CN104034353A - Computing method of digital sun sensor centroid based on detecting window - Google Patents
Computing method of digital sun sensor centroid based on detecting window Download PDFInfo
- Publication number
- CN104034353A CN104034353A CN201410250554.7A CN201410250554A CN104034353A CN 104034353 A CN104034353 A CN 104034353A CN 201410250554 A CN201410250554 A CN 201410250554A CN 104034353 A CN104034353 A CN 104034353A
- Authority
- CN
- China
- Prior art keywords
- detection window
- time
- pixel
- value
- window
- 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
- 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)
- Automation & Control Theory (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a computing method of a digital sun sensor centroid based on a detecting window, relates to the field of optical posture sensors, and solves the problems that the computing speed is slow, the data updating rate is slow, and the instantaneity is poor in the existing computing method. The computing method provided by the invention comprising the following steps of determining a self-adaption threshold value and a position of a first-time detecting window, optimizing the first-time detecting window, and determining the position of a second-time detecting window; and realizing the facula centroid computation on a picture element in the second-time detecting window. The computing method adopts a threshold value and the detecting window to match, the centroid computation on the picture element in the detecting window is realized, the computation of an automatic threshold value is simplified, the centroid computation precision is improved, and the method provides the basis for posture positioning of the sun sensor.
Description
Technical field
The present invention relates to optical attitude sensor field, be specifically related to a kind of digital sun sensor centroid computing method based on detection window.
Background technology
Digital sun sensor mainly, by detecting solar vector orientation, is determined the position of the sun in sensor body coordinate system, then by coordinates matrix, is converted and is obtained the position of the sun in satellite body coordinate system.Sun sensor attitude measurement principle as shown in Figure 1.Solar vector orientation is that the aperture height of the facula position on sensor and sun sensor by the sun is determined, the position of therefore determining solar facula is the basis of realizing digital sun sensor attitude location.
Conventional digital sun sensor centroid computing method, first to determine image threshold, make background separated with hot spot, to being greater than the picture point of threshold value, carry out centroid calculation, adopt high to the accuracy requirement of threshold value in this way, and in fact, when sun altitude is different, the hot spot brightness obtaining in image planes is not identical, is therefore difficult to determine accurately unified threshold value, has therefore produced adaptive threshold method.
Document: the design > > of the real-time hot spot centroid detection of the high speed of the < < such as Meng Fantao based on FPGA system, adopted a kind of adaptive threshold to determine method, image in one two field picture is averaged, obtain the gray level of bias light, the threshold value according to the average gray level of present frame as next frame image.The advantage of method is to realize fairly simplely, and real-time is higher, if but image planes are very large, and hot spot is when very little, and threshold value can be very little, according to current threshold value, directly to being greater than the pixel of threshold value, carrying out centroid calculation and can obviously reduce Centroid accuracy.
Document: in the digital sun sensor research > > of the < < such as Tu Binjie based on full shot, first determine an effective threshold interval and an effective threshold point number interval of crossing, program starts to travel through complete threshold interval from the minimum value of threshold interval, to meeting the threshold value of a threshold value number interval, averaged, and as final threshold value.The advantage of this method is that threshold calculations is accurate, while carrying out centroid calculation according to this threshold value, precision is high, but for realization, need to store entire image, then image be processed, this just need to increase exterior storage, volume, cost increase, computing velocity is slow, and data updating rate is slow, and real-time is poor.
Summary of the invention
There is the problems such as computing velocity is slow, data updating rate is slow, and real-time is poor for solving existing computing method in the present invention, a kind of digital sun sensor centroid computing method based on detection window is provided.
