CN112017156B - Space point target rotation period estimation method based on multispectral video - Google Patents
Space point target rotation period estimation method based on multispectral video Download PDFInfo
- Publication number
- CN112017156B CN112017156B CN202010693034.9A CN202010693034A CN112017156B CN 112017156 B CN112017156 B CN 112017156B CN 202010693034 A CN202010693034 A CN 202010693034A CN 112017156 B CN112017156 B CN 112017156B
- Authority
- CN
- China
- Prior art keywords
- spectral
- target
- frame
- spectrum
- pixel
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10032—Satellite or aerial image; Remote sensing
- G06T2207/10036—Multispectral image; Hyperspectral image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20048—Transform domain processing
- G06T2207/20056—Discrete and fast Fourier transform, [DFT, FFT]
Abstract
The invention discloses a space point target rotation period estimation method based on a multispectral video, and aims to solve the technical problem that the luminosity-based space point target rotation period estimation method in the prior art has insufficient precision under the condition of complex illumination. Firstly, acquiring a plurality of frames of spectral images, and then solving an average spectral curve of each frame of spectral image; then calculating the difference of the spectral angles; then acquiring a spectrum time-varying curve; selecting candidate points; acquiring a period corresponding to the candidate point; finally, carrying out period verification and calculation of a final rotation period; the method can solve the problem that the traditional luminosity estimation method has larger error in periodic estimation of the point target, and can better distinguish different postures of the space target by utilizing multispectral information.
Description
Technical Field
The invention relates to a space target, in particular to a space point target rotation period estimation method based on a multispectral video.
Background
The space target comprises space debris, satellites, space aircrafts and the like, and some satellites and aircrafts with abnormal or failed working states are in a free rotation state on the orbit after losing power, and the rotation period of the space target is estimated, so that the space target is beneficial to judging whether the target works normally or not. Since most of the targets, especially the high-orbit targets, have a wide distribution range, and are usually observed at a long distance (over 100 km) in order to ensure the observation efficiency and the safety of the targets, the shape information of the targets is unknown, and the targets are imaged as one point or a scattered spot in a spatial background.
Most of the existing space point target attitude state estimation methods are estimated based on photometric information, namely, the rotation period of the space point target attitude state estimation method is estimated through the periodic change of the reflected illumination intensity of the space point target attitude state estimation method. Since the luminosity of a plurality of surfaces of targets such as satellites and space debris may be similar, and the intensity of the luminosity is also easily influenced by environmental illumination factors, the luminosity-based space point target rotation period estimation method has the defect of insufficient accuracy.
Disclosure of Invention
The invention aims to solve the technical problem that the luminosity-based space point target rotation period estimation method in the prior art is insufficient in precision under the complex illumination condition, and the invention provides the target rotation period estimation method based on spectral information, which is higher in precision than the luminosity change estimation method by judging the change of the satellite attitude angle through spectral transformation in consideration that most space targets, including a satellite, a space vehicle and part of space fragments, are made of multiple materials and have different spectral reflectivities in different areas on the surfaces.
In order to achieve the purpose, the technical scheme adopted by the invention is as follows:
a space point target rotation period estimation method based on multispectral video is characterized by comprising the following steps:
1) Acquiring continuous multi-frame spectral images of a target area by using a video spectrometer, wherein each frame of spectral image comprises at least four spectra of different spectral bands;
2) Detecting pixel points occupied by a target in the first frame of spectral image by using a self-adaptive threshold method;
3) Solving the maximum pixel value of each pixel point in each spectral band to form a maximum pixel value image;
4) In the maximum pixel value image, taking m% of the maximum pixel value in all pixel values as a threshold, carrying out threshold segmentation on the maximum pixel value image, and extracting all pixel points with the pixel values larger than the threshold to form a target pixel point sequence, wherein m is more than or equal to 20 and less than or equal to 40;
5) Finding out a plurality of spectrums corresponding to any pixel point in a target pixel point sequence in an original first frame spectrum image, and solving the spectrum average value of the pixel point;
6) Repeating the step 5), and solving the average value of the spectrums of the other pixel points in the target pixel point sequence;
7) Fitting the average value of the spectrums of all the pixel points in the target pixel point sequence to obtain an average spectrum curve of the first frame of spectrum image, and recording the average spectrum curve as S 1 ；
8) Repeating the steps 2) to 7), respectively calculating the average spectrum curve of each of the rest frames of spectrum images, and recording the average spectrum curve of the t-th frame as S t ；
9) Calculating the spectral angle difference d between each frame of spectral image and the first frame of spectral image t ；
10 All spectral angle differences d) to be calculated t Connected into a one-dimensional vector D = { D = 1 ，d 2 ，d 3 …d N N is the total frame number and is marked as a target spectrum time-varying curve D;
11 Fast Fourier transform is performed on the spectrum time-varying curve D to obtain a frequency domain vector F, and the frequency domain vector F is larger than 0.5. F m As candidate points, F m Is the maximum value of F;
12 Calculate the period corresponding to the candidate points in the frequency domain vector F:
wherein n is the sequence number of each candidate point in the frequency domain vector F;
13 In a period of one period, the time-varying spectrum curve D is set according to the period L n Dividing the obtained product into K sections;
14 Respectively calculating the normalized average difference R of each of the signals from the 2 nd to the Kth section from the signal from the 1 st section n ；
15 Take all R n Period L corresponding to the maximum value of n The period of the signal to be determined is divided by the frame rate of the video spectrometer used to obtain the actual rotation period of the target.
