CN107273803A - Cloud layer image detecting method - Google Patents
Cloud layer image detecting method Download PDFInfo
- Publication number
- CN107273803A CN107273803A CN201710345541.1A CN201710345541A CN107273803A CN 107273803 A CN107273803 A CN 107273803A CN 201710345541 A CN201710345541 A CN 201710345541A CN 107273803 A CN107273803 A CN 107273803A
- Authority
- CN
- China
- Prior art keywords
- mrow
- detection means
- msub
- remote sensing
- mtd
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/13—Satellite images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/462—Salient features, e.g. scale invariant feature transforms [SIFT]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Astronomy & Astrophysics (AREA)
- Remote Sensing (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The present invention provides a kind of cloud layer image detecting method, including;Detection means receives the remote sensing images sequence for including N frame remote sensing images;Detection means obtains energy Saliency maps from the energy feature of remote sensing images sequential extraction procedures remote sensing images;Detection means calculates the luminance contrast of the energy Saliency maps, obtains luminance contrast image;Detection means extracts the textural characteristics of luminance contrast figure, obtains textural characteristics figure;Detection means utilizes image motion information, and the textural characteristics figure is carried out into intra-frame trunk, extracts area-of-interest, obtains nutritious obesity result.The requirement to sensor is reduced, while position and the size of cirrus can be accurately detected;And the algorithm used is simple, and computational efficiency is high, can meet requirement of real-time.
Description
Technical field
The present invention relates to remote sensing technology field, particularly relate to a kind of based on energy feature, textural characteristics and motion feature
Cloud layer image detecting method.
Background technology
In recent years, remote sensing technology is widely used in fields such as modern military, space-based detection and meteorologic analysis, distant
It is one of key technology of remote sensing technology to feel image interpretation.Remote sensing image interpretation refers to the geometric properties and physics according to image
Property, carries out comprehensive analysis, so that the quality and quantative attribute of object or phenomenon are disclosed, and the mutual pass between them
System, and then study its generation evolution and the regularity of distribution, that is to say, that the thing representated by recognizing them according to characteristics of image
The property of body or phenomenon.Sources for false alarms has large effect for remote sensing image interpretation.And difference is often there is in remote sensing images
Sources for false alarms, the features such as these sources for false alarms have that radiation intensity is high, changed over time.For example:High-altitude cirrus is a kind of important
Sources for false alarms.There is 1/3 to 1/2 area on the earth by cloud cover, for remote sensing images and target detection system, upper cloud layer
It is a kind of main clutter.And cirrus is because its change in shape is fast, move the feature such as changeable, the interpretation to remote sensing images is brought
Certain difficulty.Suitable cirrus detection algorithm is studied, the precision of remotely sensed image and detection system can be improved, it is advantageously implemented
Military and space usage.
In general, cirrus detection often use spectra methods, collection visible ray and it is infrared wait multi-channel data, utilization
Cloud layer is detected with the radiation difference of other ground objects.But this mode is that real-time is not high, and to imaging device
It is required that it is high, it is not easy to the development of cirrus detection.
The content of the invention
The technical problem to be solved in the present invention is to provide a kind of cloud layer image detection based on energy, texture and motion feature
Method, solves the problem of prior art high and medium cirrus is difficult to accurately be detected.
In order to solve the above technical problems, embodiments of the invention provide a kind of cloud layer image detecting method, including:Detection dress
Put the remote sensing images sequence f for receiving and including N frame remote sensing imagesn(x, y), wherein n=1 ..., N are frame number, and N is totalframes;Institute
Detection means is stated from the energy feature of the remote sensing images sequential extraction procedures remote sensing images, energy Saliency maps are obtainedInstitute
The luminance contrast that detection means calculates the energy Saliency maps is stated, luminance contrast image is obtainedThe detection
Device extracts the textural characteristics of the luminance contrast figure, obtains textural characteristics figureThe detection means utilizes image
Movable information, intra-frame trunk is carried out by the textural characteristics figure, extracts region of interest ROI, obtains nutritious obesity result.
