carrot classifying equipoment is mostly different according to the thickness of carrot, using machinery sides such as the roller bearings of different gap
Formula is classified carrot according to certain rugosity, although this hierarchical approaches greatly increase the production efficiency, saves labour turnover,
But the carrot of defective (blue or green head, fracture, bending, cracking, fibrous root etc.) cannot be identified, influence grading effect, therefore
Be not suitable for large-scale promotion application.Carrot detection method of surface flaw disclosed by the invention can provide a kind of based on machine vision
Carrot fracture, bending, cracking, fibrous root, small holes caused by worms five defective identify quantization method.
To achieve the above objectives, the technical solution adopted by the present invention is that:
A kind of carrot Surface Defect Recognition quantization method based on machine vision, comprising: the pretreatment of carrot image；
Carrot is broken defect recognition quantization method, carrot bending defect identification quantization method, the identification quantization of carrot cracking defect
Method, carrot fibrous root defect recognition quantization method and carrot small holes caused by worms defect recognition quantization method；
The pretreatment of the carrot image the following steps are included:
(1) RGB image of carrot is converted into HSV image, tri- component images of H, S, V is extracted, to H, S, V tri-
Selected threshold makes background become black to component image respectively, the bianry image for the background that is eliminated；
(2) from carrot source images obtain tri- component images of R, G, B, and according to formula 1., 2. to tri- points of R, G, B
Spirogram picture is handled, and obtains gray level image Gray and gray level image gray, recycle global threshold to gray level image Gray and
Gray level image gray carries out binaryzation respectively, and eliminating complex background influences, and it is corresponding to obtain gray level image Gray, gray level image gray
Bianry image；
Gray=G-B is 1.
Gray=2*R-G-B is 2.；
The carrot is broken defect recognition quantization method, includes the following steps:
(1) bianry image is obtained according to the H component image of carrot；
(2) edge of bianry image is obtained；
(3) according to the edge of bianry image, the minimum circumscribed rectangle of bianry image is obtained, obtains the length of minimum circumscribed rectangle
L and width W；
(4) the aspect ratio ZR=W/L of carrot is calculated；
(5) the marginal point coordinate for extracting bianry image both ends, is stored in matrix Q1With matrix Q2In；
(6) respectively to matrix Q1, matrix Q2Slope is sought, slope k is obtained1And slope k2；
(7) according to ZR, k1And k2, determine whether carrot is really to be broken, prevent the shape for judging and judging fracture curved surface by accident
Shape；
When ZR be greater than threshold value 0.25 when and k1Or k2When middle appearance is more than 5 infinities, carrot is fracture, and ZR is bigger
Fracture is more serious, works as k1Or k2When the infinitely great number of middle appearance is more, indicate that section is more precipitous；
The carrot bending defect identifies quantization method, includes the following steps:
(1) bianry image is obtained according to the H component image of carrot；
(2) area denoising is carried out to obtained bianry image, removes small spot；
(3) bianry image is corroded with circle erosion unit, eliminates burrs on edges；
(4) region of connected pixel area within a certain range, as carrot body region are extracted in bianry image
Domain；
(5) skeleton for extracting carrot body region, obtains skeleton image；
(6) hough transformation is carried out to skeleton image, obtains two fitting a straight lines of skeleton image, two fitting a straight lines
Angle is respectively θ1, θ2；
(7) according to θ1, θ2Curvature BR, BR=180/ π * is calculated | θ1-θ2|；
When curvature BR is greater than threshold value 5, carrot bending is indicated, and curvature BR is bigger, indicate that carrot is bent journey
It spends bigger；
The carrot cracking defect identifies quantization method, includes the following steps:
(1) according to tri- component images of R, G, B of carrot, gray level image Gray is 1. obtained by formula, to gray level image
Gray carries out binaryzation, obtains bianry image；
(2) region of the connected pixel area in a certain range (such as [200050000]) is extracted in bianry image；
(3) equivalent ellipsoids for obtaining the region obtain the major semiaxis a and semi-minor axis b and region area A of equivalent ellipsoids2；
(4) area of equivalent ellipsoids is calculated according to ellipse area formula S=π * a*b；
(5) zoning area and equivalent ellipsoids area ratio i=A2/ S, the ratio between major semiaxis and semi-minor axis a/b；