Digital sun sensor centroid computing method based on detection window, the method is realized by following steps:
Step 1, adaptive threshold are determined;
Image to imageing sensor output gathers, and in the time of collection, the pixel gray-scale value of image is greater than to 0 pixel and sums up and be averaging, and the result of acquisition is as the threshold value of next frame image;
Step 2, definite position of detection window for the first time;
Whole image planes are lined by line scan, and when continuous sweep m or pixel gray-scale value more than m are greater than threshold value, and the gray-scale value of next pixel is while being less than threshold value, records this pixel position Q (x
0, y
0), the Q point of take is benchmark, determines the position of detection window for the first time, the position of described detection window is for the first time:
(x
0-Lp≤x
0≤x
0+Rp,y
0-Up≤y
0≤y
0+Dp)；
In above formula: Lp is the distance of Q point train value distance measurement window leftmost side pixel train value;
Rp is the distance of Q point train value distance measurement window rightmost side pixel train value;
Up is the distance of the capable value of Q point capable value distance measurement window top side pixel;
Dp is the distance of the capable value of Q point capable value distance measurement window lower side pixel; The value of described m is less than the size in sun sensor aperture;
Step 3, detection window is for the first time optimized, determines the position of detection window for the second time;
Pixel in detection window is for the first time carried out to centroid calculation, obtain centroid position (x for the first time
f, y
f), with (x
f, y
f) centered by, determine the position of detection window for the second time, to realize the pixel in detection window is for the second time carried out to facula mass center calculating, the position of described detection window is for the second time designated as: (x
l≤ x≤x
r, y
u≤ y≤y
d);
Wherein: x
l=x
f-min (x
f-x
0+ L
p, x
0+ R
p-x
f); x
lcoordinate points as final detection window left side;
X
r=x
f+ min (x
f-x
0+ L
p, x
0+ R
p-x
f); x
rcoordinate points as final detection window right side;
Y
u=y
f-min (y
f-y
0+ U
p, y
0+ D
p-y
f); y
ucoordinate points as final detection window upside;
Y
d=y
f+ min (y
f-y
0+ U
p, y
0+ D
p-y
f); y
dcoordinate points as final detection window downside.
Beneficial effect of the present invention: the method that method of the present invention adopts threshold value and detection window to cooperatively interact, pixel in detection window is carried out to centroid calculation, simplified automatic threshold calculating, improved the precision of centroid calculation, for the attitude of sun sensor, located foundation is provided.
One, definite principle of detection window is first big after small, window is greater than spot size for the first time, dwindle for the second time window and really make hot spot be positioned at the center of window as far as possible, reduce the impact of noise on precision, guarantee that like this detection window size and location are accurate, can effectively reduce centroid position error;
Two, determine the first detection window size, with regard to not needing, two field picture has been stored, avoided large capacity storage, saved resource.
Three, due to threshold value is required to reduce, the computation process of automatic threshold is simplified, and then improves the turnover rate of data, and real-time improves.
Accompanying drawing explanation
Fig. 1 is digital sun sensor attitude measurement principle in the digital sun sensor centroid computing method based on detection window of the present invention;
Fig. 2 is that in the digital sun sensor centroid computing method based on detection window of the present invention, detection window is determined schematic diagram;
Fig. 3 is centroid calculation process flow diagram in the digital sun sensor centroid computing method based on detection window of the present invention.
Embodiment
Embodiment one, in conjunction with Fig. 1 to Fig. 3, present embodiment is described, digital sun sensor centroid computing method based on detection window, in conjunction with Fig. 1, the frame of reference that OXsYsZs is sun sensor, imaging aperture is positioned at O point coordinate, and Zs axle is perpendicular to the photosurface of detector.XC, YC are the barycentric coordinates of hot spot, and α, δ are position angle, the angle of pitch of the sun in the frame of reference.The method is realized by following steps:
Step 1: adaptive threshold is determined;
Image to imageing sensor output gathers, when gathering, pixel gray-scale value is judged, because digital sun sensor only has an aperture, it is light tight that all the other backgrounds keep as far as possible, so gray-scale value is greater than to 0 pixel, sums up and be averaging, as the threshold value of lower two field picture.
Step 2: determine detection window size for the first time;
Detection window size requirements comprises whole hot spot, if there is no diffraction diffusion, how much image patches equal hole dimension and add that aperture height H is multiplied by the sine of sun subtended angle, consider Diffraction Problems, and detection window size is greater than the actual hot spot of maximum after diffraction for the first time.
The definite of detection window is in order to guarantee that whole hot spot is included in detecting window for the first time.And can avoid large capacity storage after having determined size.
Step 3: determine the position of detection window for the first time;
Determine that detection window position detailed process is as shown in schematic diagram 2.Whole figure represents whole image planes, the upper left corner of image planes is defined as true origin, X, as shown in the figure, X represents the train value of picture point to Y-direction, and Y represents the row value of picture point, hot spot corresponding to diffraction sun sensor aperture do not considered in the rectangle frame ABCD representative of middle white, whole image planes are lined by line scan, when the continuous m of discovery is individual or m above pixel is all greater than threshold value, think that hot spot has occurred.M value is less than the size in aperture, and the object of m is to get rid of the impact of noise continuously.
