CN106408019A - Adaptive optical celestial target detection method on strong skylight background - Google Patents
Adaptive optical celestial target detection method on strong skylight background Download PDFInfo
- Publication number
- CN106408019A CN106408019A CN201610825346.4A CN201610825346A CN106408019A CN 106408019 A CN106408019 A CN 106408019A CN 201610825346 A CN201610825346 A CN 201610825346A CN 106408019 A CN106408019 A CN 106408019A
- Authority
- CN
- China
- Prior art keywords
- image
- signal
- sky brightness
- complete dictionary
- super complete
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/28—Determining representative reference patterns, e.g. by averaging or distorting; Generating dictionaries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2136—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on sparsity criteria, e.g. with an overcomplete basis
-
- 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
- G06V10/30—Noise filtering
-
- 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
- G06V10/50—Extraction of image or video features by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/07—Target detection
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Multimedia (AREA)
- Image Processing (AREA)
Abstract
The invention discloses an adaptive optical celestial target detection method on a strong skylight background. The method comprises the following steps: separately establishing a skylight background signal over-complete dictionary corresponding to the strong skylight background and a spot signal over-complete dictionary corresponding to a celestial target; separately carrying out sparse decomposition of a measured image in the established skylight background signal over-complete dictionary and the spot signal over-complete dictionary, separately calculating decomposition atoms of the measured image in the established skylight background signal over-complete dictionary and the spot signal over-complete dictionary, and obtaining a decomposition coefficient; reconstructing the image atoms, and solving residual errors of the image atoms; and calculating the mean square value of the residual errors, setting a threshold, and comparing the mean square value of the residual errors with the threshold to determine whether the image atom is a spot signal or a skylight background signal so that the celestial target can be extracted and the strong skylight background can be removed. The method can effectively extract wavefront signals reliably and remove skylight background noise.
Description
Technical field
The present invention relates to a kind of detection method of Celestial Objects, the adaptive optical under more particularly to a kind of strong sky brightness
Learn the detection method of Celestial Objects, belong to image analysis technology field.
Background technology
Adaptive optics (adaptive optics, AO) is the one-tenth that real-Time Compensation is caused by atmospheric turbulance or other factors
As during wavefront distortion prospect technology, but it detects and is affected by sky brightness and noise larger, can only operate in extremely weak
Under ambient lighting conditions, barycenter extraction is carried out to point target, complete the task of Celestial Objects observation;Under the conditions of strong sky brightness,
Because daylight light intensity is the tens of target light light intensity to arrive hundred times, existing Wavefront sensor can not work, adaptive optics system
System also cannot complete the wavefront distortion trimming process to Celestial Objects under the optical condition of the stronger sky brightness such as daytime.
For problem above, external Beckers and Gonglewski et al. did some explorations and applied AO on daytime
The work of system.Beckers et al. proposes the active wavefront sensing methods using ultra-narrow bandwidth filter technology, to launch laser
Back scattering or reflected light as beacon, by technology such as time gated and shutter controls to carry out Wavefront detecting, have obtained portion
The experimental result divided.Gonglewski et al. individually discusses the restricted problem to sky brightness for the field stop, but to solution
Certainly for Wavefront detecting problem under the conditions of strong sky brightness, effect is not apparent.China's Lee's supermacro et al. proposes by visual field
The method to detect echo signal distorted wavefront information in strong background for skew Hartmann wave front sensor (FSWFS), breaches certainly
Application under the conditions of very noisy for the adaptive optics system.
Although the proposition of the above method, improve the detectivity of ADAPTIVE OPTICS SYSTEMS to a great extent, expand
Open up systematic difference environment and working hour, but could obtain under the conditions of remaining a need for stronger signal to noise ratio in actual applications
Preferably effect, and it is larger and spend also higher to build such huge optical system difficulty.
Content of the invention
Present invention is primarily targeted at, overcome deficiency of the prior art, provide adaptive under a kind of strong sky brightness
Answer optical object's mesh object detection method, realize effectively and reliably extracting wavefront signals, remove sky brightness noise, expand adaptive
Answer the working hour of optical system.
