CN107230230A - A kind of Instrument image localization method based on composite filter - Google Patents
A kind of Instrument image localization method based on composite filter Download PDFInfo
- Publication number
- CN107230230A CN107230230A CN201710446051.0A CN201710446051A CN107230230A CN 107230230 A CN107230230 A CN 107230230A CN 201710446051 A CN201710446051 A CN 201710446051A CN 107230230 A CN107230230 A CN 107230230A
- Authority
- CN
- China
- Prior art keywords
- mrow
- instrumentation
- composite filter
- mfrac
- localization method
- 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
- 238000000034 method Methods 0.000 title claims abstract description 31
- 239000002131 composite material Substances 0.000 title claims abstract description 25
- 230000004807 localization Effects 0.000 title claims abstract description 16
- 238000012549 training Methods 0.000 claims abstract description 24
- 230000004044 response Effects 0.000 claims description 13
- 239000011159 matrix material Substances 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 4
- 230000015572 biosynthetic process Effects 0.000 claims description 2
- 230000021615 conjugation Effects 0.000 claims description 2
- 238000003786 synthesis reaction Methods 0.000 claims description 2
- 241000208340 Araliaceae Species 0.000 claims 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 claims 1
- 235000003140 Panax quinquefolius Nutrition 0.000 claims 1
- 235000008434 ginseng Nutrition 0.000 claims 1
- 230000008859 change Effects 0.000 abstract description 5
- 238000002203 pretreatment Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 230000009466 transformation Effects 0.000 description 3
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000009432 framing Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
- G06T7/75—Determining position or orientation of objects or cameras using feature-based methods involving models
-
- G06T5/70—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/80—Analysis of captured images to determine intrinsic or extrinsic camera parameters, i.e. camera calibration
Abstract
The present invention provides a kind of Instrument image localization method based on composite filter, by training in advance, composite filter is obtained, the training pattern parameter of instrumentation image is obtained, recycles the training pattern parameter of instrumentation image to obtain coordinate points of the instrumentation in panoramic picture.The present invention can accurately obtain position of the instrumentation in panoramic picture to different instrument equipment, varying environment change and different background condition.
Description
Technical field
The invention belongs to Computer Image Processing field, it is related to a kind of Instrument image localization method, in particular it relates to a kind of
Instrument image localization method based on composite filter.
Background technology
Electric inspection process robot is in Instrument image equipment identification process is carried out, it is necessary first to navigated in panoramic picture
Equipment region, carries out the operation of camera zoom to obtain clearly instrumentation image further according to the equipment region navigated to, is easy to
Image recognition.If equipment positioning failure in panoramic picture, follow-up all operations will be invalid, thus, design a kind of accurate height
The panorama sketch Instrument image localization method of effect has important practical significance.
Instrument image localization method mainly includes template matching method, Feature Points Matching method etc..Template matching method is according to mark
Fixed instrumentation template image, the similarity of image to be matched and template image is calculated using some similarity evaluation functions,
Similarity is more high, illustrate be instrument region possibility it is higher, it is generally the case that template matching method can be accurately positioned instrument
Equipment region, but when light change or complicated background, template matching method often provides multiple peak values, it is impossible to be accurately positioned
Target location;All characteristic points in Feature Points Matching method instrumentation equipment drawing picture first and panoramic picture, then carry out two-by-two
Match somebody with somebody and carry out matching checking, finally give the target location of matching, change greatly in light, background it is complicated in the case of, feature
Point match method can still obtain good effect, but when characteristic point or characteristic point is not present than sparse in instrumentation in itself
When, this method will fail, and the usual amount of calculation of Feature Points Matching method is larger.
The content of the invention
To solve the problem of prior art is present, the present invention provides a kind of instrumentation framing side of efficiently and accurately
Method, this method can accurately obtain instrumentation in panorama to different instrument equipment, varying environment change and different background condition
Position in image.
