CN107194939A - A kind of edge detection method and system based on improved edge focusing - Google Patents

A kind of edge detection method and system based on improved edge focusing Download PDF

Info

Publication number
CN107194939A
CN107194939A CN201710318071.XA CN201710318071A CN107194939A CN 107194939 A CN107194939 A CN 107194939A CN 201710318071 A CN201710318071 A CN 201710318071A CN 107194939 A CN107194939 A CN 107194939A
Authority
CN
China
Prior art keywords
edge
image
gaussian kernel
edge image
connectivity
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
Application number
CN201710318071.XA
Other languages
Chinese (zh)
Inventor
张俊勇
伍世虔
邹谜
陈鹏
徐正勤
余杭
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Wuhan University of Science and Engineering WUSE
Wuhan University of Science and Technology WHUST
Original Assignee
Wuhan University of Science and Engineering WUSE
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Wuhan University of Science and Engineering WUSE filed Critical Wuhan University of Science and Engineering WUSE
Priority to CN201710318071.XA priority Critical patent/CN107194939A/en
Publication of CN107194939A publication Critical patent/CN107194939A/en
Pending legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/10Segmentation; Edge detection
    • G06T7/13Edge detection
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/10Image acquisition modality
    • G06T2207/10004Still image; Photographic image

Abstract

The invention discloses a kind of edge detection method based on improved edge focusing, this method includes:Rim detection is carried out to original image using the first Gaussian kernel, first edge image is obtained;On the basis of first Gaussian kernel, rim detection is carried out to described first image using the second Gaussian kernel for reducing fixed step size S, second edge image is obtained, wherein, S is the difference of second Gaussian kernel and first Gaussian kernel.Weak edge and noise in the second edge figure is removed using eight connectivity method, the 3rd edge image is obtained.The present invention provides a kind of edge detection method and system based on improved edge focusing, edge-detected image is present in the prior art definition and the not high technical problem of the degree of accuracy are solved, realizes and improves the definition of edge image and the technique effect of accuracy while positioning precision is improved.