In order to achieve the above object, the technical solution adopted in the present invention is:
A kind of detection method of the adaptive optics Celestial Objects under strong sky brightness, comprises the following steps:
1) utilize the morphological differences of strong sky brightness and Celestial Objects, set up the daylight corresponding with strong sky brightness respectively
The super complete dictionary of background signal and the super complete dictionary of the spot signal corresponding with Celestial Objects;
Wherein, the super complete dictionary of sky brightness signal is only capable of the sky brightness signal of the strong sky brightness of rarefaction representation, hot spot
The super complete dictionary of signal is only capable of the spot signal of rarefaction representation Celestial Objects;
2) obtain the measuring image of the adaptive optics Celestial Objects under strong sky brightness, by measuring image in the sky set up
Respectively Its Sparse Decomposition is carried out to measuring image in the super complete dictionary of light background signal and the super complete dictionary of spot signal, calculate respectively
Decomposition atom d in the super complete dictionary of sky brightness signal and the super complete dictionary of spot signal for the measuring imagei, and decomposed
Coefficient
3) reconstructed image atom Wherein, DgaussianFor Gauss dictionary,For decomposition coefficient;
4) solve the residual error of image atom
5) calculate the mean-square value of residual error, and given threshold T;The mean-square value of residual error and threshold value T are compared, to judge this
Image atom belongs to spot signal or sky brightness signal, to extract Celestial Objects, to remove strong sky brightness;
If the mean-square value of residual error is less than threshold value T, judge that this image atom belongs to spot signal;
If the mean-square value of residual error is more than threshold value T, judge that this image atom belongs to sky brightness signal.
The present invention is further arranged to:The super complete dictionary of described spot signal, its establishment step is, using dimensional Gaussian mould
Type builds the matrix D representing the super complete dictionary of spot signal, and sample image is expanded into the m that number of samples is n2The one of × 1
Dimensional vector, whole column vectors is configured to a matrix,M is
The size of dictionary atom, the number of n decomposition coefficient, wherein, each row siFor one of the super complete dictionary of spot signal atom.
The present invention is further arranged to:The super complete dictionary of described sky brightness signal, its establishment step is to choose different letters
Make an uproar ratio, varying strength sky brightness original image, choose four corner areas of any one two field picture from original image,
The gray value of four corner areas is averaged and obtains average sky brightness image subgraph, be chosen at average sky brightness figure
As the background atom in subgraph to build the super complete dictionary of sky brightness signal.
The present invention is further arranged to:Described step 1), also include complete dictionary super to sky brightness signal and hot spot letter
Number super complete dictionary is trained in sample image sequence respectively.
The present invention is further arranged to:Described step 2) in Its Sparse Decomposition is carried out to measuring image, specifically, will survey
Image is divided into N × N number of image subblock successively, respectively in the super complete dictionary of sky brightness signal and the super complete dictionary of spot signal
The middle decomposition atom d extracting each image subblocki, and obtain decomposition coefficient
The present invention is further arranged to:Described decomposition coefficientCalculation expression be,
Wherein, D0For the super complete dictionary of sky brightness signal or the super complete dictionary of spot signal, A is coefficient matrix, k0For being
Number threshold value.
Compared with prior art, the invention has the advantages that:
According to form fractal theory, set up respectively the sky brightness signal super complete dictionary corresponding with strong sky brightness and
The super complete dictionary of the spot signal corresponding with Celestial Objects, based on super complete dictionary, measuring image is carried out Its Sparse Decomposition, from
And effectively and reliably extract wavefront signals, remove sky brightness noise, thus expanding the working hour of ADAPTIVE OPTICS SYSTEMS.
The above is only the general introduction of technical solution of the present invention, in order to be better understood upon the technological means of the present invention, under
Face combines accompanying drawing, and the invention will be further described.
Brief description
Partial target atom in the super complete dictionary of spot signal that Fig. 1 is set up by the present invention;
Fig. 1-1 is some target atoms in Fig. 1;
Fig. 1-2 is the three-dimensional energy figure of this target atoms shown in Fig. 1-1;
The signal to noise ratio that Fig. 2 sets up selected by the super complete dictionary of sky brightness signal for the present invention is the original image of 12db;
The corresponding background atom that Fig. 2-1 this original image shown in Fig. 2 is generated;
Fig. 2-2 is the graphics of this background atom shown in Fig. 2-1;
The signal to noise ratio that Fig. 3 sets up selected by the super complete dictionary of sky brightness signal for the present invention is the original image of 25db;
The corresponding background atom that Fig. 3-1 this original image shown in Fig. 3 is generated;
Fig. 3-2 is the graphics of this background atom shown in Fig. 3-1;
Fig. 4 is the process experiment carrying out wavefront light spot image using detection method;
Fig. 5 is to select the process that Shack-Hartmann image is carried out to test using detection method;
Fig. 6 is the comparison diagram before and after single sub-aperture image procossing.