The Instrument image localization method based on composite filter that the present invention is provided, comprises the following steps;
Instrumentation model training:
Define dimensional Gaussian receptance functionWherein, p (x0,y0) for instrumentation in panorama
Scheme the centre coordinate in f (x, y), σ is Gaussian response radius, 0≤x≤M-1,0≤y≤N-1, and M, N is respectively the width of image
With height.Dimensional Gaussian receptance function g (x, y) discrete Fourier transform can be expressed as
Wherein exp is the exponential function using natural constant e the bottom of as, and j is imaginary unit, x=0,1 ..., M-1, y=0,
1,…,N-1;U=0,1 ..., M-1, v=0,1 ..., N-1;U, v are nonnegative integer;Define convolution kernel h (x, y) so that set up
ThenWherein F, H are respectively f, h two-dimentional Fourier
Conversion,Operated for two-dimentional inverse Fourier transform, " " operates for two-dimensional matrix element dot product;
Convolution operation is replaced with associative operation, then G (u, v)=F (u, v) H*(u, v), wherein * grasp for the conjugation of matrix
Make, thus composite filter is defined asH*(u, v) is the training pattern parameter of instrumentation image;
Instrumentation is positioned:
For panorama sketch f to be positionede, its corresponding dimensional Gaussian, which is responded, is expressed as Ge(u, v)=Fe(u,v)·H*(u,
v);By frequency domain Ge(u, v) transforms to spatial domain ge(x, y),ge(x, y) maximum is corresponding
Coordinate points are instrumentation in panorama sketch feCenter.
In order to obtain more accurate model parameter, in instrumentation model training step, generally different images are instructed
The training pattern parameter got carries out average operation, specifically, setting training sequence of pictures as fi(1≤i≤L), wherein L is figure
Piece number, i.e. total sample number, picture fiCorresponding instrumentation centre coordinate is pi(x0,y0) (1≤i≤L), corresponding two dimension height
This response is gi(1≤i≤L), then averagely composite filter is defined as
In instrumentation positioning step, panorama sketch f is obtained using average composite filtereDimensional Gaussian response, i.e.,
In order to increase the robustness of average composite filter, it is set more to stablize, generally one regular coefficient ε of increase, i.e. flat
Equal composite filter is expressed asThe effect of regular coefficient is to avoid denominator from being led for 0
Cause numerical computations unstable.
, need to be to complete before instrumentation model training in order to reduce the image boundary effect during Fourier transformation
Scape figure f (x, y) is carried out plus Cosine Window processing, i.e. f (x, y) ← f (x, y) w (x, y), wherein,
In order to reduce the interference of the factors such as light, original panoramic images need to be carried out with some pretreatments, i.e. fe(x,y)←log
(fe(x, y)+1),WhereinFor the average of image.In instrument
In device model training step, it is also desirable to which panorama sketch f (x, y) is pre-processed i.e., f (x, y) ← log (f (x, y)+1), f
(x,y)←f(x,y)-μf,Wherein μfFor the average of image.
Instrumentation model training needs a large amount of panorama sketch, and the camera that the present invention is carried using crusing robot captures panorama
Figure, carries out instrumentation image calibration, specifically, capturing panorama sketch of the different time sections containing instrumentation using camera, obtains
Centre coordinate of the instrumentation in panoramic picture, wherein, in the coordinate and panoramic picture of the photocentre of camera in panoramic picture
Heart coordinate is overlapped, also, centre coordinate of the instrumentation in panoramic picture is overlapped with picture centre coordinate.
The invention has the advantages that:(1) the Instrument image localization method based on average synthesis accurate filtering device can
More accurately and efficiently to navigate to instrumentation position in the picture;(2) rings such as light change, background complexity can be overcome
Border factor interference, realizes that instrumentation is stable and accurately positions;(3) identification of crusing robot instrumentation can be greatly enhanced
Rate.
Brief description of the drawings
Fig. 1 panorama sketch instrumentation demarcates schematic diagram;
Fig. 2 panorama sketch positioning result figures;
Fig. 3 panorama sketch location response figures.