Optionally, detection means utilizes image motion information, and the textural characteristics figure is carried out into intra-frame trunk, extracts sense emerging
Interesting region ROI, obtaining nutritious obesity result is specially:
The detection means sets luminance threshold T, and the result to luminance contrast enters row threshold division, obtains segmentation result
For Thn(x,y):The detection means is to the segmentation result Thn
(x, y) carries out opening operation, eliminates isolated bright spot, removes part clutter, while filling hole, the segmentation result after being handled
Th′n(x,y);The detection means is by Th 'nThe zone marker that pixel value is 1 in (x, y) is region of interest ROI, calculates and obtains
To current ROI centre coordinate (xn,yn);The detection means sets movement threshold Mov, calculates the ROI of n-th frame and the (n+1)th frame
The distance between center:If Dn<Mov, the detection dress
Put, the ROI region of n frames and the (n+1)th frame is associated;The detection means extracts be mutually related ROI region, the phase
The ROI region of mutual correlation is nutritious obesity result.
Optionally, detection means obtains energy notable from the energy feature of the remote sensing images sequential extraction procedures remote sensing images
Property figureSpecially:
The detection means carries out Fourier transformation to the remote sensing images sequence:sn(ωx,ωy)=F [fn(x,y)],n
=1 ..., N, wherein, F represents Fourier transformation operator, (ωx,ωy) represent to transform to the coordinate of frequency domain;The detection means
The amplitude of Fourier transformation is calculated, and takes the logarithm and obtains logarithmic spectrum:Ln(ωx,ωy)=log [| sn(ωx,ωy) |], wherein |
| represent amplitude operator;Calculate phase spectrum P (ωx,ωy):WhereinRepresent phase operator;
The logarithmic spectrum and size are m × m mean filter mask convolution by the detection means, are smoothly composed:V(ωx,ωy)=
Ln(ωx,ωy)*hm(ωx,ωy), wherein mean filter template is:The detection means will be right
Number spectrum and smooth spectrum subtraction, obtain spectrum residual error R (ωx,ωy):R(ωx,ωy)=Ln(ωx,ωy)-V(ωx,ωy), detection dress
Residual error R (ω will be composed by puttingx,ωy) and phase spectrum P (ωx,ωy) two-dimensional discrete Fourier inverse transformation is carried out, obtain energy conspicuousness
Figure
Optionally, detection means described in detection means extracts the textural characteristics of the luminance contrast figure, obtains textural characteristics figureSpecially:Detection means builds wave filter:
Wherein, x '=a-m(xcos θ+ysin θ), y '=a-m(- xcos θ+ysin θ), a-mFor scale factor, θ represents the side of kernel function
To λ represents the wavelength of SIN function, and ψ represents phase offset, and σ represents the standard deviation of Gaussian function, and the width of γ representative functions is high
Than;Wave filter and remote sensing images convolution are obtained filter result by the detection means, and the filter result is textural characteristics figure:
The above-mentioned technical proposal of the present invention has the beneficial effect that:In such scheme, because detection means is using at image
The mode of reason carries out the nutritious obesity in remote sensing images, reduces the requirement to sensor, while volume can be accurately detected
The position of cloud and size;And the algorithm used is simple, and computational efficiency is high, can meet requirement of real-time.
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, below with presently preferred embodiments of the present invention and coordinate accompanying drawing describe in detail it is as follows.
Brief description of the drawings
Fig. 1 is the flow chart of the cloud layer image detecting method of the present invention.
Fig. 2 is one group of an example of the present invention infrared image for containing cirrus.
Fig. 3 is the energy Saliency maps to Fig. 2 infrared images.
Fig. 4 is the luminance contrast image of Fig. 3 energy Saliency maps.
Fig. 5 is the textural characteristics figure of Fig. 4 luminance contrast images.
Fig. 6 is nutritious obesity result.
Embodiment
To make the technical problem to be solved in the present invention, technical scheme and advantage clearer, below in conjunction with accompanying drawing and tool
Body embodiment is described in detail.
The present invention the problem of be difficult to accurately be detected for existing high-altitude cirrus there is provided it is a kind of be based on energy, texture and
The cloud layer image detecting method of motion feature.
As shown in figure 1, embodiments of the invention propose a kind of cloud layer image detection based on energy, texture and motion feature
Method, this method is applied to detection means, specifically includes:
Step 1:Detection means receives N frame remote sensing images, is designated as remote sensing images sequence, this remote sensing images sequence can be used
Function fn(x, y) is represented, wherein n=1 ..., N are frame number, and N is totalframes.For example:Fig. 2 is one group and contains the infrared of cirrus
Remote sensing images.
