CN106525027B - A kind of star sensor star point extracting method based on local binary patterns - Google Patents
A kind of star sensor star point extracting method based on local binary patterns Download PDFInfo
- Publication number
- CN106525027B CN106525027B CN201610952197.8A CN201610952197A CN106525027B CN 106525027 B CN106525027 B CN 106525027B CN 201610952197 A CN201610952197 A CN 201610952197A CN 106525027 B CN106525027 B CN 106525027B
- Authority
- CN
- China
- Prior art keywords
- pixel
- asterism
- star
- star chart
- local binary
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/02—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by astronomical means
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/02—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by astronomical means
- G01C21/025—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by astronomical means with the use of startrackers
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/20—Instruments for performing navigational calculations
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- General Physics & Mathematics (AREA)
- Astronomy & Astrophysics (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
- Eye Examination Apparatus (AREA)
Abstract
The star sensor star point extracting method based on local binary patterns that the invention discloses a kind of, it includes following steps: mean filter S1, is carried out to star chart, after being filtered in star chart each pixel gray value, and obtain filtered star chart；S2, the feature vector that pixel in star chart after filtering is calculated using local binary patterns LBP operator；S3, asterism pixel samples and background sample of pixels in support vector machines training real scene shooting star chart are utilized；The feature vector for the image slices vegetarian refreshments that S4, matching asterism training result and LBP operator calculate is completed asterism and is extracted.Its advantage is that: this method has fully considered the characteristics of image of asterism, asterism information as much as possible can be retained while eliminating veiling glare background pixel, therefore solve the problems, such as under light disturbance the lower pixel of gray value in veiling glare high-frequency noise point and asterism easy to be lost that traditional asterism extraction algorithm is easy to extract in star chart.
Description
Technical field
The present invention relates to star sensor technical fields, and in particular to a kind of star sensor star point based on local binary patterns
Extracting method.
Background technique
Star sensor is by observation to fixed star, the posture information for identifying, spacecraft being calculated, it is a kind of
Using fixed star as the high-precision aircraft attitude measurement sensor of observation benchmark.
Star sensor divides asterism pixel first from star chart, carries out utilizing asterism pixel extraction asterism later, further according to
The asterism of extraction carries out importance in star map recognition and Attitude Calculation.The precision that asterism extracts has been largely affected by star sensor work
Performance.
In existing star sensor star point extracting method, one is global threshold method is used, then according to global threshold
Asterism pixel is extracted, specifically, firstly, calculating the gray average and entire image of entire image in entire image
Gray standard deviation is used as global threshold plus three times times gray standard deviation using gray average, and grey scale pixel value is greater than threshold in window
The pixel of value is asterism pixel, conversely, the pixel is background pixel.There are also one is using based on Digital High Pass Filter method
Extract asterism, specifically, firstly, using each pixel of high-pass filtering formwork calculation threshold value, then, by the pixel threshold
For value compared with the gray value of pixel, pixel gray value is greater than threshold value, then the pixel is asterism pixel, conversely, being background
Pixel.
However, above-mentioned common two kinds of asterism extracting methods are all easy to be influenced by uneven illumination, asterism is being extracted
Veiling glare background pixel is extracted simultaneously.
Summary of the invention
The star sensor star point extracting method based on local binary patterns that the purpose of the present invention is to provide a kind of, by right
Star chart under light disturbance carries out mean filter processing, obtains the gray average of star chart pixel, and then calculate pixel again
LBP operator feature vector completes SVM training with sample asterism, finally utilizes SVM from star chart by asterism pixel and background pixel
It distinguishes, to complete the extraction algorithm of asterism, solves traditional asterism extraction algorithm under light disturbance and be easy to extract star chart
In veiling glare high-frequency noise point and asterism easy to be lost in gray value lower pixel the problem of.