Description

A kind of edge detection method and system based on improved edge focusing
Technical field
The present invention relates to technical field of visual measurement, more particularly to a kind of side edge detection based on improved edge focusing Method and system.
Background technology
In modern intelligent production, generally the product on streamline is connected as detection instrument using CCD camera Continuous IMAQ, analysis judgement is carried out by extracting picture edge characteristic, reaches the purpose to workpiece positioning and dimensional measurement.
Conventional a few class edge detection methods have at present:Sobel operators, Prewitt operators, Log operators and Canny sides Edge detective operators.Sobel operators can be described as with the edge that Prewitt operators are produced it is consistent, to edge position it is more accurate, but It is that edge is thicker, is not suitable for follow-up high-precision dimensional measurement.Log operators are due to using the method for finding zero crossing, easily False edge is detected, and edge is positioned by second order derivation, operand is also than larger.The rim detection phase of Canny operators For preceding several, effect preferably, is able to detect that Single pixel edge, but Canny operators are thin in order to suppress unnecessary edge , it is necessary to using larger gaussian filtering yardstick, can so make edge displacement when section and noise, edge precision is influenceed.In order to Overcome above mentioned problem, a kind of edge detection method of utilization edge focusing thought is occurred in that, so as to suppress apart from new edge farther out Partial edge details and noise.
However, prior art is utilized in edge focusing thought edge detection method, to second edge picture adjacent area Noise can not be excluded, because existing method simply instead of first edge image with second edge image, and not considered The neighbouring noise and weak edge occurred in old edge.Therefore, edge detection graph image sharpness and the degree of accuracy are still suffered from existing method Not high technical problem, therefore a kind of edge detection method of improved edge focusing is provided be particularly important.
The content of the invention
The embodiment of the present invention provides a kind of edge detection method and system based on improved edge focusing, existing to solve The not high technical problem of definition and the degree of accuracy with the presence of edge-detected image in technology.
The invention discloses a kind of edge detection method based on improved edge focusing, methods described includes:
Rim detection is carried out to original image using the first Gaussian kernel, first edge image is obtained;
On the basis of first Gaussian kernel, using diminution fixed step size S the second Gaussian kernel to described first image Rim detection is carried out, second edge image is obtained, wherein, S is the difference of second Gaussian kernel and first Gaussian kernel.
Weak edge and noise in the second edge figure is removed using eight connectivity method, the 3rd edge image is obtained.
In a kind of edge detection method based on improved edge focusing that the present invention is provided, the second Gauss is used described When checking the first edge image progress rim detection, the adjacent domain to the first edge image carries out rim detection.
Alternatively, the use eight connectivity method removes the weak edge and noise in the second edge figure, including:
Judge the pixel in the second edge figure whether in the eight connectivity region of the first edge figure;
If, judge whether the length of the connected region where the pixel is more than preset value, wherein, the connection Region is any one connected region in the eight connectivity region;
If being more than, retain the pixel;
If being not more than, the pixel is deleted, to remove the weak edge and noise in the second edge figure, institute is obtained State the 3rd edge image.
Alternatively, the weak edge and noise in the second edge figure is removed using eight connectivity method, obtains the 3rd side After edge image, including:
On the basis of second Gaussian kernel, using diminution fixed step size S the 3rd Gaussian kernel to the 3rd edge Image carries out rim detection, obtains the 4th edge image;
Weak edge and noise in 4th edge image is removed using eight connectivity method, the 5th edge image is obtained;
Judge whether the 3rd Gaussian kernel of the 5th edge image is less than default Gaussian kernel value;
If it is less, using the 5th edge image as the image of final rim detection;
If it is not, then continuing using the 4th Gaussian kernel using diminution fixed step size S to the 5th edge image Rim detection is carried out, the 6th edge image is obtained, and the weak edge in the 6th edge image is removed using eight connectivity method And noise.
Based on same inventive concept, second aspect of the present invention provides a kind of edge inspection based on improved edge focusing Examining system, the system includes:
First obtains module, for carrying out rim detection to original image using the first Gaussian kernel, obtains first edge figure Picture;
Second obtains module, on the basis of first Gaussian kernel, using the second Gauss for reducing fixed step size S Check described first image and carry out rim detection, obtain second edge image, wherein, S is second Gaussian kernel and described the The difference of one Gaussian kernel.
First processing module, for carrying out rim detection to described first image using the second Gaussian kernel, obtains second After edge image, the weak edge and noise in the second edge figure are removed using eight connectivity method, the 3rd edge graph is obtained Picture.
Alternatively, the second acquisition module is additionally operable to:The second Gaussian kernel is used to the first edge image described When carrying out rim detection, the adjacent domain to the first edge image carries out rim detection.
Alternatively, the first processing module is additionally operable to:
Judge the pixel in the second edge figure whether in the eight connectivity region of the first edge figure;
If, judge whether the length of the connected region where the pixel is more than preset value, wherein, the connection Region is any one connected region in the eight connectivity region;
If being more than, retain the pixel;