(6) degree of cracking is indicated with the angle that major semiaxis constitutes triangle with semi-minor axis, KR=2*180/ π * arctan (b/
a)；
When a/b be greater than 5 when and i=A2/ S be greater than threshold value 0.8 when, judge its for fracture area, the bigger explanation of degree of cracking KR
It cracks more serious；
The carrot fibrous root defect recognition quantization method, includes the following steps:
(1) according to the S component image of carrot, bianry image is obtained；
(2) filling of inner void is carried out to bianry image；
(3) the marginal point coordinate for obtaining bianry image, is stored in matrix P；
(4) coordinate of top edge is extracted from matrix P, and there are matrix P1In, lower edge coordinate, which is extracted, from matrix P deposits
In matrix P2In；
(5) respectively to matrix P1With matrix P2It is sampled every 50 pixels, matrix P is obtained after sampling3With matrix P4；
(6) respectively to matrix P3, matrix P4Slope is sought, slope k is obtained3, slope k4, seek k3And k4Peak value, when having one
When a peak value is greater than threshold value 1.5, then there is fibrous root at one in carrot, when there is several peak values greater than threshold value 1.5, then carrot
There are several place's fibrous roots, the corresponding point of peak value is found, and is recorded in bianry image, seeks coordinate, which answers carrot
It must location of root；
The carrot small holes caused by worms defect recognition quantization method, includes the following steps:
(1) according to tri- component images of R, G, B of carrot, gray level image gray is 2. obtained by formula, to gray level image
Gray carries out binaryzation, obtains bianry image, and the area of bianry image is A；
(2) area is extracted in bianry image in the connected region of a certain range (such as [1,000 20000])；
(3) equivalent ellipsoids for obtaining connected region, obtain the major semiaxis a, semi-minor axis b and perimeter C of equivalent ellipsoids；
(4) area of equivalent ellipsoids is calculated according to ellipse area formula S=π * a*b；
(5) the circularity e=4* π * S/C of connected region is obtained2；
(6) small holes caused by worms ratio CR=S/A is calculated；
When circularity e is greater than threshold value 0.6, judge that it, for small holes caused by worms, and calculates small holes caused by worms ratio.
Detailed description of the invention
The present invention has following attached drawing:
Defect recognition quantization flow figure Fig. 1 of the invention.
Fibrous root identification quantization flow chart in Fig. 2 present invention.
Identified amount schematic diagram is broken in Fig. 3 present invention.
Curved-ray tracing quantifies schematic diagram in Fig. 4 present invention.
Cracking identification quantization schematic diagram in Fig. 5 present invention.
Fibrous root identifies schematic diagram in Fig. 6 present invention.
Specific embodiment
Below in conjunction with attached drawing, invention is further described in detail.
Embodiment, please refers to Fig. 1-Fig. 6, and Fig. 1 gives carrot defects identification quantization main flow chart of the present invention；
Fig. 2 gives the flow chart of carrot fibrous root identification quantization detection in the present invention；Fig. 3 shows fracture identified amount in the present invention
The schematic diagram of detection；Fig. 4 shows schematically the schematic diagram of curved-ray tracing quantization detection in the present invention；Fig. 5 illustrates this hair
The schematic diagram of bright middle cracking identification quantization detection；Fig. 6 shows the schematic diagram of carrot fibrous root identification quantization detection in the present invention.
Carrot image is shot using CCD camera first, image resolution ratio is 2592*1944 and carrot is that level is put
It sets.Then successively acquired image is handled as follows according to process shown in Fig. 1:
Image is pre-processed, H, S, V image and R, G, B image of carrot image are obtained respectively, then to H component
Image uses threshold value to carry out binaryzation for 0.2, obtains the bianry image for eliminating background；Binaryzation, threshold value are carried out to S component image
It is 0.5, background is set as black；It uses threshold value to carry out binaryzation for 0.4 V component image, obtains bianry image；1. using formula
Obtain gray level image and binaryzation, threshold value 100；Gray level image is 2. obtained using formula, carries out two-value with global threshold (100)
Change, obtains the bianry image of prominent RED sector.Next according to method noted earlier carry out carrot fracture, bending,
Cracking, fibrous root and the quantization of small holes caused by worms identification.