Further, in the step 4), the value of m is 30.
Further, in step 9), the spectral angle difference d t The calculation formula of (2) is as follows:
further, in step 14), the normalized average difference R n The calculation of (c) is as follows:
further, in step 5), a weighted average method is adopted to obtain a spectrum average value.
The invention has the beneficial effects that:
1. the method can solve the problem that the traditional luminosity estimation method has larger error in periodic estimation of the point target, and can better distinguish different postures of the space target by utilizing multispectral information.
2. The method measures the difference of the postures of the point targets in different frames by calculating the spectrum angle of the video spectrum image sequence, and can be applied to video spectrometer data of various principles.
3. The method can accurately calculate the rotation period of the space target by using a mode of frequency domain estimation and time domain verification.
Drawings
FIG. 1 is a flow chart of a spatial point target rotation period estimation method based on multispectral video according to the present invention;
FIG. 2 is a graph of 5 spectral data generated by simulation at the same angle;
FIG. 3 is a three-spectral-segment synthesized pseudo-color image corresponding to a multi-spectral video frame;
FIG. 4 (a) is a calculated spectral time-varying graph for a sequence of multi-spectral video images;
FIG. 4 (b) is a graph of true rotation angle corresponding to the actual rotation period of FIG. 4 (a);
FIG. 5 is a frequency-amplitude diagram of a frequency domain transform;
fig. 6 is a cycle-amplitude diagram of a frequency domain transform.
Detailed Description
To make the objects, advantages and features of the present invention clearer, the following describes the multispectral video-based spatial point object rotation period estimation method in detail with reference to the accompanying drawings and specific embodiments. The advantages and features of the present invention will become more apparent from the following detailed description. It should be noted that: the drawings are in simplified form and are not to precise scale, the intention being solely for the convenience and clarity of illustrating embodiments of the invention; second, the structures shown in the drawings are often part of actual structures.
The invention discloses a space point target rotation period estimation method based on a multispectral video, which comprises the following steps of:
1. acquiring continuous multi-frame spectral images of a target area by using a video spectrometer, wherein each frame of spectral image comprises five spectra of different spectral bands as shown in fig. 2;
2. calculating an average spectrum curve of each frame of spectrum image;
2.1 Detecting pixel points occupied by the target in the first frame of spectral image by using an adaptive threshold method;
2.2 Solving the maximum pixel value of each pixel point in each spectral band to form a maximum pixel value image;
2.3 In the maximum pixel value image, taking 30% of the maximum pixel value in all pixel values as a threshold value, carrying out threshold segmentation on the maximum pixel value image, and extracting all pixel points with the pixel values larger than the threshold value to form a target pixel point sequence;
2.4 Finding a plurality of spectra corresponding to any pixel point in the target pixel point sequence in the original first frame spectral image, and calculating the spectral average value of the pixel point by using a weighted average method;
2.5 Repeating the step 2.4) to obtain the average value of the spectrums of the other pixel points in the target pixel point sequence;
2.6 ) fitting the average value of the spectra of all the pixel points in the target pixel point sequence to obtain the secondThe average spectral curve of a frame of spectral image is marked as S 1 ；
2.7 Repeating the steps 2.1) to 2.6), respectively calculating the average spectrum curve of each rest frame of spectrum image, and recording the average spectrum curve of the t-th frame as S t ；
3. Calculating the difference of the spectral angles;
calculating the spectral angle difference d between each frame of spectral image and the first frame of spectral image t ：
4. Acquiring a spectrum time-varying curve;
the spectral angle difference d between each frame of spectral image and the first frame of spectral image t Connected into a one-dimensional vector D = { D = 1 ，d 2 ，d 3 …d N N is the total frame number and is marked as a target spectrum time-varying curve D;
5. selecting candidate points;
performing fast Fourier transform on the spectrum time-varying curve D to obtain a frequency domain vector F, and as shown in FIG. 5, dividing the frequency domain vector F by more than 0.5. F m As shown in fig. 6, the circled points in fig. 6 are candidate points in which the final real period will be generated, F m Is the maximum value of F;
6. acquiring a period corresponding to the candidate point;
calculating the period corresponding to the candidate points in the frequency domain vector F:
wherein n is the sequence number of each candidate point in the frequency domain vector F;
7. period verification and calculation of a final rotation period;
7.1 In a period of one period, the time-varying spectrum curve D is set according to the period L n Dividing the obtained product into K sections;
7.2 Respectively meterCalculating the normalized average difference R between each of the signals from the 2 nd segment to the K-th segment and the signal from the 1 st segment n ：
7.3 Take all R n Period L corresponding to the maximum value of n The period of the signal to be obtained is divided by the frame frequency of the video spectrometer used to obtain the actual rotation period of the target.