If being not more than, the pixel is deleted, to remove the weak edge and noise in the second edge figure, institute is obtained State the 3rd edge image.
Alternatively, the system also includes Second processing module, for removing second side using eight connectivity method After weak edge and noise in edge figure, the 3rd edge image of acquisition,
On the basis of second Gaussian kernel, using diminution fixed step size S the 3rd Gaussian kernel to the 3rd edge Image carries out rim detection, obtains the 4th edge image;
Weak edge and noise in 4th edge image is removed using eight connectivity method, the 5th edge image is obtained;
Judge whether the 3rd Gaussian kernel of the 5th edge image is less than default Gaussian kernel value;
If it is less, using the 5th edge image as the image of final rim detection;
If it is not, then continuing using the 4th Gaussian kernel using diminution fixed step size S to the 5th edge image Rim detection is carried out, the 6th edge image is obtained, and the weak edge in the 6th edge image is removed using eight connectivity method And noise.
The one or more technical schemes provided in the embodiment of the present invention, have at least the following technical effects or advantages:
A kind of edge detection method based on improved edge focusing that the embodiment of the present application is provided, methods described includes: Rim detection is carried out to original image using the first Gaussian kernel, first edge image is obtained;On the basis of first Gaussian kernel On, rim detection is carried out to described first image using the second Gaussian kernel for reducing fixed step size S, second edge image is obtained, Weak edge and noise in the second edge figure is removed using eight connectivity method, the 3rd edge image is obtained.;In the above method In, due to obtaining the first side to carrying out rim detection using the first Gaussian kernel using second Gaussian kernel smaller than the first Gaussian kernel Edge image, carries out rim detection again, and the precision of obtained second edge image is higher than the precision of first edge image, and adopts The weak edge and noise of second edge are removed with eight connectivity method, the noise of second edge adjacent domain and weak side can be removed Edge, solves edge-detected image is present in the prior art definition and the not high technical problem of the degree of accuracy.Inhibit edge The weak edge and noise of image, realize and improve the definition of edge image and the skill of accuracy while positioning precision is improved Art effect.
Described above is only the general introduction of technical solution of the present invention, in order to better understand the technological means of the present invention, And can be practiced according to the content of specification, and in order to allow above and other objects of the present invention, feature and advantage can Become apparent, below especially exemplified by the embodiment of the present invention.
Brief description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing There is the accompanying drawing used required in technology description to be briefly described, it should be apparent that, drawings in the following description are this hairs Some bright embodiments, for those of ordinary skill in the art, on the premise of not paying creative work, can be with root Other accompanying drawings are obtained according to these accompanying drawings.
Fig. 1 is a kind of flow chart of the edge detection method based on improved edge focusing in the embodiment of the present invention;
Fig. 2 is a kind of structure chart of the edge inspection system based on improved edge focusing in the embodiment of the present invention.
Embodiment
The embodiments of the invention provide a kind of edge detection method based on improved edge focusing and system, to solve Edge-detected image is present in the prior art definition and the not high technical problem of the degree of accuracy.
Technical scheme in the embodiment of the present application, general thought is as follows:
A kind of edge detection method based on improved edge focusing, is carried out using the first Gaussian kernel to original image first Rim detection, obtains first edge image;Then on the basis of first Gaussian kernel, using reducing the of fixed step size S Two Gaussian kernels to described first image carry out rim detection, obtain second edge image, wherein, S be second Gaussian kernel with The difference of first Gaussian kernel;The weak edge and noise in the second edge figure are removed using eight connectivity method again, the is obtained Three edge images.
In the above-mentioned methods, due to having used second Gaussian kernel smaller than the first Gaussian kernel to being carried out using the first Gaussian kernel Rim detection obtains first edge image, and rim detection is carried out again, and the precision of obtained second edge image is higher than the first side The precision of edge image, and using the weak edge and noise of eight connectivity method removal second edge, second edge can be removed adjacent The noise of near field and weak edge, solve edge-detected image is present in the prior art definition and the not high skill of the degree of accuracy Art problem.The weak edge and noise of edge image are inhibited, realizes and improves edge image while positioning precision is improved The technique effect of definition and accuracy.
To make the purpose, technical scheme and advantage of the embodiment of the present invention clearer, below in conjunction with the embodiment of the present invention In accompanying drawing, the technical scheme in the embodiment of the present invention is clearly and completely described, it is clear that described embodiment is A part of embodiment of the present invention, rather than whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art The every other embodiment obtained under the premise of creative work is not made, belongs to the scope of protection of the invention.
Embodiment one
A kind of edge detection method based on improved edge focusing is present embodiments provided, Fig. 1, methods described is referred to Including:
Step S101:Rim detection is carried out to original image using the first Gaussian kernel, first edge image is obtained;
Step S102:On the basis of first Gaussian kernel, using diminution fixed step size S the second Gaussian kernel to described First image carries out rim detection, obtains second edge image, wherein, S is second Gaussian kernel and first Gaussian kernel Difference.