Certain pixel of supposing to line by line scan, has occurred that continuous m or m above gray-scale value are greater than the pixel with threshold value, and next pixel degree value is less than threshold value, writes down pixel position Q (x now
0, y
0), the Q point of take is benchmark, determines that detection window is designated as FEGH, the position of detection window is:
(x
0-Lp≤x
0≤x
0+Rp,y
0-Up≤y
0≤y
0+Dp)
Wherein: Lp is the distance of Q point train value distance measurement window leftmost side pixel train value;
Rp is the distance of Q point train value distance measurement window rightmost side pixel train value;
Up is the distance of the capable value of Q point capable value distance measurement window top side pixel;
Dp is the distance of the capable value of Q point capable value distance measurement window lower side pixel.
In centroid calculation process, open up two fritter storage spaces, one is used for storing the position of the capable pixel of Lp and the space of gray-scale value, and this storage space is along with the carrying out of line scanning, constantly capped, until stop storage after finding Q point.The second block space is the pixel of all the other in detection window to position, carries out the storage of position and gray-scale value.
Step 4: detection window is for the first time optimized, determines the position of detection window for the second time;
Pixel in detection window is for the first time carried out to centroid calculation, obtain centroid position (x for the first time
f, y
f), with (x
f, y
f) centered by, determine the position of detection window for the second time, the position of detection window meets two conditions, the one, (x for the second time
f, y
f) be positioned at the center of detection window, the 2nd, window is included in for the first time in window and window size needs to maximize for the second time.Condition one is in order to make hot spot be positioned at detection window center as far as possible, reduces the impact of the result of calculation of noise, and condition two is in order to comprise whole hot spot.
Using detection window is as final detection window for the second time, position is designated as: (x
l≤ x≤x
r, y
u≤ y≤y
d).
Wherein: x
l=x
f-min (x
f-x
0+ L
p, x
0+ R
p-x
f); x
lcoordinate points as final detection window left side;
X
r=x
f+ min (x
f-x
0+ L
p, x
0+ R
p-x
f); x
rcoordinate points as final detection window right side;
Y
u=y
f-min (y
f-y
0+ U
p, y
0+ D
p-y
f); y
ucoordinate points as final detection window upside;
Y
d=y
f+ min (y
f-y
0+ U
p, y
0+ D
p-y
f); y
dcoordinate points as final detection window downside;
Step 5: facula mass center calculates;
Pixel in final detection window is carried out to centroid calculation.
Embodiment two, present embodiment are the embodiment of the digital sun sensor centroid computing method based on detection window described in embodiment one:
Step 1: automatic threshold is determined;
Gray-scale value is greater than to 0 pixel and sums up and be averaging, as the threshold value of lower two field picture:
Wherein, T is threshold value; p
igray-scale value for certain pixel; N is the pixel number that gray-scale value is greater than 0.
Step 2: determine detection window size for the first time;
Detection window size requirements comprises whole hot spot, supposes that sensor aperture is square, and the length of side is 10 pixel dimension sizes, and aperture height is 4mm, considers Diffraction Problems, and selecting detection window size is for the first time 20 pixel * 20 pixels;
Step 3: determine the position of detection window for the first time;
When the continuous m of discovery is individual or m above pixel is all greater than threshold value, think that hot spot has occurred.Certain pixel of supposing to line by line scan, has occurred that continuous m or m above gray-scale value are greater than the pixel with threshold value, and next pixel degree value is less than threshold value, writes down pixel position Q (x now
0, y
0), the Q point of take is benchmark, determines that detection window is designated as FEGH, the position of detection window is:
(x
0-Lp≤x
0≤x
0+Rp,y
0-Up≤y
0≤y
0+Dp)
Wherein: m=6;
Lp=14 is the distance of Q point train value distance measurement window leftmost side pixel train value;
Rp=5 is the distance of Q point train value distance measurement window rightmost side pixel train value;
Up=4 is the distance of the capable value of Q point capable value distance measurement window top side pixel;
Dp=15 is the distance of the capable value of Q point capable value distance measurement window lower side pixel.
Step 4: determine the position of detection window for the second time
Pixel in detection window is for the first time carried out to centroid calculation, obtain centroid position M (x for the first time
f, y
f), with (x
f, y
f) centered by, determine the position ABCD of detection window for the second time, position is designated as:
(x
l≤x
k≤x
r,y
u≤y
k≤y
d)。
Step 5: facula mass center calculates;
Pixel in final detection window is carried out to centroid calculation:
Wherein:
center-of-mass coordinate for hot spot;
(x
k, y
k) be the coordinate of K pixel in image planes;
P
kit is the gray-scale value of K pixel;
N is the number of pixel in final detection window.