Step 2:Detection means extracts the energy feature of every frame remote sensing images from remote sensing images sequence respectively, distant according to every frame
Feel the energy feature of image, obtain energy Saliency maps, the energy Saliency maps are designated asFor example:
Fig. 3 is to carry out power feature extraction, the energy Saliency maps of acquisition to Fig. 2 infrared image containing cirrus.
Wherein, the energy feature of every frame remote sensing images here is using spectrum residual error feature.It is comprised the following steps that:
21, detection means carries out Fourier transformation to remote sensing images sequence:
sn(ωx,ωy)=F [fn(x, y)], n=1 ..., N
Wherein, F represents Fourier transformation operator, (ωx,ωy) represent to transform to the coordinate of frequency domain.
22, detection means calculates the amplitude of Fourier transformation, and takes the logarithm and obtain logarithmic spectrum:
Ln(ωx,ωy)=log [| sn(ωx,ωy)|]
Wherein | | represent amplitude operator.Phase spectrum is calculated simultaneously:
WhereinRepresent phase operator.
23, logarithmic spectrum obtained in the previous step and size are m × m mean filter mask convolution by detection means, are put down
Sliding spectrum:
V(ωx,ωy)=Ln(ωx,ωy)*hm(ωx,ωy)
Wherein mean filter template is:
24, logarithmic spectrum and smooth spectrum subtraction are obtained composing residual error by detection means:
R(ωx,ωy)=Ln(ωx,ωy)-V(ωx,ωy)
25, detection means will compose residual error R (ωx,ωy) and phase spectrum P (ωx,ωy) carry out two-dimensional discrete Fourier contravariant
Change, obtain energy Saliency maps:
Step 3:Detection means calculates the luminance contrast of energy Saliency maps, obtains luminance contrast image, the contrast
Degree image is designated asFor example:Fig. 4 is the carry out luminance contrast meter to Fig. 3 energy Saliency maps
The luminance contrast image obtained after calculation.
Step 4:Detection means extracts the textural characteristics of luminance contrast figure, obtains textural characteristics figure, the textural characteristics figure
It is designated asFor example:Fig. 5 is that Fig. 4 luminance contrast image is carried out to obtain after texture feature extraction
Textural characteristics figure.
Here textural characteristics can be represented using direction Gabor characteristic.Specifically obtain textural characteristics figure process bag
Include:
41, the detection means builds wave filter:
Wherein, x '=a-m(xcos θ+ysin θ), y '=a-m(-xcosθ+ysinθ)。a-mFor scale factor, θ is represented
The direction of Gabor kernel functions, λ represents the wavelength of SIN function, and ψ represents phase offset, and σ represents the standard deviation of Gaussian function, γ
The ratio of width to height of representative function.Choose different directions, you can obtain Gabor Multi-aspect filtering devices.
42, Gabor filter and input picture convolution are obtained filter result by detection means, and the filter result is line
Manage characteristic pattern:
Step 5:Detection means utilizes image motion information, and textural characteristics figure is carried out into intra-frame trunk, extracts region of interest
Domain, obtains nutritious obesity result.For example:Fig. 6 is final nutritious obesity result.
The process of above-mentioned steps 5 is described in detail below:
51, detection means setting luminance threshold T, the result to luminance contrast enters row threshold division, obtains segmentation result,
The segmentation result is designated as Thn(x,y):
52, detection means carries out opening operation to segmentation result, eliminates isolated bright spot, removes part clutter, fills simultaneously
Hole, Th ' is designated as by the segmentation result after processingn(x,y).By Th 'nThe zone marker that pixel value is 1 in (x, y) is interested
Region (Region of interest, ROI), calculates and obtains current ROI centre coordinate, and current ROI center is sat
Labeled as (xn,yn)。
53, detection means setting movement threshold Mov, calculate the distance between ROI center of n-th frame and the (n+1)th frame:
If Dn<The ROI region of n frames and the (n+1)th frame is then associated by Mov, detection means;If Dn>Mov, then associate
Failure, ROI region of the detection means not to n frames and the (n+1)th frame is associated.
54, detection means extracts the ROI region that is mutually related, and the ROI region that is mutually related is cirrus testing result.
In summary, by adopting the above-described technical solution, the beneficial effects of the invention are as follows:
The cirrus detection in remote sensing images is carried out by the way of image procossing due to detection means, is reduced to sensor
Requirement, while position and the size of cirrus can be accurately detected;And the algorithm used is simple, and computational efficiency is high, can
To meet requirement of real-time.