Step S103:Weak edge and noise in the second edge figure is removed using eight connectivity method, the 3rd side is obtained Edge image.
It should be noted that the edge detection method of improved edge focusing and existing edge focusing method in the application Difference be, improve the precision of rim detection by reducing the method for Gaussian kernel first, it is even more important that upper On the basis of stating, the weak edge and noise in second edge figure are removed using eight connectivity method, higher positioning accuracy can obtained While possess preferable weak edge and noise suppression effect, improve the definition and accuracy of edge-detected image.
Below, a kind of edge detection method based on improved edge focusing provided with reference to Fig. 1 the application carries out detailed It is thin to introduce:
Step S101 is first carried out:Rim detection is carried out to original image using the first Gaussian kernel, first edge figure is obtained Picture;
In specific implementation process, it is possible to use CCD camera connects as detection instrument to the product for needing to detect Continuous IMAQ, obtains the original image of product, carries out rim detection to original image using the first Gaussian kernel first, obtains the One edge image, wherein, the first Gaussian kernel is the larger Gaussian kernel σ of value0, occurrence can be set according to actual conditions Put, rim detection can be carried out using Canny operators, obtain first edge image, now, first edge image is made an uproar to eliminate The rough grade edge image of sound and details, with E (i, j, σ0) represent, wherein i and j represent the image coordinate of marginal point.
Then step S102 is performed:On the basis of first Gaussian kernel, using the second Gauss for reducing fixed step size S Check described first image and carry out rim detection, obtain second edge image, wherein, S is second Gaussian kernel and described the The difference of one Gaussian kernel.
In specific implementation process, fixed step size S value can be set according to actual conditions, for example can for 0.3, 0.4th, 0.5,0.6 etc., by taking S=0.5 as an example, with second Gaussian kernel σ (σ=σ0- s) obtained first edge image is entered again Row rim detection, can obtain new edge graph E (i, j, σ0- s), i.e. second edge image, because the S of selection is smaller, second Not over one pixel of offset of strong edge relative to first edge image in edge image.By front and rear edge twice After detection, we retain it is new contain noisy accurate second edge image, give up first edge image.As a result of Second Gaussian kernel smaller than the first Gaussian kernel, second edge image can be more accurate than the positioning precision of first edge image, But some noises and unnecessary edge details occur relative to first edge image.
Next step S103 is performed:Weak edge and noise in the second edge figure is removed using eight connectivity method, Obtain the 3rd edge image.
Due to performing after above-mentioned steps S102, some noises and unnecessary edge details occur in second edge image, Therefore, the weak edge and noise in the second edge figure are removed using eight connectivity method, so as to obtain the 3rd edge image.Solution Definition and the not high technical problem of the degree of accuracy that the edge-detected image in the prior art of having determined is present.Inhibit edge image Weak edge and noise, realize the technology effect of the definition that edge image is improved while positioning precision is improved and accuracy Really.
Specifically, the use eight connectivity method removes the weak edge and noise in the second edge figure, including:
Judge the pixel in the second edge figure whether in the eight connectivity region of the first edge figure;
If, judge whether the length of the connected region where the pixel is more than preset value, wherein, the connection Region is any one connected region in the eight connectivity region;
If being more than, retain the pixel;
If being not more than, the pixel is deleted, to remove the weak edge and noise in the second edge figure, institute is obtained State the 3rd edge image.
In specific implementation process, to second edge image E (i, j, σ0- s) method of eight connectivity is used to its connected region Domain is marked, if second edge image E (i, j, σ0- s) in pixel in eight adjacent ranges of first edge image, and Connected region length where the point is more than preset value T, and wherein T is rule of thumb configured, then the point retains, and otherwise removes. The weak edge and noise occurred in second edge image can be effectively removed by this method.The rest may be inferred, travels through the second side Edge image E (i, j, σ0- s) in all pixels point, the degree of precision edge graph of obtained not Noise, i.e. the 3rd edge image. It is designated as E0(i,j,σ0-s)。
In the edge detection method that the present embodiment is provided, the second Gaussian kernel is used to the first edge image described When carrying out rim detection, the adjacent domain to the first edge image carries out rim detection.
In the edge detection method that the present embodiment is provided, in the second edge figure is removed using eight connectivity method After weak edge and noise, the 3rd edge image of acquisition, including:
On the basis of second Gaussian kernel, using diminution fixed step size S the 3rd Gaussian kernel to the 3rd edge Image carries out rim detection, obtains the 4th edge image;
Weak edge and noise in 4th edge image is removed using eight connectivity method, the 5th edge image is obtained;
Judge whether the 3rd Gaussian kernel of the 5th edge image is less than default Gaussian kernel value;
If it is less, using the 5th edge image as the image of final rim detection;
If it is not, then continuing using the 4th Gaussian kernel using diminution fixed step size S to the 5th edge image Rim detection is carried out, the 6th edge image is obtained, and the weak edge in the 6th edge image is removed using eight connectivity method And noise.
