CN105261031B - The line detection method and device calculated based on gradient - Google Patents
The line detection method and device calculated based on gradient Download PDFInfo
- Publication number
- CN105261031B CN105261031B CN201510848072.6A CN201510848072A CN105261031B CN 105261031 B CN105261031 B CN 105261031B CN 201510848072 A CN201510848072 A CN 201510848072A CN 105261031 B CN105261031 B CN 105261031B
- Authority
- CN
- China
- Prior art keywords
- gradient
- pixel
- target image
- connected region
- region
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20072—Graph-based image processing
Abstract
The present invention provides a kind of line detection method and device calculated based on gradient, comprising: calculates the gradient of pixel in a target image, the gradient includes gradient direction and gradient magnitude；It filters out and corresponds to the pixel that the gradient magnitude is less than predetermined gradient amplitude thresholds in the target image；The gradient connected region of the target image is determined according to the gradient direction of pixel in the target image；Filter out the nonconforming region in region gradient direction in the gradient connected region of the target image；The endpoint of the straightway and straightway along the connected region gradient direction is obtained from the connected region；Straight-line detection is exported as a result, the straight-line detection result includes the endpoint of the straightway and the straightway.The present invention also provides a kind of straight-line detection device calculated based on gradient, solve in the prior art using computationally intensive, the straight line information of acquisition few technical problem when Hough transformation progress straight-line detection.
Description
Technical field
The present invention relates to technical field of image processing, in particular to a kind of straight-line detection side calculated based on gradient
Method and device.
Background technique
Straight line is one of essential characteristic of image, and profile of the general object in two-dimensional image can be approximately straight
Therefore the combination of line or camber line can be converted to basic elements such as these straight lines, camber lines the detection and identification of contour of object
Detection and extraction.In addition, can also realize that rigid body is rotated using line correspondences method in movement image analysis and estimation field
The measurement of amount and displacement, so the line detection algorithm of image being studied and being inquired into very important practical meaning
Justice.
For existing line detection method substantially based on Hough transformation, basic thought is to utilize the antithesis between point-line
Property carries out straight-line detection.That is: the conllinear corresponding straight line intersected inside parameter space of point in image space.Conversely, parameter
All straight lines that the same point is intersected in space have conllinear point to be corresponding to it inside image space.
Referring to Fig.1, the straight line under cartesian coordinate system can be expressed as the form of y=kx+b, and k here indicates straight line
Slope, b represent offset.So, from Fig. 1 it is not difficult to find that point P1, P2 in X-Y coordinate correspond to the straight line in k-b coordinate
L1, L2；And the point P0 in k-b coordinate then corresponds to the straight line L0 in X-Y coordinate.
It is above-mentioned to there is a situation where that slope k is infinity with the straight line that slope describes in x-y coordinate, it can be brought not to calculating
Just.Under normal conditions, straight line is indicated using polar form based on the straight-line detection of Hough transformation.That is, straight line is in polar coordinates
It is indicated in system are as follows: the form of sin θ=0 a ρ cos θ+b ρ.So, intersection point conllinear in x-y plane has just corresponded to former straight
Two parameters of line.Any linearity region can correspond some point in parameter space on image, and image is taken up an official post
The pixel of meaning all exists simultaneously on many linearity regions.Typically, each linearity region will have an accumulator
Counter corresponds to each other therewith.For each pixel on image, all linearity regions comprising it are found, and to correspond to
Summary counter add 1, to count the sum for the pixel that each accumulator is included.When some linearity region institute on image
When the pixel number for including is more than threshold value T, being considered as the corresponding region of the accumulator, there are straight lines.
Based on the line detection method of Hough transformation to noise image with certain robustness and stability, but there is
Calculation amount is huge, and memory expends the problems such as serious, meanwhile, Hough transformation can only point out the presence of certain straight line in image, not
The information more about this straight line can be provided.More multi straight can also be provided while can be reduced calculation amount by how providing one kind
The line detection method of information is technical problem urgently to be solved to those skilled in the art.
Summary of the invention
In view of this, the present invention provides a kind of line detection method and device calculated based on gradient, it is terraced using that should be based on
The line detection method and device calculated is spent, the calculation amount of straight-line detection can not only be reduced, moreover it is possible to provide more straight line information.