Claims (2)
1. the digital sun sensor centroid computing method based on detection window, is characterized in that, the method is realized by following steps:
Step 1, adaptive threshold are determined;
Image to imageing sensor output gathers, and in the time of collection, the pixel gray-scale value of image is greater than to 0 pixel and sums up and be averaging, and the result of acquisition is as the threshold value of next frame image;
Step 2, definite position of detection window for the first time;
Whole image planes are lined by line scan, and when continuous sweep m or pixel gray-scale value more than m are greater than threshold value, and the gray-scale value of next pixel is while being less than threshold value, records this pixel position Q (x
0, y
0), the Q point of take is benchmark, determines the position of detection window for the first time, the position of described detection window is for the first time:
(x
0-Lp≤x
0≤x
0+Rp,y
0-Up≤y
0≤y
0+Dp)；
In above formula: Lp is the distance of Q point train value distance measurement window leftmost side pixel train value;
Rp is the distance of Q point train value distance measurement window rightmost side pixel train value;
Up is the distance of the capable value of Q point capable value distance measurement window top side pixel;
Dp is the distance of the capable value of Q point capable value distance measurement window lower side pixel; The value of described m is less than the size in sun sensor aperture;
Step 3, detection window is for the first time optimized, determines the position of detection window for the second time;
Pixel in detection window is for the first time carried out to centroid calculation, obtain centroid position (x for the first time
f, y
f), with (x
f, y
f) centered by, determine the position of detection window for the second time, to realize the pixel in detection window is for the second time carried out to facula mass center calculating, the position of described detection window is for the second time designated as: (x
l≤ x≤x
r, y
u≤ y≤y
d);
Wherein: x
l=x
f-min (x
f-x
0+ L
p, x
0+ R
p-x
f); x
lcoordinate points as final detection window left side;
X
r=x
f+ min (x
f-x
0+ L
p, x
0+ R
p-x
f); x
rcoordinate points as final detection window right side;
Y
u=y
f-min (y
f-y
0+ U
p, y
0+ D
p-y
f); y
ucoordinate points as final detection window upside;
Y
d=y
f+ min (y
f-y
0+ U
p, y
0+ D
p-y
f); y
dcoordinate points as final detection window downside.
2. the digital sun sensor centroid computing method based on detection window according to claim 1, it is characterized in that, step 2 comprises the size of determining the first detection window, the size of described detection window for the first time will comprise whole hot spot, if hot spot does not have diffraction diffusion, the size of how much image patches equal sun sensor aperture size and aperture height be multiplied by the sinusoidal of sun subtended angle with, if hot spot has diffraction diffusion, detection window size is greater than maximum actual hot spot after diffraction for the first time.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410250554.7A CN104034353A (en) | 2014-06-06 | 2014-06-06 | Computing method of digital sun sensor centroid based on detecting window |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410250554.7A CN104034353A (en) | 2014-06-06 | 2014-06-06 | Computing method of digital sun sensor centroid based on detecting window |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104034353A true CN104034353A (en) | 2014-09-10 |
Family
ID=51465217
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410250554.7A Pending CN104034353A (en) | 2014-06-06 | 2014-06-06 | Computing method of digital sun sensor centroid based on detecting window |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104034353A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105046724A (en) * | 2015-07-21 | 2015-11-11 | 长春理工大学 | Dual-image-point centroid extraction method in atmospheric coherent parameter test based on differential image motion |
| CN108168541A (en) * | 2017-12-20 | 2018-06-15 | 北京遥感设备研究所 | A kind of improved sub-pixed mapping asterism method for positioning mass center |
| US11579240B2 (en) | 2016-08-22 | 2023-02-14 | The University Of Sussex | Attitude determination system |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1907809A (en) * | 2006-08-01 | 2007-02-07 | 北京航空航天大学 | High precision APS sun sensor |
| CN1987346A (en) * | 2006-11-01 | 2007-06-27 | 北京航空航天大学 | Method and device for quick high precision positioning light spot image mass center |