In order to achieve the above object, the invention is realized by the following technical scheme:
A kind of star sensor star point extracting method based on local binary patterns, characterized in that comprise the steps of:
S1, mean filter is carried out to star chart, after being filtered in star chart each pixel gray value, and obtain filtered
Star chart；
S2, the feature vector that pixel in star chart after filtering is calculated using local binary patterns LBP operator；
S3, using support vector machines training real scene shooting star chart in the feature of asterism pixel samples and background sample of pixels to
Amount；
The feature vector for the image slices vegetarian refreshments that S4, matching asterism training result and LBP operator calculate is completed asterism and is extracted.
The above-mentioned star sensor star point extracting method based on local binary patterns, wherein the step S1 is specifically wrapped
Contain:
Pass through calculation formulaObtain the gray value of multiple pixels, in formula, I (i, j)
For the gray value of pixel (i, j), m (i, j) is the gray average of (i, j) of pixel.
The above-mentioned star sensor star point extracting method based on local binary patterns, wherein the step S2 is specifically wrapped
Contain:
According to the gray value of each pixel obtained in step S1, the feature at asterism pixel is calculated using LBP operator
Vector, comprising:
In formula, gcFor center pixel value, gPFor neighborhood territory pixel value, P is neighborhood point number, and R is neighborhood point radius.
The above-mentioned star sensor star point extracting method based on local binary patterns, wherein the step S4 is specifically wrapped
Contain:
It the feature vector of star chart pixel and will be calculated after S41, the filtering obtained using LBP operator in step S2
Input condition of the feature vector as support vector machines；
Star after S42, support vector machines will be filtered according to the feature vector for the asterism sampled pixel that training obtains in step S3
Pixel is divided into asterism pixel and background pixel in figure, to complete asterism extraction.
Compared with the prior art, the present invention has the following advantages: it has fully considered the characteristics of image of asterism, can eliminate
Retain asterism information as much as possible while veiling glare background pixel, therefore solves traditional asterism extraction algorithm under light disturbance
The problem of being easy to extract the lower pixel of gray value in veiling glare high-frequency noise point and the asterism easy to be lost in star chart.
Detailed description of the invention
Fig. 1 is flow chart of the method for the present invention；
Fig. 2 is LBP circle shaped neighborhood region schematic diagram in the embodiment of the present invention；
Fig. 3 is star chart used in the embodiment of the present invention；
Fig. 4 is the experimental result for carrying out asterism extraction to the star chart of Fig. 3 using the global threshold algorithm of the prior art；
Fig. 5 is the experimental result for carrying out asterism extraction to the star chart of Fig. 3 using method of the invention.
Specific embodiment
The present invention is further elaborated by the way that a preferable specific embodiment is described in detail below in conjunction with attached drawing.
As shown in Figure 1, the invention discloses a kind of star sensor star point extracting method based on local binary patterns, packet
Containing following steps:
S1, mean filter is carried out to star chart, after being filtered in star chart each pixel gray value, and obtain filtered
Star chart；This step is background pixel in order to prevent, and noise pixel interferes asterism pixel extraction, in the present embodiment, first using equal
Value filtering template, Filtering Template are typically sized to 3 × 3；
S2, the feature vector that pixel in star chart after filtering is calculated using local binary patterns LBP operator；
S3, using support vector machines training real scene shooting star chart in the feature of asterism pixel samples and background sample of pixels to
Amount；
The feature vector for the image slices vegetarian refreshments that S4, matching asterism training result and LBP operator calculate is completed asterism and is extracted.
The step S1 specifically includes to carry out mean filter to star chart using following formula, the gray scale for obtaining star chart midpoint is equal
Value:
In above formula, m (i, j) indicates that the gray average of pixel (i, j) after image filtering, I (i, j) indicate point (i, j)
Grey scale pixel value.