Described above is the preferred embodiment of the present invention, it is noted that for those skilled in the art
For, on the premise of principle of the present invention is not departed from, some improvements and modifications can also be made, these improvements and modifications
It should be regarded as protection scope of the present invention.
Claims (4)
1. a kind of cloud layer image detecting method, it is characterised in that including:
Detection means receives the remote sensing images sequence f for including N frame remote sensing imagesn(x, y), wherein n=1 ..., N are frame number, and N is
Totalframes;
The detection means obtains energy Saliency maps from the energy feature of the remote sensing images sequential extraction procedures remote sensing images
The detection means calculates the luminance contrast of the energy Saliency maps, obtains luminance contrast image
The detection means extracts the textural characteristics of the luminance contrast figure, obtains textural characteristics figure
The detection means utilizes image motion information, and the textural characteristics figure is carried out into intra-frame trunk, extracts area-of-interest
ROI, obtains nutritious obesity result.
2. the method as described in claim 1, it is characterised in that the detection means utilizes image motion information, by the line
Manage characteristic pattern and carry out intra-frame trunk, extract region of interest ROI, obtaining nutritious obesity result is specially:
The detection means sets luminance threshold T, and the result to the luminance contrast enters row threshold division, obtains segmentation result
For Thn(x,y):
<mrow>
<msub>
<mi>Th</mi>
<mi>n</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfenced open = "{" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<mn>1</mn>
<mo>,</mo>
</mrow>
</mtd>
<mtd>
<mrow>
<mi>i</mi>
<mi>f</mi>
</mrow>
</mtd>
<mtd>
<mrow>
<msubsup>
<mi>f</mi>
<mi>n</mi>
<mn>3</mn>
</msubsup>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>&GreaterEqual;</mo>
<mi>T</mi>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<mn>0</mn>
<mo>,</mo>
</mrow>
</mtd>
<mtd>
<mrow>
<mi>i</mi>
<mi>f</mi>
</mrow>
</mtd>
<mtd>
<mrow>
<msubsup>
<mi>f</mi>
<mi>n</mi>
<mn>3</mn>
</msubsup>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo><</mo>
<mi>T</mi>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
<mo>,</mo>
<mi>n</mi>
<mo>=</mo>
<mn>1</mn>
<mo>,</mo>
<mo>...</mo>
<mo>,</mo>
<mi>N</mi>
<mo>;</mo>
</mrow>
The detection means is to the segmentation result Thn(x, y) carries out opening operation, eliminates isolated bright spot, removes part clutter,
Hole, the segmentation result Th ' after being handled are filled simultaneouslyn(x,y);
The detection means is by the Th 'nThe zone marker that pixel value is 1 in (x, y) is region of interest ROI, obtains current
ROI centre coordinate (xn,yn);
The detection means sets movement threshold Mov, calculates the distance between the ROI center of n-th frame and the (n+1)th frame:
<mrow>
<msub>
<mi>D</mi>
<mi>n</mi>
</msub>
<mo>=</mo>
<msqrt>
<mrow>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>x</mi>
<mi>n</mi>
</msub>
<mo>-</mo>
<msub>
<mi>x</mi>
<mrow>
<mi>n</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
<mo>+</mo>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>y</mi>
<mi>n</mi>
</msub>
<mo>-</mo>
<msub>
<mi>y</mi>
<mrow>
<mi>n</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
</mrow>
</msqrt>
<mo>,</mo>
<mi>n</mi>
<mo>=</mo>
<mn>1</mn>
<mo>,</mo>
<mo>...</mo>
<mo>,</mo>
<mi>N</mi>
<mo>-</mo>
<mn>1</mn>
<mo>;</mo>
</mrow>
If Dn<The ROI region of n frames and the (n+1)th frame is then associated by Mov, the detection means;
The detection means extracts the ROI region that is mutually related, and the ROI region that is mutually related is nutritious obesity result.