Carrot fault recognition quantization step is as follows:
1) Threshold segmentation is carried out to the H component image of carrot, threshold value 0.2 obtains bianry image；
2) noise is removed with the mode that area denoises, obtains the edge of bianry image；
3) according to the edge of bianry image, the minimum circumscribed rectangle of bianry image is obtained, obtains the long L of minimum circumscribed rectangle
With wide W；
4) the aspect ratio ZR=W/L of carrot is calculated；
5) edge coordinate for extracting bianry image both ends, is stored in Q1And Q2In (shown in such as Fig. 3 (Q)；
6) Q is calculated1And Q2Slope k1、k2(such as Fig. 3 (K1) shown in)；
When aspect ratio ZR is greater than 0.25 and k1Or k2In when having no less than 5 infinities, that is, be judged as fracture, the present embodiment
In aspect ratio ZR be 0.27 and k1In have 15 it is infinite a little bigger, so the present embodiment can be determined that as fracture, and the plane of disruption is very
It is precipitous.
It is as follows that carrot curved-ray tracing quantifies implementation steps:
1) Threshold segmentation is carried out to the H component image of carrot, threshold value 0.2 obtains bianry image；
2) area denoising is carried out to obtained bianry image, removes small spot；
3) the round erosion unit for being 3 with radius carries out corrosion treatment to bianry image, eliminates burrs on edges；
4) region that area is greater than 100000 elemental areas, as carrot body region are extracted in bianry image
Domain；
5) skeleton for extracting carrot body region, obtains skeleton image；
6) hough transformation is carried out to skeleton image, obtains two fitting a straight lines as shown in Figure 4, two fitting a straight lines
Angle is respectively θ1, θ2；
6) according to θ1, θ2Curvature BR=180/ π * is calculated | θ1-θ2|；
In the present embodiment, curvature BR is 11.24, and curvature BR is bigger, indicates that carrot bending degree is more serious.
Carrot cracking identification quantized result is as shown in Figure 5, the specific steps are as follows:
1) after obtaining carrot RGB image, gray level image, and binaryzation is 1. calculated with formula, obtains bianry image,
Threshold value is 100；
2) area is extracted in bianry image in the connected region of [2,000 50000]；
3) equivalent ellipsoids of connected region and the area A of connected region are obtained2；
4) the major semiaxis a and semi-minor axis b of equivalent ellipsoids are obtained, ellipse area is S=π * a*b；
5) i=A is calculated2The ratio between/S, major semiaxis and semi-minor axis a/b and degree of cracking KR=2*180/ π * arctan (b/a)；
A/b=6.53 and i=0.97 in this example, so judging it for cracking, degree of cracking KR=17.41.
Steps are as follows for carrot fibrous root identification quantization:
1) flow chart, selection S component image carry out binaryzation, obtain bianry image, threshold value 0.5 according to Fig.2,；
2) filling of inner void is carried out to bianry image；
3) the marginal point coordinate for extracting bianry image, is saved in P, as shown in Fig. 6 (P)；
4) by coordinate (Fig. 6 (P in top edge is extracted from P clockwise1)) and lower edge coordinate, it is respectively present P1And P2(figure
6(P2))；
5) to P1And P2It is sampled every 50 pixels, P is obtained after sampling3And P4,
6) respectively to matrix P3, matrix P4Slope is sought, slope k is obtained3, slope k4, seek k3And k4Peak value, such as Fig. 6 (K)
It is shown；
Having a peak value in this example is 2.07, is greater than 1.5, so carrot has a fibrous root, palpus location of root is on the diagram
Coordinate be (1417,377)；
Steps are as follows for carrot small holes caused by worms identification quantization:
1) 2. the RGB image for obtaining carrot is obtained by formula and emphasizes red gray level image, and binaryzation, is obtained
Bianry image, threshold value 100；The area of bianry image is A=480936；
2) in such a way that area denoises, area is obtained in bianry image in the connected region of [1,000 20000], i.e.,
It is small holes caused by worms region；
3) equivalent ellipsoids in small holes caused by worms region are obtained and obtain the major semiaxis a=34.16 of equivalent ellipsoids, semi-minor axis b=
28.57, perimeter C=201.87；
4) area of equivalent ellipsoids, as small holes caused by worms area are calculated according to ellipse area formula S=π * a*b；
5) circularity calculation formula e=4* π * S/C2；
6) small holes caused by worms ratio CR=S/A is calculated；
Ellipse area S is 3065.8 elemental areas in this example, and small holes caused by worms ratio CR=0.0064, small holes caused by worms circularity is 0.945.