In the edge detection method that the present embodiment is provided, then with the 3rd Gaussian kernel σ (σ=σ0- 2s) in second edge Image E0(i,j,σ0- s) and its adjacent domain progress rim detection, obtain the 4th edge image E (i, j, σ0- 2s), then use Eight connectivity method removes the 4th weak edge of edge image and noise, so as to obtain high-precision edge graph E0(i,j,σ0- 2s), i.e., Five edge images.Then repeat the above steps, whether the 3rd Gaussian kernel is less than default Gaussian kernel value, specifically, Ke Yitong Cross and judge the 3rd Gaussian kernel σ=σ0Whether-ns is sufficiently small, and default Gaussian kernel value can be chosen according to actual conditions, for example Default Gaussian kernel value is 1, if the 3rd Gaussian kernel is less than 1, using the 5th edge image as the figure of final rim detection Picture;If the 3rd Gaussian kernel is not less than 1, continue using the 4th Gaussian kernel using diminution fixed step size S to the described 5th Edge image carries out rim detection, obtains the 6th edge image, and remove in the 6th edge image using eight connectivity method Weak edge and noise.Gaussian kernel σ is steadily decreasing by the above method0Iteration optimization edge reaches approaching to reality edge, together The method of Shi Caiyong eight connectivities removes the noise and weak edge at new edge, finally gives the clear and standard at not Noise and weak edge True edge image.
Embodiment two
Based on the inventive concept same with embodiment one, the embodiment of the present invention two provides a kind of poly- based on improved edge Burnt edge inspection system, the system includes:
First obtains module 201, for carrying out rim detection to original image using the first Gaussian kernel, obtains first edge Image;
Second obtains module 202, on the basis of first Gaussian kernel, using the second of diminution fixed step size S Gaussian kernel carries out rim detection to described first image, obtains second edge image, wherein, S is second Gaussian kernel and institute State the difference of the first Gaussian kernel.
First processing module 203, for carrying out rim detection to described first image using the second Gaussian kernel, obtains the After two edge images, the weak edge and noise in the second edge figure are removed using eight connectivity method, the 3rd edge is obtained Image.
In the system that the present embodiment is provided, the second acquisition module is additionally operable to:The second Gaussian kernel pair is used described When the first edge image carries out rim detection, the adjacent domain to the first edge image carries out rim detection.
In the system that the present embodiment is provided, the first processing module is additionally operable to:
Judge the pixel in the second edge figure whether in the eight connectivity region of the first edge figure;
If, judge whether the length of the connected region where the pixel is more than preset value, wherein, the connection Region is any one connected region in the eight connectivity region;
If being more than, retain the pixel;
If being not more than, the pixel is deleted, to remove the weak edge and noise in the second edge figure, institute is obtained State the 3rd edge image.
The system provided in the present embodiment, in addition to Second processing module, for described using the removal of eight connectivity method After weak edge and noise in second edge figure, the 3rd edge image of acquisition,
On the basis of second Gaussian kernel, using diminution fixed step size S the 3rd Gaussian kernel to the 3rd edge Image carries out rim detection, obtains the 4th edge image;
Weak edge and noise in 4th edge image is removed using eight connectivity method, the 5th edge image is obtained;
Judge whether the 3rd Gaussian kernel of the 5th edge image is less than default Gaussian kernel value;
If it is less, using the 5th edge image as the image of final rim detection;
If it is not, then continuing using the 4th Gaussian kernel using diminution fixed step size S to the 5th edge image Rim detection is carried out, the 6th edge image is obtained, and the weak edge in the 6th edge image is removed using eight connectivity method And noise.
In embodiment one based on various change mode and the instantiation system that is equally applicable to the present embodiment, pass through Foregoing pair of detailed description, those skilled in the art are clear that in the present embodiment, so for the letter of specification It is clean, it will not be described in detail herein.
The one or more technical schemes provided in the embodiment of the present invention, have at least the following technical effects or advantages:
A kind of edge detection method based on improved edge focusing that the embodiment of the present application is provided, methods described includes: Rim detection is carried out to original image using the first Gaussian kernel, first edge image is obtained;On the basis of first Gaussian kernel On, rim detection is carried out to described first image using the second Gaussian kernel for reducing fixed step size S, second edge image is obtained, Weak edge and noise in the second edge figure is removed using eight connectivity method, the 3rd edge image is obtained.In the above method In, due to obtaining the first side to carrying out rim detection using the first Gaussian kernel using second Gaussian kernel smaller than the first Gaussian kernel Edge image, carries out rim detection again, and the precision of obtained second edge image is higher than the precision of first edge image, and adopts The weak edge and noise of second edge are removed with eight connectivity method, the noise of second edge adjacent domain and weak side can be removed Edge, solves edge-detected image is present in the prior art definition and the not high technical problem of the degree of accuracy.Inhibit edge The weak edge and noise of image, realize and improve the definition of edge image and the skill of accuracy while positioning precision is improved Art effect.
, but those skilled in the art once know basic creation although preferred embodiments of the present invention have been described Property concept, then can make other change and modification to these embodiments.So, appended claims are intended to be construed to include excellent Select embodiment and fall into having altered and changing for the scope of the invention.
Obviously, those skilled in the art can carry out various changes and modification without departing from this hair to the embodiment of the present invention The spirit and scope of bright embodiment.So, if these modifications and variations of the embodiment of the present invention belong to the claims in the present invention And its within the scope of equivalent technologies, then the present invention is also intended to comprising including these changes and modification.