Specific embodiment
With reference to Figure of description, the present invention is further illustrated.
The present invention provides a kind of detection method of the adaptive optics Celestial Objects under strong sky brightness, walks including following
Suddenly:
1) utilize the morphological differences of strong sky brightness and Celestial Objects, set up the daylight corresponding with strong sky brightness respectively
The super complete dictionary of background signal and the super complete dictionary of the spot signal corresponding with Celestial Objects;And to the sky brightness set up
The super complete dictionary of signal and the super complete dictionary of spot signal are trained in sample image sequence respectively.
Wherein, the super complete dictionary of sky brightness signal is only capable of the sky brightness signal of the strong sky brightness of rarefaction representation, hot spot
The super complete dictionary of signal is only capable of the spot signal of rarefaction representation Celestial Objects;It should be noted that super complete dictionary is a kind of complete
New signal representation theory, it replaces complete basic function with super complete redundancy functions storehouse, and the element in dictionary is referred to as former
Son.
The super complete dictionary of described spot signal, its establishment step is to represent spot signal using dimensional Gaussian model construction
The matrix D of super complete dictionary, sample image is expanded into the m that number of samples is n2× 1 dimensional vector, by whole row
Vector is configured to a matrix,M is the size of dictionary atom, and n decomposes
The number of coefficient, wherein, each row siFor one of the super complete dictionary of spot signal atom.
Set up the super complete dictionary of spot signal according to simulated conditions and establishment step, be illustrated in figure 1 spot signal super complete
For the partial target atom in dictionary, simulated conditions are:Image size is 41pixel × 41pixel, spot center (x0,y0),
Coordinate is (20,20), peak value 15ADU, and equivalent Gaussian width is σA=1.25pixel;The target hot spot so simulating 5 ×
More than 80% energy has been concentrated it can be seen that itself and actual target are closer in 5 pixels.It is in Fig. 1 as Figure 1-1
Some target atoms, Fig. 1-2 be Fig. 1-1 shown in this target atoms three-dimensional energy figure.
The super complete dictionary of described sky brightness signal, its establishment step is to choose different signal to noise ratios, the daylight of varying strength
The original image of background, as shown in Figures 2 and 3, chooses four corner areas of any one two field picture, by four from original image
The gray value of individual corner areas is averaged and is obtained average sky brightness image subgraph, is chosen at average sky brightness image
Background atom in image building the super complete dictionary of sky brightness signal, as shown in Fig. 2-1 and Fig. 3-1.
As can be seen that sky brightness area grayscale has had from Fig. 2, Fig. 2-1, Fig. 2-2, and Fig. 3, Fig. 3-1, Fig. 3-2
Volt, but spot signal relatively is more steady, and the sky brightness of varying strength only has difference on average gray value, other form bases
This is identical.
2) obtain the measuring image of the adaptive optics Celestial Objects under strong sky brightness, measuring image is divided into successively
N × N number of image subblock, respectively to actual measurement in the super complete dictionary of sky brightness signal set up and the super complete dictionary of spot signal
Each image subblock of image carries out Its Sparse Decomposition, calculates each image subblock respectively in the super complete dictionary of sky brightness signal and hot spot
Decomposition atom d in the super complete dictionary of signali, and obtain decomposition coefficient
Described decomposition coefficientCalculation expression be,
Wherein, D0For the super complete dictionary of sky brightness signal or the super complete dictionary of spot signal, A is coefficient matrix, k0For being
Number threshold value.
3) reconstructed image atom Wherein, DgaussianFor Gauss dictionary,For decomposition coefficient;
4) solve the residual error of image atom
5) calculate the mean-square value of residual error, and given threshold T;The mean-square value of residual error and threshold value T are compared, to judge this
Image atom belongs to spot signal or sky brightness signal, to extract Celestial Objects, to remove strong sky brightness;
If the mean-square value of residual error is less than threshold value T, judge that this image atom belongs to spot signal;
If the mean-square value of residual error is more than threshold value T, judge that this image atom belongs to sky brightness signal.