Embodiment
Most highly preferred embodiment of the invention is illustrated below in conjunction with accompanying drawing:
By taking certain pointer-type gauges equipment in certain transformer station as an example, the Instrument image positioning side based on average composite filter
Method, is carried out according to the following steps:
1) instrumentation image calibration:Calibration for cameras photocentre is carried out in advance to rectify standard, it is ensured that camera photocentre is in the picture
Coordinate is overlapped with picture centre coordinate.Crusing robot is during patrol task, by adjusting head parameter so that instrument is set
Standby centre coordinate is moved to picture centre, then carries out camera zoom so that target is clear, and details is visible.In whole zoom process
In, the centre coordinate of instrumentation image only need to be obtained, thus only need gauged instrument to set during instrumentation image calibration
The centre coordinate of standby image, the dimension information without instrument.
Instrumentation image calibration is exactly the panorama sketch for collecting the candid photograph of different time sections crusing robot, and obtains equipment and exist
Centre coordinate in image, as shown in Figure 1.
2) instrumentation model training:Panorama sketch f (x, y) (0≤x≤M-1,0≤y≤N-1), wherein M, N is respectively figure
The centre coordinate of instrumentation in the picture is p (x in the width of picture and height, panorama sketch0,y0), i.e. step 1) in demarcation
Centre coordinate.Dimensional Gaussian receptance function g (x, y) (0≤x≤M-1,0≤y≤N-1) is defined,
σ represents Gaussian response radius, and value is bigger, and response radius is bigger, and the tolerance to noise is higher, and σ value is according to warp
Selection is tested, those skilled in the art will know that obtaining selection σ=2 in suitable value, the present embodiment by way of experiment, two dimension is high
This receptance function g (x, y) is in centre coordinate p (x0,y0) functional value at place reaches maximum 1, away from p (x0,y0) place function
It is worth close to 0;Dimensional Gaussian receptance function g (x, y) discrete Fourier transform can be expressed as
Wherein exp is the exponential function using natural constant e the bottom of as, and j is imaginary unit, x=0,1 ..., M-1, y=0,
1,…,N-1;U=0,1 ..., M-1, v=0,1 ..., N-1;U, v are nonnegative integer.Define convolution kernel h (x, y) (0≤x≤M-
1,0≤y≤N-1) so that set upWhereinFor two-dimensional convolution operator.It is desirable that complete
One dimensional Gaussian of output is responded after scape figure f (x, y) is operated by convolution kernel h (x, y), and the response is sat at the center of instrumentation
Punctuate reaches peak, and the response away from central point is close to 0.According to two-dimensional convolution theorem, set upWherein F, H are respectively f and h two-dimensional Fourier transform,For
Two-dimentional inverse Fourier transform operation, is the operation of two-dimensional matrix element dot product.
Convolution operation is replaced with associative operation, then sets up G (u, v)=F (u, v) H*(u, v), wherein G are g two-dimentional Fu
In leaf transformation, * be matrix conjugate operation, accordingly, composite filter may be defined as
H*(u, v) is the training pattern parameter of the instrumentation image.
In the present embodiment, in order to reduce the image boundary effect during Fourier transformation, in instrumentation model
Before training, panorama sketch need to be carried out plus Cosine Window processing, i.e.,
F (x, y) ← f (x, y) w (x, y),
Wherein,
In the present embodiment, in order to reduce the interference of the factors such as light, some pretreatments need to be carried out to panoramic picture f (x, y),
That is,
F (x, y) ← log (f (x, y)+1),
f(x,y)←f(x,y)-μf,
Wherein μfFor the average of image.
In a practical situation, in order to obtain more accurate model parameter, different images are generally trained to obtained model
Parameter carries out average operation, specifically, if training sequence of pictures is fi(1≤i≤L), wherein L is picture number, the present embodiment choosing
It is 8, picture f to take L=8, i.e. total sample numberiCorresponding instrumentation centre coordinate is pi(x0,y0) (1≤i≤L), corresponding two
Dimension Gaussian response is gi(1≤i≤L), then averagely composite filter is defined as
In order to increase the robustness of average composite filter, it is set more to stablize, generally one regular coefficient ε of increase,
It is vertical
The effect of regular coefficient is to avoid denominator from causing numerical computations unstable for 0, those skilled in the art will know that according to
Need to be chosen, suitable value is chosen by testing, the present embodiment chooses ε=0.1.