The core technology of the invention is that the difference of the average spectrum curve of a space point target in a multispectral video is utilized to establish a one-dimensional signal reflecting the target spin period, and the target rotation period is finally determined by the methods of frequency domain peak value selection and time domain verification. The invention relates to a method for calculating a point target spectrum difference curve and calculating a target rotation period through time-frequency combination. Compared with the traditional estimation method based on the luminosity difference, the method has better accuracy, can be applied to judging a space failure target through a spinning cycle, and has application value in the aspects of space service, on-orbit maintenance of a spacecraft and the like.
As shown in fig. 3, there are 8 small images in the figure, each of which is a pseudo color synthesized by three spectral bands of a frame of spectral data, and it can be seen from the figure that there is a difference in corresponding multispectral images under different attitudes of the satellite, and this difference in periodic variation is the basis for performing the estimation of the rotation period.
As shown in fig. 4, fig. 4 (a) is a spectral time-varying curve D generated using a simulated sequence of multispectral video frames; fig. 4 (b) shows the true rotation angle of the satellite when used for simulation, and a comparison between fig. 4 (a) and fig. 4 (b) shows that the time-varying spectrum curve D has a significant periodicity, and the period of the time-varying spectrum curve D is substantially consistent with the true period, and the true period can be estimated from the time-varying spectrum curve D.
Claims (5)
1. A space point target rotation period estimation method based on multispectral video is characterized by comprising the following steps:
1) Acquiring continuous multi-frame spectral images of a target area by using a video spectrometer, wherein each frame of spectral image comprises at least four spectra of different spectral bands;
2) Detecting pixel points occupied by a target in the first frame of spectral image by using a self-adaptive threshold method;
3) Solving the maximum pixel value of each pixel point in each spectral band to form a maximum pixel value image;
4) In the maximum pixel value image, taking m% of the maximum pixel value in all pixel values as a threshold, carrying out threshold segmentation on the maximum pixel value image, and extracting all pixel points with the pixel values larger than the threshold to form a target pixel point sequence, wherein m is more than or equal to 20 and less than or equal to 40;
5) Finding out a plurality of spectrums corresponding to any pixel point in a target pixel point sequence in an original first frame spectrum image, and solving the spectrum average value of the pixel point;
6) Repeating the step 5), and solving the average value of the spectrums of the other pixel points in the target pixel point sequence;
7) Fitting the spectrum average values of all pixel points in the target pixel point sequence to obtain an average spectrum curve of the first frame of spectrum image, and recording the average spectrum curve as S 1 ；
8) Repeating the steps 2) to 7), respectively calculating the average spectrum curve of each of the rest frames of spectrum images, and recording the average spectrum curve of the t-th frame as S t ；
9) Calculating the spectral angle difference d between each frame of spectral image and the first frame of spectral image t ；
10 All spectral angular differences d) to be calculated t Connected into a one-dimensional vector D = { D = 1 ，d 2 ，d 3 …d N N is the total frame number and is marked as a target spectrum time-varying curve D;
11 Fast Fourier transform is performed on the spectrum time-varying curve D to obtain a frequency domain vector F, and the frequency domain vector F is larger than 0.5. F m As candidate points, F m Is the maximum value of F;
12 Calculate the period corresponding to the candidate point in the frequency domain vector F:
wherein n is the sequence number of each candidate point in the frequency domain vector F;
13 One period is taken as a section, and the spectral time-varying curve D is arranged according to the period L n Dividing the obtained product into K sections;
14 Respectively calculating the normalized average difference R of each of the signals from the 2 nd to the Kth section from the signal from the 1 st section n ；
15 Take all R n Period L corresponding to the maximum value of n The period of the signal to be determined is divided by the frame rate of the video spectrometer used to obtain the actual rotation period of the target.
2. The method according to claim 1, wherein the spatial point target rotation period estimation method based on multispectral video comprises: in the step 4), the value of m is 30.