The detection method of the adaptive optics Celestial Objects under the strong sky brightness being provided using the present invention, carries out wavefront light
The process experiment of spot image, as shown in figure 4, and obtain treatment effect figure.
Choose the wavefront light spot image containing different sky brightness levels from laboratory environment, wavefront light spot image is divided
Do not carry out Its Sparse Decomposition in the super complete dictionary of target and the super complete dictionary of background, as shown in Figure 4.
In Fig. 4 (a) figure be the original light spot image containing strong sky brightness it can be seen that sky brightness is stronger, hot spot believe
Number flood.
In Fig. 4 (b) figure be by subtract thresholding algorithm carry out processing obtain subtract threshold effect figure it can be seen that subtracting threshold value
Signal to noise ratio slightly improves afterwards, but because background and noise bounce are larger, still has larger background to remain, target not yet preferably carries
Take.
In Fig. 4, (c) figure and (d) figure are respectively under the super complete dictionary of spot signal and the super complete dictionary of sky brightness signal
Image is carried out with the result of Its Sparse Decomposition, there it can be seen that in the super complete dictionary of spot signal, target area energy gathers
Collection, forms obvious peak value;And contrast the Its Sparse Decomposition of the super complete dictionary of sky brightness signal it can be seen that in corresponding position
There are more obvious peak-to-valley value, the peak value contrast in complete dictionary super with target herein.
In Fig. 4, (e) figure is by detection method and to take the treatment effect figure obtained by 0.6 using given threshold T,
It can be seen that sky brightness is almost filtered completely, spot signal retains.
The detection method of the adaptive optics Celestial Objects under the strong sky brightness being provided using the present invention, selection Shack-
Hartmann's image carries out processing experiment, as shown in figure 5, and obtaining treatment effect figure.
In experiment, target sample size is 16 × 16, and it is 768 × 484pixel that threshold value T takes 0.6, CCD target surface, sub-aperture
Size is 20 × 20pixels, microlens array a size of 23 × 23, effective sub-aperture number 400, and each sub-aperture is one
The square hole of 0.51mm, in this experiment, each sub-aperture corresponds to 19 pixels of focal plane.
In Fig. 5, (a) figure is by being collected the Shack-Hartmann image containing strong sky brightness, it can be seen that
Signal to noise ratio is 18.9dB.
In Fig. 5, (b) figure is design sketch after over subtraction threshold process it can be seen that still having stronger after over subtraction threshold process
Noise residual, the signal to noise ratio after process is about 24.56dB, slightly improves.
In Fig. 5, (c) figure is after being processed using detection method proposed by the invention as a result, it is possible to find out a sub- aperture light
Spot extraction effect is preferable, and sky brightness noise filtering is more clean, and now signal to noise ratio is about 48.56dB.
Fig. 6 is the comparison diagram before and after single sub-aperture image procossing, and in Fig. 6, (a) figure is before processing single sub-aperture image,
In Fig. 6, (b) figure is to subtract single sub-aperture design sketch after threshold process, after in Fig. 6, (c) figure is using detection method process
Single sub-aperture design sketch.
(c) figure in Fig. 6 and (b) figure are contrasted, it can be seen that before processing, sky brightness is stronger, hot spot quilt
Background is flooded, and after subtracting thresholding algorithm process, noise bounce is still larger, after being processed using detection method, hot spot Objective extraction
Preferably.From fig. 6, it can be seen that when signal to noise ratio is relatively low, detection method relatively subtracts thresholding algorithm larger advantage.
In addition, under the conditions of to different signal to noise ratios, using detection method and the barycenter subtracting after thresholding algorithm process
Average deviation RMS value, PV value and SNR are compared, as shown in table 1.
Table 1
Can see from table 1, compared to subtracting thresholding algorithm, the barycenter after being processed using detection method is risen and fallen relatively
Little, the precision after explanation is processed is preferable;And, had using the SNR that detection method relatively subtracts image after thresholding algorithm is processed
Larger raising, especially when signal to noise ratio is relatively low, effect becomes apparent from.