3) instrumentation is positioned:Obtaining the training pattern parameter of instrumentation imageAfterwards, for be positioned
Panorama sketch fe, its corresponding two dimension, which is responded, to be represented by
In the present embodiment, need to be to original panoramic images f in order to reduce the interference of the factors such as lighteSome pretreatments are carried out,
That is,
fe(x,y)←log(fe(x, y)+1),
WhereinFor the average of image.
Again by Ge(u, v) transforms to spatial domain ge(x, y),
geThe corresponding coordinate points of (x, y) maximum are instrumentation in panorama sketch feCenter, panorama sketch positioning knot
Fruit is as shown in Fig. 2 corresponding response is as shown in Figure 3.After the completion of instrumentation positioning, you can the numerical value on instrumentation is carried out
Read.
Claims (7)
1. a kind of Instrument image localization method based on composite filter, it is characterised in that comprise the following steps;
Instrumentation model training:
Define dimensional Gaussian receptance functionWherein, p (x0,y0) for instrumentation panorama sketch f (x,
Y) centre coordinate in, σ be Gaussian response radius, 0≤x≤M-1,0≤y≤N-1, M, N be respectively image width with height;
Dimensional Gaussian receptance function g (x, y) discrete Fourier transform is expressed as
<mrow>
<mi>G</mi>
<mrow>
<mo>(</mo>
<mi>u</mi>
<mo>,</mo>
<mi>v</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>x</mi>
<mo>=</mo>
<mn>0</mn>
</mrow>
<mrow>
<mi>M</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</munderover>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>y</mi>
<mo>=</mo>
<mn>0</mn>
</mrow>
<mrow>
<mi>N</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</munderover>
<mi>g</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<msup>
<mi>exp</mi>
<mrow>
<mo>-</mo>
<mi>j</mi>
<mn>2</mn>
<mi>&pi;</mi>
<mrow>
<mo>(</mo>
<mfrac>
<mrow>
<mi>x</mi>
<mi>u</mi>
</mrow>
<mi>M</mi>
</mfrac>
<mo>+</mo>
<mfrac>
<mrow>
<mi>y</mi>
<mi>v</mi>
</mrow>
<mi>N</mi>
</mfrac>
<mo>)</mo>
</mrow>
</mrow>
</msup>
<mo>,</mo>
</mrow>
Wherein exp is the exponential function using natural constant e the bottom of as, and j is imaginary unit, x=0,1 ..., M-1, y=0,1 ..., N-
1;U=0,1 ..., M-1, v=0,1 ..., N-1;U, v are nonnegative integer;Define convolution kernel h (x, y) so that set up
<mrow>
<mi>f</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>&CircleTimes;</mo>
<mi>h</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mi>g</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>,</mo>
</mrow>
ThenWherein F, H are respectively f, h two-dimensional Fourier transform,Operated for two-dimentional inverse Fourier transform, " " operates for two-dimensional matrix element dot product;
Convolution operation is replaced with associative operation, then sets up G (u, v)=F (u, v) H*(u, v), wherein * grasp for the conjugation of matrix
Make, thus, composite filter is defined asH*(u, v) is the training pattern ginseng of instrumentation image
Number;
Instrumentation is positioned:
For panorama sketch f to be positionede, its corresponding dimensional Gaussian, which is responded, is expressed as Ge(u, v)=Fe(u,v)·H*(u,v);Will
Frequency domain Ge(u, v) transforms to spatial domain ge(x, y),geThe corresponding coordinate points of (x, y) maximum
As instrumentation is in panorama sketch feCenter.