3. The method according to claim 1, wherein the spatial point target rotation period estimation method based on multispectral video comprises: in step 9), the spectral angle difference d t The calculation formula of (c) is:
4. the method according to claim 1, wherein the multispectral video-based method for estimating the rotation period of the spatial point target comprises: in step 14), the normalized average difference R n The calculation method of (c) is as follows:
5. the method according to claim 1, wherein the multispectral video-based method for estimating the rotation period of the spatial point target comprises: in step 5), a weighted average method is adopted to calculate the spectrum average value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010693034.9A CN112017156B (en) | 2020-07-17 | 2020-07-17 | Space point target rotation period estimation method based on multispectral video |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010693034.9A CN112017156B (en) | 2020-07-17 | 2020-07-17 | Space point target rotation period estimation method based on multispectral video |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112017156A CN112017156A (en) | 2020-12-01 |
| CN112017156B true CN112017156B (en) | 2023-02-14 |
Family
ID=73500023
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010693034.9A Active CN112017156B (en) | 2020-07-17 | 2020-07-17 | Space point target rotation period estimation method based on multispectral video |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112017156B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103646244A (en) * | 2013-12-16 | 2014-03-19 | 北京天诚盛业科技有限公司 | Methods and devices for face characteristic extraction and authentication |
| CN103776540A (en) * | 2013-12-30 | 2014-05-07 | 华中科技大学 | Multiband common-optical-path spectrum combined remote sensing measurement system and method thereof |
| CN108627667A (en) * | 2018-05-15 | 2018-10-09 | 中国人民解放军战略支援部队航天工程大学 | Based on luminosity sequence while estimation space unstability target precession and spin rate method |
| US10361802B1 (en) * | 1999-02-01 | 2019-07-23 | Blanding Hovenweep, Llc | Adaptive pattern recognition based control system and method |
-
2020
- 2020-07-17 CN CN202010693034.9A patent/CN112017156B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10361802B1 (en) * | 1999-02-01 | 2019-07-23 | Blanding Hovenweep, Llc | Adaptive pattern recognition based control system and method |
| CN103646244A (en) * | 2013-12-16 | 2014-03-19 | 北京天诚盛业科技有限公司 | Methods and devices for face characteristic extraction and authentication |
| CN103776540A (en) * | 2013-12-30 | 2014-05-07 | 华中科技大学 | Multiband common-optical-path spectrum combined remote sensing measurement system and method thereof |
| CN108627667A (en) * | 2018-05-15 | 2018-10-09 | 中国人民解放军战略支援部队航天工程大学 | Based on luminosity sequence while estimation space unstability target precession and spin rate method |
Non-Patent Citations (2)
| Title |
|---|
| 《Imaging using limited-angle backscattered data from real targets》;D Pommet et al;《IEEE》;19970430;第39卷(第2期);全文 * |
| 《基于微多普勒效应激光探测的目标分类研究》;王云鹏;《中国优秀硕士学位论文全文数据库 信息科技辑》;20200215;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112017156A (en) | 2020-12-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109741356B (en) | Sub-pixel edge detection method and system | |
| JP6016552B2 (en) | Antireflection system | |
| CN107808138B (en) | Communication signal identification method based on FasterR-CNN | |
| CN113970743A (en) | Method and system for detecting objects near vehicle | |
| CN110598613B (en) | Expressway agglomerate fog monitoring method | |
| CN105139375A (en) | Satellite image cloud detection method combined with global DEM and stereo vision | |
| CN110488283B (en) | Error correction method for multi-channel HRWS-SAR channel | |
| CN106056625A (en) | Airborne infrared moving target detection method based on geographical homologous point registration | |
| US20140098098A1 (en) | Method for accelerating monte carlo renders | |
| AU2007287418A1 (en) | Target orientation | |
| CN114972083A (en) | Image restoration method based on measured data under complex optical imaging condition | |
| CN106895794B (en) | A kind of method and device obtaining laser beam scan path | |
| CA3038176C (en) | Object motion mapping from single-pass electro-optical satellite imaging sensors | |
| CN113446998B (en) | Hyperspectral target detection data-based dynamic unmixing method | |
| CN112017156B (en) | Space point target rotation period estimation method based on multispectral video | |
| CN105389582B (en) | ISAR contour extraction of objects method based on CLEAN algorithm scattering centers extraction | |
| CN110766657A (en) | Laser interference image quality evaluation method | |
| CN112784777B (en) | Unsupervised hyperspectral image change detection method based on countermeasure learning | |
| CN112836707B (en) | ISAR image aerial target length feature extraction method | |
| CN114549448B (en) | Complex multi-type defect detection evaluation method based on infrared thermal imaging data analysis | |
| CN111047525A (en) | Method for translating SAR remote sensing image into optical remote sensing image | |
| CN111199251A (en) | Multi-scale hyperspectral image classification method based on weighted neighborhood | |
| CN113093164B (en) | Translation-invariant and noise-robust radar image target identification method | |
| CN112686880B (en) | Method for detecting abnormity of railway locomotive component | |
| CN114596271A (en) | Method for extracting corrosion characteristics of high-light-reflection surface |
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 |