To achieve the goals above, technical solution used in the embodiment of the present invention is as follows:
In a first aspect, the embodiment of the present invention provides a kind of line detection method calculated based on gradient, comprising: calculate a mesh
The gradient of pixel in logo image, the gradient include gradient direction and gradient magnitude；It filters out and corresponds to institute in the target image
State the pixel that gradient magnitude is less than predetermined gradient amplitude thresholds；It is determined according to the gradient direction of pixel in the target image
The gradient connected region of the target image；It is non-uniform to filter out region gradient direction in the gradient connected region of the target image
The region of property；The endpoint of the straightway and straightway along the connected region gradient direction is obtained from the connected region；It is defeated
Straight-line detection is as a result, the straight-line detection result includes the endpoint of the straightway and the straightway out.
Second aspect, the embodiment of the present invention also provide a kind of straight-line detection device calculated based on gradient, comprising: calculate mould
Block, for calculating the gradient of pixel in a target image, the gradient includes gradient direction and gradient magnitude；Except module of making an uproar,
For filtering out the pixel for corresponding to the gradient magnitude in the target image and being less than predetermined gradient amplitude thresholds；Gradient connected region
Domain determining module, for determining the gradient connected region of the target image according to the gradient direction of pixel in the target image
Domain；Module is filtered out, the nonconforming region in region gradient direction in the gradient connected region for filtering out the target image；It obtains
Modulus block, for obtaining the endpoint of straightway and straightway along the connected region gradient direction from the connected region；
Output module, for exporting straight-line detection as a result, the straight-line detection result includes the end of the straightway and the straightway
Point.
By the above-mentioned line detection method and device calculated based on gradient, by filtering out the noise in image, to image
It carries out connected region to detect to obtain connected region, then filters out the poor region of gradient direction consistency in connected region to obtain ladder
The connected region for spending orientation consistency obtains the end of straightway and straightway to the connected region processing of the gradient direction consistency
Point.Line detection method relative to Hough transformation not only reduces calculation amount, can also provide more multi straight information.
To enable the above objects, features and advantages of the present invention to be clearer and more comprehensible, preferred embodiment is cited below particularly, and cooperate
Appended attached drawing, is described in detail below.
Detailed description of the invention
Illustrate the embodiment of the present invention or technical solution in the prior art in order to clearer, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention without creative efforts, may be used also for ordinary skill user person
To obtain other relevant attached drawings according to these attached drawings.
Fig. 1 shows corresponding relationship of the straight line in cartesian coordinate system and parameter coordinate system.
Fig. 2 shows the processes for the line detection method calculated based on gradient that the first specific embodiment of the invention provides
Figure.
Fig. 3, which is shown, determines gradient side in gradient connected region step in the line detection method that Fig. 2 is calculated based on gradient
To ideal actual value with the relational graph between the true observation of gradient direction.
Fig. 4-A shows the schematic diagram of the input picture of a pending straight-line detection.
Fig. 4-B shows the schematic diagram of the connected region of gradient direction after input picture is handled.
Fig. 4-C shows the schematic diagram for the gradient image that gradient direction is more dispersed in above-mentioned connected region.
Fig. 5 is the structural block diagram for the straight-line detection device calculated based on gradient that the second specific embodiment of the invention provides.
Specific embodiment
Below in conjunction with attached drawing in the embodiment of the present invention, technical solution in the embodiment of the present invention carries out clear, complete
Ground description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.Usually exist
The component of the embodiment of the present invention described and illustrated in attached drawing can be arranged and be designed with a variety of different configurations herein.Cause
This, is not intended to limit claimed invention to the detailed description of the embodiment of the present invention provided in the accompanying drawings below
Range, but it is merely representative of selected embodiment of the invention.Based on the embodiment of the present invention, those skilled in the art are not doing
Every other embodiment obtained under the premise of creative work out, shall fall within the protection scope of the present invention.
First specific embodiment
Fig. 2 provides the flow chart of the line detection method calculated based on gradient for the first specific embodiment of the invention.Such as Fig. 2
Shown, the flow chart of the line detection method calculated based on gradient in the present embodiment may comprise steps of.