| CN101363732A (en) * | 2008-09-17 | 2009-02-11 | 北京航空航天大学 | High frame frequency sun sensor and implementing method thereof |
-
2014
- 2014-06-06 CN CN201410250554.7A patent/CN104034353A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1907809A (en) * | 2006-08-01 | 2007-02-07 | 北京航空航天大学 | High precision APS sun sensor |
| CN1987346A (en) * | 2006-11-01 | 2007-06-27 | 北京航空航天大学 | Method and device for quick high precision positioning light spot image mass center |
| CN101363732A (en) * | 2008-09-17 | 2009-02-11 | 北京航空航天大学 | High frame frequency sun sensor and implementing method thereof |
Non-Patent Citations (5)
| Title |
|---|
| XIAOMING YIN,ET AL: "Adaptive thresholding and dynamic windowing method for automatic centroid detection of digital Shack–Hartmann wavefront sensor", 《OPTICAL SOCIETY OF AMERICA》 * |
| 夏爱利，马彩文: "提高波前探测精度的高阶矩方法", 《红外与激光工程》 * |
| 孟范涛等: "基于FPGA的高速实时光斑质心检测***的设计", 《仪器仪表用户》 * |
| 王薇，陈怀新: "基于优化探测窗口的光斑质心探测方法", 《强激光与粒子束》 * |
| 郑晓云等: "基于FPGA的数字太阳敏感器质心算法研究", 《中国空间科学学会2013年空间光学与机电技术研讨会》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105046724A (en) * | 2015-07-21 | 2015-11-11 | 长春理工大学 | Dual-image-point centroid extraction method in atmospheric coherent parameter test based on differential image motion |
| CN105046724B (en) * | 2015-07-21 | 2017-10-24 | 长春理工大学 | Double image point barycenter extracting method in Differential Image motion air relevant parameters test |
| US11579240B2 (en) | 2016-08-22 | 2023-02-14 | The University Of Sussex | Attitude determination system |
| CN108168541A (en) * | 2017-12-20 | 2018-06-15 | 北京遥感设备研究所 | A kind of improved sub-pixed mapping asterism method for positioning mass center |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2021233029A1 (en) | Simultaneous localization and mapping method, device, system and storage medium | |
| CN112132972B (en) | Three-dimensional reconstruction method and system for fusing laser and image data | |
| CN107063197B (en) | Reservoir characteristic curve extraction method based on spatial information technology | |
| CN103115614B (en) | Associated parallel matching method for multi-source multi-track long-strip satellite remote sensing images | |
| EP3132231B1 (en) | A method and system for estimating information related to a vehicle pitch and/or roll angle | |
| CN112396664A (en) | Monocular camera and three-dimensional laser radar combined calibration and online optimization method | |
| CN104392435A (en) | Fisheye camera calibration method and device | |
| CN103578088A (en) | Method for processing star image | |
| CN103295239A (en) | Laser-point cloud data automatic registration method based on plane base images | |
| CN109118577B (en) | Underwater laser scanning reconstruction system and method based on manned submersible | |
| CN114419130A (en) | Bulk cargo volume measurement method based on image characteristics and three-dimensional point cloud technology | |
| CN112487912B (en) | Arbitrary direction ship detection method based on improved YOLOv3 | |
| CN115880364A (en) | Robot pose estimation method based on laser point cloud and visual SLAM | |
| CN104034353A (en) | Computing method of digital sun sensor centroid based on detecting window | |
| CN113763549B (en) | Simultaneous positioning and mapping method and device integrating laser radar and IMU and storage medium | |
| CN103646389A (en) | SAR slant range image match automatic extraction method based on geometric model | |
| CN110221312A (en) | A method of quickly detecting ground point cloud based on laser radar | |
| CN114660602A (en) | Wide-area InSAR deformation rate adaptive splicing fusion method, device, equipment and medium | |
| CN104180794B (en) | The disposal route in digital orthoimage garland region | |
| CN103674904A (en) | Rapid atmospheric transmission correction method in infrared characteristic measurement | |
| CN110276043A (en) | A kind of regional aim access calculation method calculated based on boundary point access | |
| CN204924273U (en) | Three -dimensional laser sensor data calibrating device of panorama | |
| CN105352482A (en) | Bionic compound eye microlens technology-based 3-3-2 dimension object detection method and system | |
| CN116844124A (en) | Three-dimensional object detection frame labeling method, three-dimensional object detection frame labeling device, electronic equipment and storage medium | |
| CN116879886A (en) | Early warning method for graded communication of micro-change monitoring radar on landslide surface |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20140910 |
|
| WD01 | Invention patent application deemed withdrawn after publication |