2. the Instrument image localization method as claimed in claim 2 based on composite filter, it is characterised in that in instrumentation
In model training step, obtain training pattern parameter for the training of different panorama sketch and carry out average operation, specifically, setting training figure
Piece sequence is fi(1≤i≤L), wherein L is picture number, i.e. total sample number, picture fiCorresponding instrumentation centre coordinate is pi
(x0,y0) (1≤i≤L), corresponding dimensional Gaussian response is gi(1≤i≤L), then averagely composite filter is defined as
<mrow>
<msubsup>
<mi>H</mi>
<mi>&mu;</mi>
<mo>*</mo>
</msubsup>
<mrow>
<mo>(</mo>
<mi>u</mi>
<mo>,</mo>
<mi>v</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<mi>L</mi>
</mfrac>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>L</mi>
</munderover>
<msup>
<mi>H</mi>
<mo>*</mo>
</msup>
<mrow>
<mo>(</mo>
<mi>u</mi>
<mo>,</mo>
<mi>v</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<mi>L</mi>
</mfrac>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>L</mi>
</munderover>
<mfrac>
<mrow>
<msub>
<mi>G</mi>
<mi>i</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>u</mi>
<mo>,</mo>
<mi>v</mi>
<mo>)</mo>
</mrow>
<mo>&CenterDot;</mo>
<msubsup>
<mi>F</mi>
<mi>i</mi>
<mo>*</mo>
</msubsup>
<mrow>
<mo>(</mo>
<mi>u</mi>
<mo>,</mo>
<mi>v</mi>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<msub>
<mi>F</mi>
<mi>i</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>u</mi>
<mo>,</mo>
<mi>v</mi>
<mo>)</mo>
</mrow>
<mo>&CenterDot;</mo>
<msubsup>
<mi>F</mi>
<mi>i</mi>
<mo>*</mo>
</msubsup>
<mrow>
<mo>(</mo>
<mi>u</mi>
<mo>,</mo>
<mi>v</mi>
<mo>)</mo>
</mrow>
</mrow>
</mfrac>
<mo>,</mo>
</mrow>
In instrumentation positioning step, panorama sketch f is obtained using average composite filtereDimensional Gaussian response, i.e.,
3. the Instrument image localization method as claimed in claim 2 based on composite filter, it is characterised in that average synthesis filter
Ripple device is changed toWherein ε is regular coefficient.
4. the Instrument image localization method as claimed in claim 1 based on composite filter, it is characterised in that in instrumentation
Before model training, first panoramic picture f (x, y) is carried out plus Cosine Window processing, i.e. f (x, y) ← f (x, y) w (x, y), wherein,
5. the Instrument image localization method as claimed in claim 1 based on composite filter, it is characterised in that to panorama sketch fe
Pre-processed, i.e. fe(x,y)←log(fe(x, y)+1), Its
InFor the average of image.
6. the Instrument image localization method as claimed in claim 1 based on composite filter, it is characterised in that in instrumentation
In model training step, panorama sketch f (x, y) is pre-processed i.e., f (x, y) ← log (f (x, y)+1), f (x, y) ← f (x,
y)-μf,Wherein μfFor the average of image.
7. the Instrument image localization method based on composite filter as described in claim 1-6 any claims, its feature
It is, in addition to instrumentation image calibration step, specifically, capturing panorama of the different time sections containing instrumentation using camera
Figure, obtains centre coordinate of the instrumentation in panoramic picture, wherein, the coordinate and panorama of the photocentre of camera in panoramic picture
Picture centre coordinate is overlapped, also, centre coordinate of the instrumentation in panoramic picture is overlapped with picture centre coordinate.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710446051.0A CN107230230B (en) | 2017-06-14 | 2017-06-14 | Instrument image positioning method based on synthesis filter |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710446051.0A CN107230230B (en) | 2017-06-14 | 2017-06-14 | Instrument image positioning method based on synthesis filter |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107230230A true CN107230230A (en) | 2017-10-03 |
CN107230230B CN107230230B (en) | 2020-03-24 |
Family
ID=59934940
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710446051.0A Active CN107230230B (en) | 2017-06-14 | 2017-06-14 | Instrument image positioning method based on synthesis filter |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107230230B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108921177A (en) * | 2018-06-22 | 2018-11-30 | 重庆邮电大学 | The instrument localization method of Intelligent Mobile Robot |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20120051423A1 (en) * | 2010-09-01 | 2012-03-01 | Electronics And Telecommunications Research Institute | Video processing method and apparatus based on multiple texture images using video excitation signals |