3. the method as described in claim 1, it is characterised in that the detection means is from the remote sensing images sequential extraction procedures remote sensing
The energy feature of image, obtains energy Saliency mapsSpecially:
The detection means carries out Fourier transformation to the remote sensing images sequence:
sn(ωx,ωy)=F [fn(x, y)], n=1 ..., N, wherein, F represents Fourier transformation operator, (ωx,ωy) represent to become
Change to the coordinate of frequency domain;
The detection means calculates the amplitude of Fourier transformation, and takes the logarithm and obtain logarithmic spectrum:
Ln(ωx,ωy)=log [| sn(ωx,ωy) |], wherein | | amplitude operator is represented,
Calculate phase spectrum P (ωx,ωy):
WhereinRepresent phase operator;
The logarithmic spectrum and size are m × m mean filter mask convolution by the detection means, are smoothly composed:
V(ωx,ωy)=Ln(ωx,ωy)*hm(ωx,ωy), wherein mean filter template is:
The detection means will obtain spectrum residual error R (ω to the number spectrum and the smooth spectrum subtractionx,ωy):
R(ωx,ωy)=Ln(ωx,ωy)-V(ωx,ωy),
The detection means is by the spectrum residual error R (ωx,ωy) and the phase spectrum P (ωx,ωy) carry out two-dimensional discrete Fourier
Inverse transformation, obtains energy Saliency maps
<mrow>
<msubsup>
<mi>f</mi>
<mi>n</mi>
<mn>1</mn>
</msubsup>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mo>|</mo>
<msup>
<mi>F</mi>
<mrow>
<mo>-</mo>
<mn>1</mn>
</mrow>
</msup>
<mo>&lsqb;</mo>
<mi>exp</mi>
<mo>{</mo>
<mi>R</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>&omega;</mi>
<mi>x</mi>
</msub>
<mo>,</mo>
<msub>
<mi>&omega;</mi>
<mi>y</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>+</mo>
<mi>i</mi>
<mi>P</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>&omega;</mi>
<mi>x</mi>
</msub>
<mo>,</mo>
<msub>
<mi>&omega;</mi>
<mi>y</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>}</mo>
<mo>&rsqb;</mo>
<msup>
<mo>|</mo>
<mn>2</mn>
</msup>
<mo>.</mo>
</mrow>
4. the method as described in claim 1, it is characterised in that detection means described in the detection means extracts the brightness pair
Than the textural characteristics of degree figure, textural characteristics figure is obtainedSpecially:
The detection means builds wave filter:
Wherein, x '=a-m(xcosθ+
Ysin θ), y '=a-m(-xcosθ+ysinθ);a-mFor scale factor, θ represents the direction of kernel function, and λ represents the ripple of SIN function
Long, ψ represents phase offset, and σ represents the standard deviation of Gaussian function, the ratio of width to height of γ representative functions;
The wave filter and the remote sensing images convolution are obtained filter result by the detection means, and the filter result is line
Manage characteristic pattern:
<mrow>
<msubsup>
<mi>f</mi>
<mi>n</mi>
<mn>3</mn>
</msubsup>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<msubsup>
<mi>f</mi>
<mi>n</mi>
<mn>2</mn>
</msubsup>
<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>
<mi>n</mi>
<mo>=</mo>
<mn>1</mn>
<mo>,</mo>
<mo>...</mo>
<mo>,</mo>
<mi>N</mi>
<mo>.</mo>
</mrow>
2
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710345541.1A CN107273803B (en) | 2017-05-16 | 2017-05-16 | Cloud layer image detection method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710345541.1A CN107273803B (en) | 2017-05-16 | 2017-05-16 | Cloud layer image detection method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107273803A true CN107273803A (en) | 2017-10-20 |
CN107273803B CN107273803B (en) | 2020-04-24 |
Family
ID=60065156
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710345541.1A Active CN107273803B (en) | 2017-05-16 | 2017-05-16 | Cloud layer image detection method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107273803B (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108648184A (en) * | 2018-05-10 | 2018-10-12 | 电子科技大学 | A kind of detection method of remote sensing images high-altitude cirrus |
CN110667847A (en) * | 2019-10-17 | 2020-01-10 | 于刚 | Unmanned aerial vehicle intelligent flying height control platform |
CN111812106A (en) * | 2020-09-15 | 2020-10-23 | 沈阳风驰软件股份有限公司 | Method and system for detecting glue overflow of appearance surface of wireless earphone |
CN111967508A (en) * | 2020-07-31 | 2020-11-20 | 复旦大学 | Time series abnormal point detection method based on saliency map |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20120183225A1 (en) * | 2010-11-24 | 2012-07-19 | Indian Statistical Institute | Rough wavelet granular space and classification of multispectral remote sensing image |