The step S2 specifically includes, according to the gray value of each pixel obtained in step S1, to utilize LBP operator
Calculate the feature vector at asterism pixel；
Usual LBP operator takes 3 × 3 neighborhoods, and traditional LBP operator has that sampling number is very few, this can reduce LBP
The robustness of operator.Therefore, the LBP operator of standard is extended to the calculation that any radius and neighborhood points can be obtained after circle shaped neighborhood region
Son increases number of sampling points, improves LBP operator robustness.It is adjacent that Fig. 2 gives under round field operator when (P, R) takes different value
Domain schematic diagram,
In formula, gcFor center pixel value；gPFor neighborhood territory pixel value；P is neighborhood point number；R is neighborhood point radius.In star chart,
Asterism size is generally 3 × 3 pixel sizes, in order to enable LBP operator accurately to calculate the feature vector of asterism pixel, this implementation
It is 1.5, P 12 that LBP operator radius R is chosen in example.According to above-mentioned LBP operator calculate the feature of pixel in star chart after filtering to
Amount.
The step S3 is specially that LBP operator is utilized to calculate asterism pixel samples and background sample of pixels in real scene shooting star chart
Feature vector, and specifically in the present embodiment, selected from real scene shooting star chart with the feature vector of SVM training sample asterism
200 asterism pixel samples and 200 background sample of pixels calculate asterism pixel samples with the LBP operator in S2 step first
With the feature vector of background sample of pixels, the feature of asterism pixel and background sample of pixels in real scene shooting star chart is then trained with SVM
Vector.
Using sample asterism training SVM, α value is solved, SVM training principle is as follows:
The optimization object function of SVM can be write as:
In above formula, label is tag variable, if data point is in positive direction position, label value is 1；Conversely,
Label value is -1.Angle brackets indicate x(i)And x(j)The inner product of two vectors；
Its constraint condition are as follows:
The He of C >=α >=0
Constant C, which is used to control, to be maximized interval and guarantees the function interval largely put less than 1.0.The groundwork of SVM
It is just to solve for α, solves α with SVM training sample asterism, then segmentation hyperplane can be expressed with these α.
The step S4 specifically includes:
S41, filtered using LBP operator in step S2 after star chart pixel feature vector and the spy that will be calculated
Levy input condition of the vector as support vector machines；
Star after S42, support vector machines will be filtered according to the feature vector for the asterism sampled pixel that training obtains in step S3
Pixel is divided into asterism pixel and background pixel in figure, to complete asterism extraction.
Asterism pixel is extracted from star chart using SVM, specifically, after the filtering calculated in conjunction with LBP operator in S2 in star chart
Trained support vector machines is completed in the feature vector and step S3 of pixel, asterism pixel is extracted from star chart.
Star sensor star point extracting method provided by the invention based on local binary patterns, by being filtered to star chart
Processing, obtains the gray average of pixel in star chart, so recycle LBP operator calculate the feature of asterism pixel in star chart to
Amount, and using SVM training asterism sample, asterism pixel is finally extracted from star chart using the SVM that training is completed, the present invention fills
Divide the characteristics of image that asterism is utilized, asterism information as much as possible can be retained while eliminating veiling glare background pixel, because
This solves under light disturbance traditional asterism extraction algorithm and is easy to extract veiling glare high-frequency noise point and star easy to be lost in star chart
In point the problem of gray value lower pixel.
When in order to verify star sensor star chart under present invention processing light disturbance, eliminates veiling glare background pixel and retain asterism
The ability of Pixel Information, here are to be with the real scene shooting star chart that moonlight enters star sensor visual field according to above-mentioned specific embodiment
Example, and the comparing result that the asterism extracting method actual experiment for combining marginal data of the invention obtains:
Fig. 3 is star sensor star chart under light disturbance used in the present embodiment, uses global threshold algorithm and office respectively
Portion's thresholding algorithm handles the figure, and experimental result is as shown in Figure 4,5, respectively the experimental result schematic diagram of veiling glare pixel, wherein Fig. 4
For the experimental result that Global Algorithm asterism extracts, Fig. 5 is the experimental result that inventive algorithm asterism extracts, and can be seen according to result
Out, the present invention can retain asterism information as much as possible while eliminating veiling glare background pixel, be conducive to improve subsequent step
The precision of middle center coordination.