Step S110 calculates the gradient of pixel in a target image, and the gradient includes gradient direction and gradient magnitude.
Calculate the gradient of each pixel in target image.The expression formula of the gradient of pixel is as follows:
Here gx、gyRespectively represent pixel level, the gradient of vertical direction.The gradient of pixel includes the pixel
Gradient direction and gradient magnitude.
The expression formula of gradient direction is as follows:
The expression formula of gradient magnitude is as follows:
Step S120 is filtered out and is corresponded to the pixel that the gradient magnitude is less than predetermined gradient amplitude thresholds in the target image
Point.
Pixel in image with smaller gradient magnitude means that the pixel is in flat site or change of gradient
Slow region.Simultaneously as, more or less there are several noises in calculating gradient procedure in the quantification problem of image pixel value
Point, therefore, it is necessary to which these pixels with small gradient magnitude are filtered out.
Using following expression, filtering out for the pixel of small gradient magnitude is carried out.
Wherein, T represents the threshold value of gradient magnitude, and M (x, y) represents the mask Prototype drawing generated after gradient magnitude filtering
Picture, subsequent operation handle the coordinate points for being 1 in mask template image value.
Step S130 determines that the gradient of the target image is connected to according to the gradient direction of pixel in the target image
Region.
Assuming that there are the ideal gradient direction i of image slices vegetarian refreshments and its quantizing noise n, then to the ideal of image slices vegetarian refreshments
Expression formula should be as follows:
Wherein,It is the actual gradient direction value of image slices vegetarian refreshments.
Using signal processing knowledge we learn, it is above only ideal situation, and it is actual observation data expression formula
Should be as follows:
Wherein,It is the observation to noise signal,It is to gradient direction
Ideal actual value,It is the true observation of gradient direction.Obviously, the ideal actual value of gradient direction is with the true of gradient direction
There are certain errors between real observation.As shown in figure 3, q here isRadius and our ladders allowed
Spend range existing for the noise of direction, it may be assumed that q represents error radius (range) to a certain extent.
Consider that the pixel by gradient direction difference greater than predetermined angle threshold value τ is excluded, and by gradient disparities in τ model
Pixel in enclosing is summarized as gradient connected domain.That is expression formula are as follows: | angle error |≤τ, we obtainHere threshold value ρ is set using greatest hope, and τ represents the angle tolerance of connected region detection judgement
Degree.Error radius q is set based on experience value, and for the value of usual q between [2,4], user can also be according to the actual situation to q
Value make adjustment appropriate, the value of τ is usually between [5,25].
Step S140 filters out the nonconforming region in region gradient direction in the gradient connected region of the target image.
In the ideal case, divided by the detection that previous step carries out connected region to gradient direction, we can will scheme
As being divided into the connected region that several gradient directions reach unanimity, referring to Fig. 4-A and Fig. 4-B, wherein Fig. 4-A indicates input figure
Picture, Fig. 4-B indicate the connected region of gradient direction.
But in view of setting an angle tolerance threshold value when to the calculating of gradient direction connected area segmentation is calculated, therefore
Here there is a possibility that certain, although so that in the connected region that above-mentioned steps are calculated all neighborhood pixels gradient direction
Very close to, but the case where the gradient direction in whole region still disperses very much, referring to shown in Fig. 4-C.