Claims (8)

1. a kind of edge detection method based on improved edge focusing, it is characterised in that methods described includes:
Rim detection is carried out to original image using the first Gaussian kernel, first edge image is obtained;
On the basis of first Gaussian kernel, described first image is carried out using the second Gaussian kernel for reducing fixed step size S Rim detection, obtains second edge image, wherein, S is the difference of second Gaussian kernel and first Gaussian kernel.
Weak edge and noise in the second edge figure is removed using eight connectivity method, the 3rd edge image is obtained.
2. the method as described in claim 1, it is characterised in that use the second Gaussian kernel to the first edge image described When carrying out rim detection, the adjacent domain to the first edge image carries out rim detection.
3. the method as described in claim 1, it is characterised in that the use eight connectivity method is removed in the second edge figure Weak edge and noise, including:
Judge the pixel in the second edge figure whether in the eight connectivity region of the first edge figure;
If, judge whether the length of the connected region where the pixel is more than preset value, wherein, the connected region For any one connected region in the eight connectivity region;
If being more than, retain the pixel;
If being not more than, the pixel is deleted, to remove the weak edge and noise in the second edge figure, described the is obtained Three edge images.
4. the method as described in claim 1, it is characterised in that in the second edge figure is removed using eight connectivity method After weak edge and noise, the 3rd edge image of acquisition, including:
On the basis of second Gaussian kernel, using diminution fixed step size S the 3rd Gaussian kernel to the 3rd edge image Rim detection is carried out, the 4th edge image is obtained;
Weak edge and noise in 4th edge image is removed using eight connectivity method, the 5th edge image is obtained;
Judge whether the 3rd Gaussian kernel of the 5th edge image is less than default Gaussian kernel value;
If it is less, using the 5th edge image as the image of final rim detection;
If it is not, then continuing to carry out the 5th edge image using using the 4th Gaussian kernel for reducing fixed step size S Rim detection, obtains the 6th edge image, and remove the weak edge in the 6th edge image using eight connectivity method and make an uproar Sound.
5. a kind of edge inspection system based on improved edge focusing, it is characterised in that the system includes:
First obtains module, for carrying out rim detection to original image using the first Gaussian kernel, obtains first edge image;
Second obtains module, on the basis of first Gaussian kernel, using the second Gaussian kernel pair for reducing fixed step size S Described first image carries out rim detection, obtains second edge image, wherein, S is that second Gaussian kernel and described first are high The difference of this core.
First processing module, for carrying out rim detection to described first image using the second Gaussian kernel, obtains second edge After image, the weak edge and noise in the second edge figure are removed using eight connectivity method, the 3rd edge image is obtained.
6. system as claimed in claim 5, it is characterised in that the second acquisition module is additionally operable to:Second is used described When Gaussian kernel carries out rim detection to the first edge image, the adjacent domain to the first edge image carries out edge inspection Survey.
7. system as claimed in claim 5, it is characterised in that the first processing module is additionally operable to:
Judge the pixel in the second edge figure whether in the eight connectivity region of the first edge figure;
If, judge whether the length of the connected region where the pixel is more than preset value, wherein, the connected region For any one connected region in the eight connectivity region;
If being more than, retain the pixel;
If being not more than, the pixel is deleted, to remove the weak edge and noise in the second edge figure, described the is obtained Three edge images.
8. system as claimed in claim 5, it is characterised in that also including Second processing module, for using eight connectivity side Method is removed after the weak edge and noise in the second edge figure, the 3rd edge image of acquisition,
On the basis of second Gaussian kernel, using diminution fixed step size S the 3rd Gaussian kernel to the 3rd edge image Rim detection is carried out, the 4th edge image is obtained;
Weak edge and noise in 4th edge image is removed using eight connectivity method, the 5th edge image is obtained;
Judge whether the 3rd Gaussian kernel of the 5th edge image is less than default Gaussian kernel value;
If it is less, using the 5th edge image as the image of final rim detection;
If it is not, then continuing to carry out the 5th edge image using using the 4th Gaussian kernel for reducing fixed step size S Rim detection, obtains the 6th edge image, and remove the weak edge in the 6th edge image using eight connectivity method and make an uproar Sound.
CN201710318071.XA 2017-05-08 2017-05-08 A kind of edge detection method and system based on improved edge focusing Pending CN107194939A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710318071.XA CN107194939A (en) 2017-05-08 2017-05-08 A kind of edge detection method and system based on improved edge focusing