Advantages of the present invention:
The defects of carrot disclosed in this invention fracture, bending, cracking, lateral root and small holes caused by worms identification quantified detection method benefit
Carrot realtime graphic is obtained with CCD camera, defect recognition quantization is carried out based on realtime graphic of the image processing techniques to acquisition,
The subjectivity for overcoming artificial detection improves the objectivity and accuracy of defect recognition quantization.It is processed applied to agricultural production
And classification, it can be improved production efficiency, reduce production cost.
The content being not described in detail in this specification belongs to the prior art well known to professional and technical personnel in the field.
Claims (2)
1. a kind of carrot Surface Defect Recognition quantization method based on machine vision characterized by comprising carrot image
Pretreatment；Carrot is broken defect recognition quantization method, carrot bending defect identifies quantization method, carrot cracking defect
Identify quantization method, carrot fibrous root defect recognition quantization method and carrot small holes caused by worms defect recognition quantization method；
The pretreatment of the carrot image the following steps are included:
(1) RGB image of carrot is converted into HSV image, extracts tri- component images of H, S, V, to tri- components of H, S, V
Selected threshold makes background become black to image respectively, the bianry image for the background that is eliminated；
(2) from carrot source images obtain tri- component images of R, G, B, and according to formula 1., 2. to tri- component maps of R, G, B
As being handled, gray level image Gray and gray level image gray are obtained, recycles global threshold to gray level image Gray and gray scale
Image gray carries out binaryzation respectively, and eliminating complex background influences, and obtains gray level image Gray, gray level image gray corresponding two
It is worth image；
Gray=G-B is 1.
Gray=2*R-G-B is 2.；
The carrot is broken defect recognition quantization method, includes the following steps:
(1) bianry image is obtained according to the H component image of carrot；
(2) edge of bianry image is obtained；
(3) according to the edge of bianry image, obtain the minimum circumscribed rectangle of bianry image, obtain minimum circumscribed rectangle long L and
Wide W；
(4) the aspect ratio ZR=W/L of carrot is calculated；
(5) the marginal point coordinate for extracting bianry image both ends, is stored in matrix Q1With matrix Q2In；
(6) respectively to matrix Q1, matrix Q2Slope is sought, slope k is obtained1And slope k2；
(7) according to ZR, k1And k2, determine whether carrot is really to be broken, and judge the shape for being broken curved surface；
When ZR be greater than threshold value 0.25 when and k1Or k2When middle appearance is more than 5 infinities, carrot is fracture, and ZR gets over major rupture
It is more serious, work as k1Or k2When the infinitely great number of middle appearance is more, indicate that section is more precipitous；
The carrot bending defect identifies quantization method, includes the following steps:
(1) bianry image is obtained according to the H component image of carrot；
(2) area denoising is carried out to obtained bianry image, removes small spot；
(3) bianry image is corroded with circle erosion unit, eliminates burrs on edges；
(4) region of connected pixel area within a certain range, as carrot body region are extracted in bianry image；
(5) skeleton for extracting carrot body region, obtains skeleton image；
(6) hough transformation is carried out to skeleton image, obtains two fitting a straight lines of skeleton image, the angle of two fitting a straight lines
Respectively θ1, θ2；
(7) according to θ1, θ2Curvature BR, BR=180/ π * is calculated | θ1-θ2|；
When curvature BR is greater than threshold value 5, carrot bending is indicated, and curvature BR is bigger, indicate that carrot bending degree is got over
Greatly；
The carrot cracking defect identifies quantization method, includes the following steps:
(1) according to tri- component images of R, G, B of carrot, gray level image Gray is 1. obtained by formula, to gray level image Gray
Binaryzation is carried out, bianry image is obtained；
(2) region of connected pixel area within a certain range is extracted in bianry image；
(3) equivalent ellipsoids for obtaining the region obtain the major semiaxis a and semi-minor axis b and region area A of equivalent ellipsoids2；
(4) area of equivalent ellipsoids is calculated according to ellipse area formula S=π * a*b；
(5) zoning area and equivalent ellipsoids area ratio i=A2/ S, the ratio between major semiaxis and semi-minor axis a/b；
(6) degree of cracking is indicated with the angle that major semiaxis constitutes triangle with semi-minor axis, KR=2*180/ π * arctan (b/a)；
When a/b be greater than 5 when and i=A2When/S is greater than threshold value 0.8, it is judged for fracture area, and degree of cracking KR is bigger to illustrate cracking more
Seriously.