It should be noted that those of ordinary skill in the art will appreciate that: realize above-mentioned each method embodiment whole or
This can be accomplished by hardware associated with program instructions for part steps.Program above-mentioned can store computer-readable deposits in one
In storage media.When being executed, execution includes the steps that above-mentioned each method embodiment to the program；And storage medium above-mentioned includes:
The various media that can store program code such as ROM, RAM, magnetic or disk.
It is discussed in detail although the contents of the present invention have passed through above preferred embodiment, but it should be appreciated that above-mentioned
Description is not considered as limitation of the present invention.After those skilled in the art have read above content, for of the invention
A variety of modifications and substitutions all will be apparent.Therefore, protection scope of the present invention should be limited to the appended claims.
Claims (4)
1. a kind of star sensor star point extracting method based on local binary patterns, which is characterized in that comprise the steps of:
S1, mean filter is carried out to star chart, after being filtered in star chart each pixel gray value, and obtain filtered star
Figure；
S2, the feature vector that pixel in star chart after filtering is calculated using local binary patterns LBP operator；
S3, the feature vector of asterism pixel samples and background sample of pixels in support vector machines training real scene shooting star chart is utilized；
The feature vector for the image slices vegetarian refreshments that S4, matching asterism training result and LBP operator calculate is completed asterism and is extracted.
2. the star sensor star point extracting method based on local binary patterns as described in claim 1, which is characterized in that described
Step S1 specifically include:
Pass through calculation formulaObtain the gray average of multiple pixels, in formula, I (i, j) is
The gray value of pixel (i, j), m (i, j) are the gray average of (i, j) of pixel.
3. the star sensor star point extracting method based on local binary patterns as described in claim 1, which is characterized in that described
Step S2 specifically include:
According to the gray value of each pixel obtained in step S1, the feature vector at asterism pixel is calculated using LBP operator,
Include:
In formula, gcFor center pixel value, gPFor neighborhood territory pixel value, P is neighborhood point number, and R is neighborhood point radius.
4. the star sensor star point extracting method based on local binary patterns as described in claim 1, which is characterized in that described
Step S4 specifically include:
The feature vector of star chart pixel and the feature that will be calculated after S41, the filtering obtained using LBP operator in step S2
Input condition of the vector as support vector machines；
After S42, support vector machines will be filtered according to the feature vector for the asterism sampled pixel that training obtains in step S3 in star chart
Pixel is divided into asterism pixel and background pixel, to complete asterism extraction.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610952197.8A CN106525027B (en) | 2016-11-02 | 2016-11-02 | A kind of star sensor star point extracting method based on local binary patterns |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610952197.8A CN106525027B (en) | 2016-11-02 | 2016-11-02 | A kind of star sensor star point extracting method based on local binary patterns |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106525027A CN106525027A (en) | 2017-03-22 |
| CN106525027B true CN106525027B (en) | 2019-04-09 |
Family
ID=58325294
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610952197.8A Active CN106525027B (en) | 2016-11-02 | 2016-11-02 | A kind of star sensor star point extracting method based on local binary patterns |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106525027B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107909073A (en) * | 2017-10-18 | 2018-04-13 | 天津大学 | Multidimensional local binary patterns and the hand-written music score spectral line delet method of machine learning |
| CN108805050B (en) * | 2018-05-28 | 2021-01-01 | 上海交通大学 | Electric wire detection method based on local binary pattern |