CN103093241A (en) * | 2013-01-23 | 2013-05-08 | 北京理工大学 | Optical remote sensing image non-homogeneous cloud layer discriminating method based on homogenization processing |
CN105868745A (en) * | 2016-06-20 | 2016-08-17 | 重庆大学 | Weather identifying method based on dynamic scene perception |
-
2017
- 2017-05-16 CN CN201710345541.1A patent/CN107273803B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20120183225A1 (en) * | 2010-11-24 | 2012-07-19 | Indian Statistical Institute | Rough wavelet granular space and classification of multispectral remote sensing image |
CN103093241A (en) * | 2013-01-23 | 2013-05-08 | 北京理工大学 | Optical remote sensing image non-homogeneous cloud layer discriminating method based on homogenization processing |
CN105868745A (en) * | 2016-06-20 | 2016-08-17 | 重庆大学 | Weather identifying method based on dynamic scene perception |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108648184A (en) * | 2018-05-10 | 2018-10-12 | 电子科技大学 | A kind of detection method of remote sensing images high-altitude cirrus |
CN110667847A (en) * | 2019-10-17 | 2020-01-10 | 于刚 | Unmanned aerial vehicle intelligent flying height control platform |
CN110667847B (en) * | 2019-10-17 | 2020-08-18 | 安徽省徽腾智能交通科技有限公司泗县分公司 | Unmanned aerial vehicle intelligent flying height control platform |
CN111967508A (en) * | 2020-07-31 | 2020-11-20 | 复旦大学 | Time series abnormal point detection method based on saliency map |
CN111812106A (en) * | 2020-09-15 | 2020-10-23 | 沈阳风驰软件股份有限公司 | Method and system for detecting glue overflow of appearance surface of wireless earphone |
CN111812106B (en) * | 2020-09-15 | 2020-12-08 | 沈阳风驰软件股份有限公司 | Method and system for detecting glue overflow of appearance surface of wireless earphone |
Also Published As
Publication number | Publication date |
---|---|
CN107273803B (en) | 2020-04-24 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11244197B2 (en) | Fast and robust multimodal remote sensing image matching method and system | |
CN103729848B (en) | High-spectrum remote sensing small target detecting method based on spectrum saliency | |
CN103971115B (en) | Automatic extraction method for newly-increased construction land image spots based on NDVI and PanTex index | |
CN104376330B (en) | Polarimetric SAR Image Ship Target Detection method based on super-pixel scattering mechanism | |
CN104835175B (en) | Object detection method in a kind of nuclear environment of view-based access control model attention mechanism | |
CN107273803A (en) | Cloud layer image detecting method | |
CN109684925B (en) | Depth image-based human face living body detection method and device | |
CN103914847A (en) | SAR image registration method based on phase congruency and SIFT | |
CN105205858A (en) | Indoor scene three-dimensional reconstruction method based on single depth vision sensor | |
CN105139412A (en) | Hyperspectral image corner detection method and system | |
CN106296638A (en) | Significance information acquisition device and significance information acquisition method | |
CN105184779A (en) | Rapid-feature-pyramid-based multi-dimensioned tracking method of vehicle | |
CN102129573A (en) | SAR (Synthetic Aperture Radar) image segmentation method based on dictionary learning and sparse representation | |
CN103593832A (en) | Method for image mosaic based on feature detection operator of second order difference of Gaussian | |
CN103955701A (en) | Multi-level-combined multi-look synthetic aperture radar image target recognition method | |
CN102629380B (en) | Remote sensing image change detection method based on multi-group filtering and dimension reduction | |
Zhu et al. | Robust registration of aerial images and LiDAR data using spatial constraints and Gabor structural features | |
CN105488541A (en) | Natural feature point identification method based on machine learning in augmented reality system | |
CN110110618A (en) | A kind of SAR target detection method based on PCA and global contrast | |
CN110533025A (en) | The millimeter wave human body image detection method of network is extracted based on candidate region | |
CN103533332B (en) | A kind of 2D video turns the image processing method of 3D video | |
CN108510531A (en) | SAR image registration method based on PCNCC and neighborhood information | |
Zhang et al. | Optical and SAR image dense registration using a robust deep optical flow framework | |
Misra et al. | SPRINT: Spectra Preserving Radiance Image Fusion Technique using holistic deep edge spatial attention and Minnaert guided Bayesian probabilistic model | |
CN102136060A (en) | Method for detecting population density |
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 |