2. the carrot Surface Defect Recognition quantization method based on machine vision as described in claim 1, which is characterized in that institute
Carrot fibrous root defect recognition quantization method is stated, is included the following steps:
(1) according to the S component image of carrot, bianry image is obtained；
(2) filling of inner void is carried out to bianry image；
(3) the marginal point coordinate for obtaining bianry image, is stored in matrix P；
(4) coordinate that top edge is extracted from matrix P, is stored in matrix P1In, lower edge coordinate is extracted from matrix P, is saved
In matrix P2In；
(5) respectively to matrix P1With matrix P2It is sampled every 50 pixels, matrix P is obtained after sampling3With matrix P4；
(6) respectively to matrix P3, matrix P4Slope is sought, slope k is obtained3, slope k4, seek k3And k4Peak value, when there is a peak
Value be greater than threshold value 1.5 when, then carrot exist one at fibrous root, when have several peak values be greater than threshold value 1.5 when, then carrot presence
Several place's fibrous roots find the corresponding point of peak value, and are recorded in bianry image, seek coordinate, which answers carrot fibrous root
Position；
The carrot small holes caused by worms defect recognition quantization method, includes the following steps:
(1) according to tri- component images of R, G, B of carrot, gray level image gray is 2. obtained by formula, to gray level image gray
Binaryzation is carried out, bianry image is obtained, the area of bianry image is A；
(2) area is extracted in bianry image in the connected region of a certain range；
(3) equivalent ellipsoids for obtaining connected region, obtain the major semiaxis a, semi-minor axis b and perimeter C of equivalent ellipsoids；
(4) area of equivalent ellipsoids is calculated according to ellipse area formula S=π * a*b；
(5) the circularity e=4* π * S/C of connected region is obtained2；
(6) small holes caused by worms ratio CR=S/A is calculated；
When circularity e is greater than threshold value 0.6, judge that it, for small holes caused by worms, and calculates small holes caused by worms ratio.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811586927 | 2018-12-25 | ||
| CN2018115869272 | 2018-12-25 |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109752391A true CN109752391A (en) | 2019-05-14 |
| CN109752391B CN109752391B (en) | 2020-06-30 |
Family
ID=66406784
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910162568.6A Expired - Fee Related CN109752391B (en) | 2018-12-25 | 2019-03-05 | Carrot surface defect identification and quantification method based on machine vision |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109752391B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110434090A (en) * | 2019-08-23 | 2019-11-12 | 中山德著智能科技有限公司 | A kind of carrot surface flaw detecting method |
| CN114354628A (en) * | 2022-01-05 | 2022-04-15 | 威海若维信息科技有限公司 | Root agricultural product defect detection method based on machine vision |
Citations (22)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH08171634A (en) * | 1994-12-16 | 1996-07-02 | Iseki & Co Ltd | Fruits recognizing device |
| JPH09163853A (en) * | 1995-12-19 | 1997-06-24 | Iseki & Co Ltd | Image processing device for fruit |
| JP2007071789A (en) * | 2005-09-08 | 2007-03-22 | Yanmar Co Ltd | Chestnut quality inspection method |
| TWI277911B (en) * | 2005-03-01 | 2007-04-01 | Ta-Te Lin | An image processing method for selected fruit by the X-ray scanner to automatically identify the region of injury within the scanned fruit |
| CN101281112A (en) * | 2008-04-30 | 2008-10-08 | 浙江理工大学 | Image type automatic analysis method for mesh adhesion rice corn |
| EP2466380A1 (en) * | 2010-12-20 | 2012-06-20 | Agfa Graphics N.V. | A curable jettable fluid for making a flexographic printing master |
| CN102829735A (en) * | 2012-08-31 | 2012-12-19 | 哈尔滨工业大学 | Machine vision based detection method of defect of geometrical shape of back surface of E type magnet |
| CN103394472A (en) * | 2013-07-04 | 2013-11-20 | 中国农业大学 | Method for detecting and grading greening potatoes based on machine vision |