| CN113205063A (en) * | 2021-05-19 | 2021-08-03 | 云南电网有限责任公司电力科学研究院 | Visual identification and positioning method for defects of power transmission conductor |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103968845A (en) * | 2014-04-15 | 2014-08-06 | 北京控制工程研究所 | DSP and FPGA parallel multi-mode star image processing method for star sensor |
| WO2014186081A1 (en) * | 2013-05-14 | 2014-11-20 | The Charles Stark Draper Laboratory, Inc. | Star tracker with steerable field-of-view baffle coupled to wide field-of-view camera |
| WO2014209510A1 (en) * | 2013-06-28 | 2014-12-31 | The Charles Stark Draper Laboratory, Inc. | Chip-scale star tracker |
| CN104931044A (en) * | 2015-06-16 | 2015-09-23 | 上海新跃仪表厂 | Star sensor image processing method and system |
| CN105374035A (en) * | 2015-11-03 | 2016-03-02 | 北京航空航天大学 | Star sensor star point extraction method under stray light interference |
-
2016
- 2016-11-02 CN CN201610952197.8A patent/CN106525027B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2014186081A1 (en) * | 2013-05-14 | 2014-11-20 | The Charles Stark Draper Laboratory, Inc. | Star tracker with steerable field-of-view baffle coupled to wide field-of-view camera |
| WO2014209510A1 (en) * | 2013-06-28 | 2014-12-31 | The Charles Stark Draper Laboratory, Inc. | Chip-scale star tracker |
| CN103968845A (en) * | 2014-04-15 | 2014-08-06 | 北京控制工程研究所 | DSP and FPGA parallel multi-mode star image processing method for star sensor |
| CN104931044A (en) * | 2015-06-16 | 2015-09-23 | 上海新跃仪表厂 | Star sensor image processing method and system |
| CN105374035A (en) * | 2015-11-03 | 2016-03-02 | 北京航空航天大学 | Star sensor star point extraction method under stray light interference |
Non-Patent Citations (1)
| Title |
|---|
| 基于支持向量机的导航星选取算法研究;郑胜等;《计算机应用研究》;20050110(第1期);全文 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106525027A (en) | 2017-03-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108765325B (en) | Small unmanned aerial vehicle blurred image restoration method | |
| CN109146963B (en) | Image position offset detection method based on rapid feature matching | |
| CN104268519B (en) | Image recognition terminal and its recognition methods based on pattern match | |
| CN105931255A (en) | Method for locating target in image based on obviousness and deep convolutional neural network | |
| CN107368792A (en) | A kind of finger vein identification method and system based on wave filter and Bone Edge | |
| CN105512657B (en) | Character identifying method and equipment | |
| CN106525027B (en) | A kind of star sensor star point extracting method based on local binary patterns | |
| CN104156918B (en) | SAR image noise suppression method based on joint sparse representation and residual fusion | |
| CN107133929A (en) | Low quality file and picture binary coding method based on background estimating and energy minimization | |
| CN112950685B (en) | Infrared and visible light image registration method, system and storage medium | |
| CN104268853A (en) | Infrared image and visible image registering method | |
| CN105138983B (en) | The pedestrian detection method divided based on weighting block model and selective search | |
| CN102324099A (en) | Step edge detection method oriented to humanoid robot | |
| Zhao et al. | Automatic blur region segmentation approach using image matting | |
| US20190043168A1 (en) | An image processing device, an image processing method, and computer-readable recording medium | |
| CN109064402B (en) | Single image super-resolution reconstruction method based on enhanced non-local total variation model prior | |
| CN103632137A (en) | Human iris image segmentation method | |
| CN109410246B (en) | Visual tracking method and device based on correlation filtering | |
| CN109117746A (en) | Hand detection method and machine readable storage medium | |
| CN108564019A (en) | A kind of face identification method and system | |
| CN110276788B (en) | Method and apparatus for infrared imaging seeker target tracking | |
| CN106156691A (en) | The processing method of complex background image and device thereof | |
| CN112926516B (en) | Robust finger vein image region-of-interest extraction method | |
| CN112990368B (en) | Polygonal structure guided hyperspectral image single sample identification method and system | |
| CN116342653A (en) | Target tracking method, system, equipment and medium based on correlation filter |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | 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 |