In response to this, it is necessary to carry out handling the gradient direction connection of this falseness using region consistency judgement
Region.In this embodiment, it is preferred to use fuzzy mathematics to carry out consistency judgement to the gradient direction of connected region.Mould
Paste mathematics has broken in ordinary set opinion element to the membership of the absolutization of set, it is believed that " belonging to " and " being not belonging to " this
Except two states, the case where there are certain " middle transitions ", indicates the subjection degree of each element x, be equal to specify one
A fuzzy set.Here, a fuzzy membership function is defined.The expression formula of fuzzy membership function is as follows:
Wherein, Δ x (i, j) is the gradient direction of pixel (x, y) and the difference of its surrounding pixel point gradient direction, above
Formula illustrate the subjection degree of pixel and its affiliated area in image.If the gradient side of a pixel and its affiliated area
Smaller to difference, then the degree of membership of the pixel is bigger, on the contrary then degree of membership is smaller.By the connected region of region consistency to be verified
Gather as one, can be closed in the collection and define a fuzzy entropy.The expression formula of fuzzy entropy is as follows:
Hm(um(x (i, j)))=
-(um(x (i, j))) log2(um(x (i, j)))-(1-um(x (i, j))) log2(1-um(x (i, j)))
For the Fog property of the fuzzy set of measure definitions on the image, need to define a fuzzy entropy measure.It is fuzzy
The expression formula of entropy measure is as follows:
Wherein, M, N represent the width and height in region to be determined, Em(A) it is consistent to represent region gradient direction to be determined
The fuzzy entropy measure of the connected region of property.By to the fuzzy entropy measure being calculated.A fuzzy entropy measure threshold value is set, when
When the fuzzy entropy measure in region to be determined is not more than the given threshold, decide that the region to be determined is gradient direction consistency
Region decides that the region to be determined is gradient direction non-one when the fuzzy entropy measure in region to be determined is greater than given threshold
The region of cause property.Connected region inside gradient direction nonuniformity region is filtered out from connected region.
Step S150 obtains straightway and straightway along the connected region gradient direction from the connected region
Endpoint.
In this step, the connected region of the gradient direction consistency obtained with previous step carries out data processing, from this
The connected region of gradient direction consistency obtains the pixel of corresponding maximum coordinate and minimum coordinate, and the maximum is sat
The endpoint of mark and the corresponding pixel of minimum coordinate as the straightway, by the maximum coordinate and minimum coordinate pair
The line between pixel answered is as the straightway.
Assuming that extreme coordinates are (x0,y0),(x1,y1) can be obtained straight line slope (k) and length (d) parameter.Slope k
Expression formula is as follows:
The expression formula of line segment length d is as follows:
Step S160 exports straight-line detection as a result, the straight-line detection result includes the straightway and the straightway
Endpoint.
Image is connected by filtering out the noise in image by the above-mentioned line detection method calculated based on gradient
Logical region detection obtains connected region, then filters out the poor region of gradient direction consistency in connected region to obtain gradient direction
The connected region of consistency obtains the endpoint of straightway and straightway to the connected region processing of the gradient direction consistency.Phase
Calculation amount is not only reduced for the line detection method of Hough transformation, more multi straight information can also be provided.
Second specific embodiment
Fig. 5 is a kind of structure of the straight-line detection device calculated based on gradient provided for the second specific embodiment of the invention
Block diagram should be calculated in the first specific embodiment based on gradient straight based on the straight-line detection device 200 that gradient calculates
The method of line detection, referring to figure 5., the straight-line detection device 200 calculated based on gradient includes
Computing module 210, for calculating the gradient of pixel in a target image, the gradient includes gradient direction and ladder
Spend amplitude.
Except module 220 of making an uproar, the gradient magnitude is corresponded in the target image less than predetermined gradient amplitude threshold for filtering out
The pixel of value.
Gradient connected region determining module 230, for determining institute according to the gradient direction of pixel in the target image
State the gradient connected region of target image.
In this embodiment, it is preferable that gradient connected region determining module 230 is by pixel in the target image
It is less than the target figure of predetermined angle tolerance in the true observation of gradient direction and the absolute value of the difference of ideal actual value
Pixel is concluded into gradient connected region as in.
Wherein, pixel in the ideal actual value of gradient direction is set as picture in the target image in the target image
Maximum expected value of the vegetarian refreshments on gradient direction.
Module 240 is filtered out, region gradient direction nonuniformity in the gradient connected region for filtering out the target image
Region.
In this embodiment, it is preferable that filter out module 240 and define a fuzzy entropy in the gradient connected region
Measure function is calculated the fuzzy entropy measure in each region in the gradient connected region using the fuzzy entropy measure function, works as institute
When stating fuzzy entropy measure greater than a preset threshold, filter out from the gradient connected region of the target image greater than the preset threshold
The corresponding region of fuzzy entropy measure.
Obtain module 250, for obtained from the connected region along the connected region gradient direction straightway and
The endpoint of straightway.