| CN104056789A (en) * | 2013-03-19 | 2014-09-24 | 青岛农业大学 | Carrot defect image quantitative detection method and carrot sorting apparatus |
| CN104215639A (en) * | 2013-06-05 | 2014-12-17 | 江南大学 | Pear surface defect detection method based on machine vision |
| CN104280349A (en) * | 2014-10-28 | 2015-01-14 | 南京农业大学 | Method for identifying hollowness of white radishes based on hyperspectral image |
| CN105092597A (en) * | 2015-09-14 | 2015-11-25 | 广东电网有限责任公司电力科学研究院 | Detection method for cracks on surface of hard-plastic material |
| CN105354847A (en) * | 2015-11-10 | 2016-02-24 | 浙江大学 | Fruit surface defect detection method based on adaptive segmentation of sliding comparison window |
| CN105404869A (en) * | 2015-11-20 | 2016-03-16 | 陕西科技大学 | Computer vision based fruit shape grading method |
| CN105891231A (en) * | 2015-01-26 | 2016-08-24 | 青岛农业大学 | Carrot surface defect detection method based on image processing |
| CN106204589A (en) * | 2016-07-11 | 2016-12-07 | 陕西科技大学 | A kind of fruit type quality stage division based on Digital Image Processing |
| CN106251361A (en) * | 2016-08-30 | 2016-12-21 | 兰州交通大学 | A kind of rail surface defects image self-adapting division method |
| CN106355185A (en) * | 2016-08-30 | 2017-01-25 | 兰州交通大学 | Method for rapidly extracting steel rail surface area under condition of vibration |
| CN107256549A (en) * | 2017-06-06 | 2017-10-17 | 滁州市天达汽车部件有限公司 | A kind of bamboo strip defect detection method based on machine vision |
| CN108171693A (en) * | 2017-12-27 | 2018-06-15 | 合肥市雅视智能科技有限公司 | A kind of method of automatic detection poor quality mushroom |
| CN108985363A (en) * | 2018-07-03 | 2018-12-11 | 长安大学 | A kind of cracks in reinforced concrete bridge classifying identification method based on RBPNN |
| CN109002821A (en) * | 2018-07-19 | 2018-12-14 | 武汉科技大学 | A kind of Internetbank shield digit recognition method based on connected domain and tangent slope |
-
2019
- 2019-03-05 CN CN201910162568.6A patent/CN109752391B/en not_active Expired - Fee Related
Patent Citations (22)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH08171634A (en) * | 1994-12-16 | 1996-07-02 | Iseki & Co Ltd | Fruits recognizing device |
| JPH09163853A (en) * | 1995-12-19 | 1997-06-24 | Iseki & Co Ltd | Image processing device for fruit |
| TWI277911B (en) * | 2005-03-01 | 2007-04-01 | Ta-Te Lin | An image processing method for selected fruit by the X-ray scanner to automatically identify the region of injury within the scanned fruit |
| JP2007071789A (en) * | 2005-09-08 | 2007-03-22 | Yanmar Co Ltd | Chestnut quality inspection method |
| CN101281112A (en) * | 2008-04-30 | 2008-10-08 | 浙江理工大学 | Image type automatic analysis method for mesh adhesion rice corn |
| EP2466380A1 (en) * | 2010-12-20 | 2012-06-20 | Agfa Graphics N.V. | A curable jettable fluid for making a flexographic printing master |
| CN102829735A (en) * | 2012-08-31 | 2012-12-19 | 哈尔滨工业大学 | Machine vision based detection method of defect of geometrical shape of back surface of E type magnet |
| CN104056789A (en) * | 2013-03-19 | 2014-09-24 | 青岛农业大学 | Carrot defect image quantitative detection method and carrot sorting apparatus |
| CN104215639A (en) * | 2013-06-05 | 2014-12-17 | 江南大学 | Pear surface defect detection method based on machine vision |
| CN103394472A (en) * | 2013-07-04 | 2013-11-20 | 中国农业大学 | Method for detecting and grading greening potatoes based on machine vision |
| CN104280349A (en) * | 2014-10-28 | 2015-01-14 | 南京农业大学 | Method for identifying hollowness of white radishes based on hyperspectral image |
| CN105891231A (en) * | 2015-01-26 | 2016-08-24 | 青岛农业大学 | Carrot surface defect detection method based on image processing |
| CN105092597A (en) * | 2015-09-14 | 2015-11-25 | 广东电网有限责任公司电力科学研究院 | Detection method for cracks on surface of hard-plastic material |
| CN105354847A (en) * | 2015-11-10 | 2016-02-24 | 浙江大学 | Fruit surface defect detection method based on adaptive segmentation of sliding comparison window |
| CN105404869A (en) * | 2015-11-20 | 2016-03-16 | 陕西科技大学 | Computer vision based fruit shape grading method |