General principle, principal character and the advantage of the present invention have been shown and described above.The technical staff of the industry should
Understand, the present invention is not restricted to the described embodiments, the simply explanation present invention's described in above-described embodiment and specification is former
Reason, without departing from the spirit and scope of the present invention, the present invention also has various changes and modifications, these changes and improvements
Both fall within scope of the claimed invention.Claimed scope is by appending claims and its equivalent circle.
Fixed.
Claims (6)
1. a kind of detection method of the adaptive optics Celestial Objects under strong sky brightness is it is characterised in that comprise the following steps:
1) utilize the morphological differences of strong sky brightness and Celestial Objects, set up the sky brightness corresponding with strong sky brightness respectively
The super complete dictionary of signal and the super complete dictionary of the spot signal corresponding with Celestial Objects;
Wherein, the super complete dictionary of sky brightness signal is only capable of the sky brightness signal of the strong sky brightness of rarefaction representation, spot signal
Super complete dictionary is only capable of the spot signal of rarefaction representation Celestial Objects;
2) obtain the measuring image of the adaptive optics Celestial Objects under strong sky brightness, by measuring image in the daylight back of the body set up
Respectively Its Sparse Decomposition is carried out to measuring image in the super complete dictionary of scape signal and the super complete dictionary of spot signal, calculate actual measurement respectively
Decomposition atom d in the super complete dictionary of sky brightness signal and the super complete dictionary of spot signal for the imagei, and obtain decomposition coefficient
3) reconstructed image atom Wherein, DgaussianFor Gauss dictionary,For decomposition coefficient;
4) solve the residual error of image atom
5) calculate the mean-square value of residual error, and given threshold T;The mean-square value of residual error and threshold value T are compared, to judge this image
Atom belongs to spot signal or sky brightness signal, to extract Celestial Objects, to remove strong sky brightness;
If the mean-square value of residual error is less than threshold value T, judge that this image atom belongs to spot signal;
If the mean-square value of residual error is more than threshold value T, judge that this image atom belongs to sky brightness signal.
2. the detection method of the adaptive optics Celestial Objects under strong sky brightness according to claim 1, its feature exists
In:The super complete dictionary of described spot signal, its establishment step is,
Represent the matrix D of the super complete dictionary of spot signal using dimensional Gaussian model construction, sample image is expanded into sample number
Mesh is the m of n2× 1 dimensional vector, whole column vectors is configured to a matrix,
1≤i≤n, m are the size of dictionary atom, the number of n decomposition coefficient, wherein, each row siFor in the super complete dictionary of spot signal
An atom.
3. the detection method of the adaptive optics Celestial Objects under strong sky brightness according to claim 1, its feature exists
In:The super complete dictionary of described sky brightness signal, its establishment step is to choose different signal to noise ratios, the sky brightness of varying strength
Original image, choose four corner areas of any one two field picture from original image, by the gray value of four corner areas
Average and obtain average sky brightness image subgraph, the background atom being chosen in average sky brightness image subgraph comes
Build the super complete dictionary of sky brightness signal.
4. the detection method of the adaptive optics Celestial Objects under strong sky brightness according to claim 1, its feature exists
In:Described step 1), also include complete dictionary super to sky brightness signal and the super complete dictionary of spot signal in sample image sequence
It is trained respectively in row.