In this embodiment, it is preferable that the acquisition module 250 is calculated in the gradient connected region corresponding
The pixel of maximum coordinate and minimum coordinate, using the maximum coordinate and the corresponding pixel of minimum coordinate as institute
The endpoint for stating straightway, using the line between the maximum coordinate and the corresponding pixel of minimum coordinate as the straight line
Section.
Output module 260, for exporting straight-line detection as a result, the straight-line detection result includes the straightway and described
The endpoint of straightway.
It should be noted that all the embodiments in this specification are described in a progressive manner, each embodiment weight
Point explanation is the difference from other embodiments, and the same or similar parts between the embodiments can be referred to each other.
In addition, the flow chart and block diagram in the drawings show system, method and the meters of multiple embodiments according to the present invention
The architecture, function and operation in the cards of calculation machine program product.In this regard, each box in flowchart or block diagram
Can represent a part of a module, section or code, a part of the module, section or code include one or
Multiple executable instructions for implementing the specified logical function.It should also be noted that in some implementations as replacements, box
Middle marked function can also occur in a different order than that indicated in the drawings.For example, two continuous boxes are actually
It can be basically executed in parallel, they can also be executed in the opposite order sometimes, and this depends on the function involved.Also it to infuse
Meaning, the combination of each box in block diagram and or flow chart and the box in block diagram and or flow chart can be with holding
The dedicated hardware based system of function or movement as defined in row is realized, or can use specialized hardware and computer instruction
Combination realize.
It should be noted that, in this document, relational terms such as first and second and the like are used merely to a reality
Body or operation are distinguished with another entity or operation, are deposited without necessarily requiring or implying between these entities or operation
In any actual relationship or order or sequence.Moreover, the terms "include", "comprise" or its any other variant are intended to
Non-exclusive inclusion, so that the process, method, article or equipment including a series of elements is not only wanted including those
Element, but also including other elements that are not explicitly listed, or further include for this process, method, article or equipment
Intrinsic element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that
There is also other identical elements in process, method, article or equipment including the element.
The foregoing is only a preferred embodiment of the present invention, is not intended to restrict the invention, for the skill of this field
For art personnel, the invention may be variously modified and varied.All within the spirits and principles of the present invention, made any to repair
Change, equivalent replacement, improvement etc., should all be included in the protection scope of the present invention.It should also be noted that similar label and letter exist
Similar terms are indicated in following attached drawing, therefore, once being defined in a certain Xiang Yi attached drawing, are then not required in subsequent attached drawing
It is further defined and explained.
Claims (8)
1. a kind of line detection method calculated based on gradient characterized by comprising
The gradient of pixel in a target image is calculated, the gradient includes gradient direction and gradient magnitude；
It filters out and corresponds to the pixel that the gradient magnitude is less than predetermined gradient amplitude thresholds in the target image；
The gradient connected region of the target image is determined according to the gradient direction of pixel in the target image；
Filter out the nonconforming region in region gradient direction in the gradient connected region of the target image；
The endpoint of the straightway and straightway along the connected region gradient direction is obtained from the connected region；
Straight-line detection is exported as a result, the straight-line detection result includes the endpoint of the straightway and the straightway；
Wherein, the gradient direction according to pixel in the target image determines the gradient connected region of the target image
The step of include:
Absolute value of the difference by pixel in the target image in the true observation and ideal actual value of gradient direction is less than
Pixel is concluded into gradient connected region in the target image of predetermined angle tolerance.
2. a kind of line detection method calculated based on gradient as described in claim 1, it is characterised in that:
Pixel is set as in the target image pixel in ladder in the ideal actual value of gradient direction in the target image
Spend the maximum expected value on direction.
3. a kind of line detection method calculated based on gradient as described in claim 1, it is characterised in that: filter out the target
The step of nonconforming region in region gradient direction, includes: in the gradient connected region of image
A fuzzy entropy measure function is defined in the gradient connected region, using described in fuzzy entropy measure function calculating
The fuzzy entropy measure in each region in gradient connected region, when the fuzzy entropy measure is greater than a preset threshold, from the target
The gradient connected region of image filters out the corresponding region of fuzzy entropy measure greater than the preset threshold.