| CN106204589A (en) * | 2016-07-11 | 2016-12-07 | 陕西科技大学 | A kind of fruit type quality stage division based on Digital Image Processing |
| CN106251361A (en) * | 2016-08-30 | 2016-12-21 | 兰州交通大学 | A kind of rail surface defects image self-adapting division method |
| CN106355185A (en) * | 2016-08-30 | 2017-01-25 | 兰州交通大学 | Method for rapidly extracting steel rail surface area under condition of vibration |
| CN107256549A (en) * | 2017-06-06 | 2017-10-17 | 滁州市天达汽车部件有限公司 | A kind of bamboo strip defect detection method based on machine vision |
| CN108171693A (en) * | 2017-12-27 | 2018-06-15 | 合肥市雅视智能科技有限公司 | A kind of method of automatic detection poor quality mushroom |
| CN108985363A (en) * | 2018-07-03 | 2018-12-11 | 长安大学 | A kind of cracks in reinforced concrete bridge classifying identification method based on RBPNN |
| CN109002821A (en) * | 2018-07-19 | 2018-12-14 | 武汉科技大学 | A kind of Internetbank shield digit recognition method based on connected domain and tangent slope |
Non-Patent Citations (10)
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110434090A (en) * | 2019-08-23 | 2019-11-12 | 中山德著智能科技有限公司 | A kind of carrot surface flaw detecting method |
| CN114354628A (en) * | 2022-01-05 | 2022-04-15 | 威海若维信息科技有限公司 | Root agricultural product defect detection method based on machine vision |
| CN114354628B (en) * | 2022-01-05 | 2023-10-27 | 威海若维信息科技有限公司 | Rhizome agricultural product defect detection method based on machine vision |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109752391B (en) | 2020-06-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Yiyang | The design of glass crack detection system based on image preprocessing technology | |
| CN111310558B (en) | Intelligent pavement disease extraction method based on deep learning and image processing method | |
| CN109682839B (en) | Online detection method for surface defects of metal arc-shaped workpiece | |
| CN116721106A (en) | Profile flaw visual detection method based on image processing | |
| CN107369136B (en) | Visual detection method for surface cracks of polycrystalline diamond compact | |
| CN107607554A (en) | A kind of Defect Detection and sorting technique of the zinc-plated stamping parts based on full convolutional neural networks | |
| CN109187581A (en) | The bearing finished products plate defects detection method of view-based access control model | |
| CN107610104A (en) | Crack detecting method at a kind of elevator compensation chain R based on machine vision | |
| CN109472788B (en) | Method for detecting flaw on surface of airplane rivet | |
| CN105445277A (en) | Visual and intelligent detection method for surface quality of FPC (Flexible Printed Circuit) | |
| CN107490582B (en) | Assembly line workpiece detection system | |
| CN115330645B (en) | Welding image enhancement method | |
| CN108596930A (en) | It must algorithm and the morphologic ray image weld seam extracting method of mathematics based on longicorn | |
| CN106934801A (en) | A kind of fluorescentmagnetic particle(powder) defect inspection method based on Laws texture filterings | |
| CN105069778B (en) | Based on the industrial products detection method of surface flaw that target signature notable figure builds | |
| CN108038838A (en) | A kind of cotton fibriia species automatic testing method and system | |
| CN111080636A (en) | CNN semantic segmentation self-learning detection method for surface defects of color steel tiles | |
| CN115294144B (en) | Method for identifying surface defects of furniture composite board | |
| CN109752391A (en) | A kind of carrot Surface Defect Recognition quantization method based on machine vision | |
| CN112258444A (en) | Elevator steel wire rope detection method | |
| CN113221881A (en) | Multi-level smart phone screen defect detection method | |
| CN115471466A (en) | Steel surface defect detection method and system based on artificial intelligence | |
| CN113436102A (en) | Method for detecting surface defects of machined part | |
| CN115619799B (en) | Grain image segmentation method and system based on transfer learning | |
| CN111860500A (en) | Shoe print wear area detection and edge tracing method |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20200630 |