5. the detection method of the adaptive optics Celestial Objects under strong sky brightness according to claim 1, its feature exists
In:Described step 2) in Its Sparse Decomposition is carried out to measuring image, specifically, measuring image is divided into successively N × N number of image
Sub-block, the decomposition extracting each image subblock respectively in the super complete dictionary of sky brightness signal and the super complete dictionary of spot signal is former
Sub- di, and obtain decomposition coefficient
6. the detection method of the adaptive optics Celestial Objects under strong sky brightness according to claim 1, its feature exists
In:Described decomposition coefficientCalculation expression be,
Wherein, D0For the super complete dictionary of sky brightness signal or the super complete dictionary of spot signal, A is coefficient matrix, k0For coefficient threshold
Value.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610825346.4A CN106408019A (en) | 2016-09-14 | 2016-09-14 | Adaptive optical celestial target detection method on strong skylight background |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610825346.4A CN106408019A (en) | 2016-09-14 | 2016-09-14 | Adaptive optical celestial target detection method on strong skylight background |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106408019A true CN106408019A (en) | 2017-02-15 |
Family
ID=57997122
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610825346.4A Pending CN106408019A (en) | 2016-09-14 | 2016-09-14 | Adaptive optical celestial target detection method on strong skylight background |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106408019A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106952233A (en) * | 2017-03-24 | 2017-07-14 | 深圳大学 | Fluorescence polymolecular localization method, device and super-resolution imaging method, system |
CN113295287A (en) * | 2021-05-26 | 2021-08-24 | 中国科学院光电技术研究所 | Hartmann subaperture threshold value reduction method for pupil dynamic intensity distribution |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104899567A (en) * | 2015-06-05 | 2015-09-09 | 重庆大学 | Small weak moving target tracking method based on sparse representation |
CN105118035A (en) * | 2015-08-20 | 2015-12-02 | 南京信息工程大学 | Self-adaptive optical spot signal extraction method based on sparse representation |
-
2016
- 2016-09-14 CN CN201610825346.4A patent/CN106408019A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104899567A (en) * | 2015-06-05 | 2015-09-09 | 重庆大学 | Small weak moving target tracking method based on sparse representation |
CN105118035A (en) * | 2015-08-20 | 2015-12-02 | 南京信息工程大学 | Self-adaptive optical spot signal extraction method based on sparse representation |
Non-Patent Citations (3)
Title |
---|
唐峥远: "基于图像稀疏表示的红外小目标检测与跟踪算法研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
李正周 等: "基于空时稀疏表示的红外小目标检测算法", 《兵工学报》 * |
韩金辉: "基于超完备图像稀疏表示的自适应红外小目标检测算法", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106952233A (en) * | 2017-03-24 | 2017-07-14 | 深圳大学 | Fluorescence polymolecular localization method, device and super-resolution imaging method, system |
CN106952233B (en) * | 2017-03-24 | 2020-03-31 | 深圳大学 | Fluorescent multi-molecule positioning method and device and super-resolution imaging method and system |
CN113295287A (en) * | 2021-05-26 | 2021-08-24 | 中国科学院光电技术研究所 | Hartmann subaperture threshold value reduction method for pupil dynamic intensity distribution |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111325748B (en) | Infrared thermal image nondestructive testing method based on convolutional neural network | |
CN107527352B (en) | Remote sensing ship target contour segmentation and detection method based on deep learning FCN network | |
CN103942803B (en) | SAR (Synthetic Aperture Radar) image based automatic water area detection method | |
CN110120020A (en) | A kind of SAR image denoising method based on multiple dimensioned empty residual error attention network | |
CN108510467A (en) | SAR image target recognition method based on variable depth shape convolutional neural networks | |
CN110097129A (en) | Remote sensing target detection method based on profile wave grouping feature pyramid convolution | |
CN104766334B (en) | Small IR targets detection tracking and its device | |
CN107480611A (en) | A kind of crack identification method based on deep learning convolutional neural networks | |
CN106156744A (en) | SAR target detection method based on CFAR detection with degree of depth study | |
CN108596880A (en) | Weld defect feature extraction based on image procossing and welding quality analysis method | |
Cui et al. | Improved swin transformer-based semantic segmentation of postearthquake dense buildings in urban areas using remote sensing images | |
CN102629378A (en) | Remote sensing image change detection method based on multi-feature fusion | |
CN105374026A (en) | A maritime infrared small target detection method suitable for coastal defense monitoring | |
CN110728668B (en) | Airspace high-pass filter for maintaining small target form | |
CN105405132A (en) | SAR image man-made target detection method based on visual contrast and information entropy | |
CN106156758A (en) | A kind of tidal saltmarsh method in SAR coast chart picture | |
CN106408019A (en) | Adaptive optical celestial target detection method on strong skylight background | |
Yao et al. | Application of improved MobileNet-SSD on underwater sea cucumber detection robot | |
CN105118035A (en) | Self-adaptive optical spot signal extraction method based on sparse representation | |
CN103942792B (en) | Impurity detecting method in medicine detection robot based on time domain features of sequence images | |
CN107358162B (en) | Polarization SAR image change detection method based on depth stack network | |
CN109919215A (en) | The object detection method of feature pyramid network is improved based on clustering algorithm | |
CN105551048A (en) | Space surface patch-based three-dimensional corner detection method | |
CN108491796A (en) | A kind of time domain period point target detecting method | |
CN104008365A (en) | Method for detecting sparse degree of fruit tree leaves based on image processing technology |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170215 |