CN105678344A (en) * | 2016-02-29 | 2016-06-15 | 浙江群力电气有限公司 | Intelligent classification method for power instrument equipment |
CN105741266A (en) * | 2016-01-22 | 2016-07-06 | 北京航空航天大学 | Pathological image cell nucleus quick location method |
CN105825215A (en) * | 2016-03-15 | 2016-08-03 | 云南大学 | Instrument positioning method based on local neighbor embedded kernel function and carrier of method |
-
2017
- 2017-06-14 CN CN201710446051.0A patent/CN107230230B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20120051423A1 (en) * | 2010-09-01 | 2012-03-01 | Electronics And Telecommunications Research Institute | Video processing method and apparatus based on multiple texture images using video excitation signals |
CN105741266A (en) * | 2016-01-22 | 2016-07-06 | 北京航空航天大学 | Pathological image cell nucleus quick location method |
CN105678344A (en) * | 2016-02-29 | 2016-06-15 | 浙江群力电气有限公司 | Intelligent classification method for power instrument equipment |
CN105825215A (en) * | 2016-03-15 | 2016-08-03 | 云南大学 | Instrument positioning method based on local neighbor embedded kernel function and carrier of method |
Non-Patent Citations (4)
Title |
---|
JOHN S.等: "TRANSFORMS,FILTERS AND EDGE DETECTORS IN IMAGE PROCESSING", 《INTERNATIONAL JOURNAL OF PUREAND APPLIED MATHEMATICS》 * |
XIAOPING L.等: "Comparative study on wavelet transform and traditional image processing", 《THE 3RD INTERNATIONAL CONFERENCE ON INFORMATION SCIENCES AND INTERACTION SCIENCES》 * |
吕瑞峰 等: "变电站智能巡检系统仪表定位算法设计与实现", 《电子制作》 * |
李辉: "基于机器视觉的仪表数字识别研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108921177A (en) * | 2018-06-22 | 2018-11-30 | 重庆邮电大学 | The instrument localization method of Intelligent Mobile Robot |
Also Published As
Publication number | Publication date |
---|---|
CN107230230B (en) | 2020-03-24 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11244197B2 (en) | Fast and robust multimodal remote sensing image matching method and system | |
CN103955694B (en) | Image recognition meter reading system and method | |
CN102541954B (en) | Method and system for searching trademarks | |
CN109872305B (en) | No-reference stereo image quality evaluation method based on quality map generation network | |
CN110706293B (en) | SURF feature matching-based electronic component positioning and detecting method | |
CN103914847A (en) | SAR image registration method based on phase congruency and SIFT | |
CN103218825B (en) | Quick detection method of spatio-temporal interest points with invariable scale | |
CN107122787A (en) | A kind of image scaling quality evaluating method of feature based fusion | |
CN102509304A (en) | Intelligent optimization-based camera calibration method | |
CN106327422A (en) | Image stylized reconstruction method and device | |
CN110895697A (en) | Transformer nameplate information acquisition method and intelligent acquisition system | |
CN116448769A (en) | Multi-mode information fusion plate defect detection system and detection method thereof | |
CN112749752A (en) | Hyperspectral image classification method based on depth transform | |
CN109993116B (en) | Pedestrian re-identification method based on mutual learning of human bones | |
CN107230230A (en) | A kind of Instrument image localization method based on composite filter | |
Panetta et al. | Unrolling post-mortem 3D fingerprints using mosaicking pressure simulation technique | |
CN106709915A (en) | Image resampling operation detection method | |
CN110321869A (en) | Personnel's detection and extracting method based on Multiscale Fusion network | |
CN106204593A (en) | A kind of Feature Points Extraction based on sequence image synthesis | |
CN105956530A (en) | Image correction method and image correction device | |
CN107622476A (en) | Image Super-resolution processing method based on generative probabilistic model | |
CN110084736A (en) | A kind of method of detecting watermarks and system based on SURF and pyramid algorith | |
CN113111850B (en) | Human body key point detection method, device and system based on region-of-interest transformation | |
CN110599398A (en) | Online image splicing and fusing method based on wavelet technology | |
CN112184835A (en) | Method and system for restoring deformation of piano keyboard image |
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 |