4. a kind of line detection method calculated based on gradient as described in claim 1, which is characterized in that from the connected region
In domain obtain along the connected region gradient direction straightway and straightway endpoint the step of include:
The pixel that maximum coordinate and minimum coordinate are corresponded in the gradient connected region is calculated, by the maximum
The endpoint of coordinate and the corresponding pixel of minimum coordinate as the straightway, by the maximum coordinate and minimum coordinate
Line between corresponding pixel is as the straightway.
5. a kind of straight-line detection device calculated based on gradient characterized by comprising
Computing module, for calculating the gradient of pixel in a target image, the gradient includes gradient direction and gradient magnitude；
Except module of making an uproar, for filtering out the pixel for corresponding to the gradient magnitude in the target image and being less than predetermined gradient amplitude thresholds
Point；
Gradient connected region determining module, for determining the target figure according to the gradient direction of pixel in the target image
The gradient connected region of picture；
Module is filtered out, the nonconforming region in region gradient direction in the gradient connected region for filtering out the target image；
Module is obtained, for obtaining straightway and straightway along the connected region gradient direction from the connected region
Endpoint；
Output module, for exporting straight-line detection as a result, the straight-line detection result includes the straightway and the straightway
Endpoint；
Wherein, the gradient connected region determining module by pixel in the target image gradient direction true observation
It is less than pixel in the target image of predetermined angle tolerance with the absolute value of the difference of ideal actual value to conclude into gradient company
Logical region.
6. a kind of straight-line detection device calculated based on gradient as claimed in claim 5, it is characterised in that: the target image
Middle pixel is set as maximum phase of the pixel on gradient direction in the target image in the ideal actual value of gradient direction
Prestige value.
7. a kind of straight-line detection device calculated based on gradient as claimed in claim 5, it is characterised in that:
The module that filters out defines a fuzzy entropy measure function in the gradient connected region, using the fuzzy entropy measure
Function calculates the fuzzy entropy measure in each region in the gradient connected region, when the fuzzy entropy measure is greater than a preset threshold
When, the corresponding region of fuzzy entropy measure greater than the preset threshold is filtered out from the gradient connected region of the target image.
8. a kind of straight-line detection device calculated based on gradient as claimed in claim 5, it is characterised in that:
The pixel that maximum coordinate and minimum coordinate are corresponded in the gradient connected region is calculated in the acquisition module,
Using the maximum coordinate and the corresponding pixel of minimum coordinate as the endpoint of the straightway, by the maximum coordinate
And the line between the corresponding pixel of minimum coordinate is as the straightway.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510848072.6A CN105261031B (en) | 2015-11-26 | 2015-11-26 | The line detection method and device calculated based on gradient |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510848072.6A CN105261031B (en) | 2015-11-26 | 2015-11-26 | The line detection method and device calculated based on gradient |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105261031A CN105261031A (en) | 2016-01-20 |
| CN105261031B true CN105261031B (en) | 2019-01-15 |
Family
ID=55100703
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510848072.6A Active CN105261031B (en) | 2015-11-26 | 2015-11-26 | The line detection method and device calculated based on gradient |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105261031B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106056619A (en) * | 2016-06-13 | 2016-10-26 | 长安大学 | Unmanned aerial vehicle vision wire patrol method based on gradient constraint Radon transform |
| CN106980851B (en) * | 2017-03-21 | 2020-02-14 | 浙江华睿科技有限公司 | Method and device for positioning data matrix DM code |
| CN108010020B (en) * | 2017-11-29 | 2020-05-26 | 浙江华睿科技有限公司 | Silicon wafer slide detection method and device |