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710318071.XA CN107194939A (en) 2017-05-08 2017-05-08 A kind of edge detection method and system based on improved edge focusing

Publications (1)

Publication Number Publication Date
CN107194939A true CN107194939A (en) 2017-09-22

Family

ID=59872927

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710318071.XA Pending CN107194939A (en) 2017-05-08 2017-05-08 A kind of edge detection method and system based on improved edge focusing

Country Status (1)

Country Link
CN (1) CN107194939A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108759661A (en) * 2018-03-27 2018-11-06 松下电子部品(江门)有限公司 The straight line offset method of edge detection vision system

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105528584A (en) * 2015-12-23 2016-04-27 浙江宇视科技有限公司 Method and device for detecting frontal face image

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105528584A (en) * 2015-12-23 2016-04-27 浙江宇视科技有限公司 Method and device for detecting frontal face image

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
SHIQIAN WU, ET AL.: "An Edge-Based Method for Aligning Differently Exposed Images", 《IEEE CONFERENCE ON INDUSTRIAL ELECTRONICS AND APPLICATIONS (ICIEA)》 *
李芳等: "基于快速8-连通域标记的视频字幕提取新算法", 《电视技术》 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108759661A (en) * 2018-03-27 2018-11-06 松下电子部品(江门)有限公司 The straight line offset method of edge detection vision system

Similar Documents

Publication Publication Date Title
Dhankhar et al. A review and research of edge detection techniques for image segmentation
CN109141232B (en) Online detection method for disc castings based on machine vision
CN105787902B (en) Utilize the image denoising method of block sorting detection noise
CN107678192B (en) Mura defect detection method based on machine vision
CN107368792A (en) A kind of finger vein identification method and system based on wave filter and Bone Edge
CN107341793A (en) A kind of target surface image processing method and device
CN106373098B (en) Method for suppressing random impulsive noise based on non-similar pixel statistics
CN106934768A (en) A kind of method and device of image denoising
CN104732530A (en) Image edge detection method
CN107194939A (en) A kind of edge detection method and system based on improved edge focusing
CN102663706A (en) Adaptive weighted mean value filtering method based on diamond template
CN107767372B (en) Chip pin online visual detection system and method for layered parallel computing
CN108827979A (en) A kind of module group lens appearance detecting method
JP5889778B2 (en) Automatic unevenness detection apparatus and automatic unevenness detection method for flat panel display
CN105184792B (en) A kind of saw blade wear extent On-line Measuring Method
CN104102911A (en) Image processing for AOI (automated optical inspection)-based bullet appearance defect detection system
CN111489389A (en) Light spot center detection method
CN110378902A (en) A kind of scratch detection method under strong noise background
JP6114559B2 (en) Automatic unevenness detector for flat panel display
Zheng et al. Measurement of laser welding pool geometry using a closed convex active contour model
Čisar et al. Kernel sets in compass edge detection
US11068740B2 (en) Particle boundary identification
CN103914861B (en) Image processing method and device
CN109544571A (en) A kind of metallic phase image edge detection method based on mathematical morphology
CN109448012A (en) A kind of method for detecting image edge and device

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20170922