| CN110874844A (en) | 2018-08-30 | 2020-03-10 | 京东方科技集团股份有限公司 | Line segment detection method, device and equipment |
| CN111429399B (en) * | 2019-01-09 | 2023-08-01 | 银河水滴科技（北京）有限公司 | Linear detection method and device |
| CN111652825B (en) * | 2020-08-06 | 2020-11-13 | 上海海栎创微电子有限公司 | Edge tracking straight line segment rapid detection device and method based on gradient direction constraint |
| CN112541914B (en) * | 2020-12-25 | 2022-06-03 | 三峡大学 | Short line filtering method for image recognition |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103020967A (en) * | 2012-12-07 | 2013-04-03 | 青岛经纬蓝图信息技术有限公司 | Unmanned aerial vehicle aerial image accurate matching method based on island edge characteristics |
| CN103177261A (en) * | 2013-03-06 | 2013-06-26 | 北方民族大学 | Image-recognition-technology-based cow milk yield auto-metering system and image recognition method therefor |
| CN103279755A (en) * | 2013-03-25 | 2013-09-04 | 东莞中国科学院云计算产业技术创新与育成中心 | Vehicle bottom shadow characteristics-based rapid license plate positioning method |
-
2015
- 2015-11-26 CN CN201510848072.6A patent/CN105261031B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103020967A (en) * | 2012-12-07 | 2013-04-03 | 青岛经纬蓝图信息技术有限公司 | Unmanned aerial vehicle aerial image accurate matching method based on island edge characteristics |
| CN103177261A (en) * | 2013-03-06 | 2013-06-26 | 北方民族大学 | Image-recognition-technology-based cow milk yield auto-metering system and image recognition method therefor |
| CN103279755A (en) * | 2013-03-25 | 2013-09-04 | 东莞中国科学院云计算产业技术创新与育成中心 | Vehicle bottom shadow characteristics-based rapid license plate positioning method |
Non-Patent Citations (3)
| Title |
|---|
| LSD：a Line Segment Detector;Rafael Grompone von Gioi等;《Image Processing on line》;20120324;35-55 |
| 一种基于梯度的直线段检测算法;覃勋辉等;《光子学报》;20120229;第41卷(第2期);第205-209页 |
| 基于图像模糊熵邻域非一致性的过渡区提取与分割;王彦春等;《电子学报》;20081231;第36卷(第12期);第2445-2449页 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105261031A (en) | 2016-01-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105261031B (en) | The line detection method and device calculated based on gradient | |
| CN104268857B (en) | A kind of fast sub-picture element rim detection and localization method based on machine vision | |
| CN104376319B (en) | A kind of method based on anisotropic Gaussian core extraction closed edge image outline | |
| CN103116875B (en) | Self-adaptation bilateral filtering image de-noising method | |
| CN104915963B (en) | A kind of detection and localization method for PLCC elements | |
| CN105849774B (en) | Image contrast device, imaging sensor, processing system, image contrast method | |
| CN106197612B (en) | A kind of transparent bottled liquid-level detecting method based on machine vision | |
| CN103292701A (en) | Machine-vision-based online dimensional measurement method of precise instrument | |
| CN104408711A (en) | Multi-scale region fusion-based salient region detection method | |
| Chen et al. | Guidelines for DIC in geotechnical engineering research | |
| CN103426184A (en) | Optical flow tracking method and device | |
| CN105387801B (en) | A kind of subway tunnel section of jurisdiction faulting of slab ends quantity measuring method | |
| CN106340010B (en) | A kind of angular-point detection method based on second order profile difference | |
| CN104657979B (en) | A kind of features of ultrasound pattern detection method and system | |
| CN103150725B (en) | Based on SUSAN edge detection method and the system of non-local mean | |
| CN105469380A (en) | Method and device for detecting shielding against object | |
| CN105761256A (en) | Image sub-pixel edge line acquisition method and device | |
| Liao et al. | Enclosed laplacian operator of nonlinear anisotropic diffusion to preserve singularities and delete isolated points in image smoothing | |
| CN105662412B (en) | A kind of method for shimming first order of magnetic resonance system, device and equipment | |
| Awrangjeb et al. | An affine resilient curvature scale-space corner detector | |
| CN104949621A (en) | Boundary positioning method of stripes of grating ruler | |
| CN113221371A (en) | Method and device for determining critical sliding surface of side slope and terminal equipment | |
| CN105425296B (en) | Geologic body recognition methods and device | |
| KR101208647B1 (en) | Method and apparatus for detecting obstacle on road | |
| Zhang et al. | Stereo matching cost computation based on nonsubsampled contourlet 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 |