CN102799665B - Unmanned aerial vehicle video data processing method - Google Patents
Unmanned aerial vehicle video data processing method Download PDFInfo
- Publication number
- CN102799665B CN102799665B CN201210243293.7A CN201210243293A CN102799665B CN 102799665 B CN102799665 B CN 102799665B CN 201210243293 A CN201210243293 A CN 201210243293A CN 102799665 B CN102799665 B CN 102799665B
- Authority
- CN
- China
- Prior art keywords
- data
- attitude
- filtering
- carried out
- uav
- 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
Abstract
The invention relates to an unmanned aerial vehicle video data processing method, in particular relates to a method for matching and fusing respective processing results after respective processing on unmanned aerial vehicle video data and remote measuring data. The method disclosed by the invention can solve the problem of synchronization between ground video data and the remote measuring data on an unmanned aerial vehicle, especially a microminiature unmanned aerial vehicle platform, and matching correlation processing is carried out for improving the synchronization accuracy and observation accuracy of the video data and the remote measuring data; and post processing error caused by non synchronization of the video data and the remote measuring data is effectively improved by adopting the method disclosed by the invention.
Description
Technical field
The present invention relates to a kind of UAV Video data processing method, may be used for unmanned plane particularly Small and micro-satellite platform, for improving unmanned aerial vehicle vision audio data and telemetry synchronization accuracy and data processing precision.
Background technology
At present, unmanned aerial vehicle platform is as the aerial platform of investigation, strike.Its data processing and relevant Decision are undertaken by ground handling operator.The mainly unmanned plane target of investication processed, the particular location struck target may be geographical position or the relative position relative to strike firepower; This just needs the positional information being obtained unmanned plane target of investication by data processing algorithm.And the effect determining investigation by the target positioning error that Processing Algorithm obtains, hit.The positioning calculation precision improving unmanned plane target of investication is the key of ground data process.
In prior art, two kinds of processing modes are adopted usually to UAV Video data processing:
1 non-data with same frame mode, that is to say that video data and telemetry adopt different transmissions to process to ground; This mode does not consider the stationary problem of data completely, and ground data process is also carry out synchronization process according to the acquisition moment of video data and telemetry; This processing mode can produce comparatively big error, because do not consider the factors such as video generation, compression, decompress(ion), link transmission time delay.
2 data with same frame modes, that is to say that sending to the moment of On-Board Processor to carry out packing at airborne end according to video data and telemetry processes; Larger improvement is had than non-its synchronism of same frame method.Ground data process carries out simultaneously match according to the video data in sync packet and telemetry; But the method does not consider the time required for video compression, does not make full use of video data information yet.
Summary of the invention
Technology of the present invention is dealt with problems and is: overcome the deficiencies in the prior art, provides a kind of UAV Video data processing method, makes the synchronism of unmanned aerial vehicle vision audio data and telemetry, precision all increases, improve the quality of data.
Technical solution of the present invention is:
A kind of UAV Video data processing method, the data that unmanned plane passes down comprise video data and telemetry, and telemetry comprises UAV Attitude data and load attitude data, and the step of described UAV Video data processing method is as follows:
(1) video data passed down unmanned plane carries out decompress(ion) and obtains image data, carries out feature point extraction to each frame image data;
(2) characteristic point that front and back frame image data extracts is mated;
(3) relative attitude of front and back frame image is resolved according to the characteristic point after coupling;
(4) UAV Attitude data in telemetry are passed down to unmanned plane and load attitude data carries out compound;
(5) filtering is carried out to the data after the compound obtained in step (4),
(6) data after filtering in the relative attitude obtained in step (3) and step (5) are carried out correlation registration;
(7) according to the result after correlation registration, by the relative attitude obtained in step (3), update process is carried out to the data after filtering in step (5).
The characteristic point extracted front and back frame image data in described step (2) is carried out coupling and is specially: by the expression same clarification of objective Point matching composition characteristic point pair in the frame image data of front and back, rejects the feature point pairs of matching error.
The feature point pairs of rejecting matching error can be realized by least squares estimate or stochastical sampling Uniform estimates method.
The relative attitude of front and back frame image is resolved by relative orientation algorithm realization according to the characteristic point after coupling in described step (3).
Pass down UAV Attitude data in telemetry and load attitude data to unmanned plane in described step (4) to carry out compound and be specially:
Adopt the mode of three dimensional space coordinate conversion, UAV Attitude data are expressed as spin matrix, load attitude data is also expressed as spin matrix, then the matrix of consequence that unmanned plane spin matrix premultiplication load spin matrix obtains is the compound result of UAV Attitude data and load attitude data.
In described step (6), the data after relative attitude and filtering are carried out correlation registration to be specially:
First the data after filtering in step (5) are carried out difference processing, according to correlation coefficient process, correlation registration is carried out to the relative attitude in the result after described difference processing and step (3) afterwards.
Described step (7) is carried out update process and is result by correlation registration in step (6), carries out adjustment processing according to relative attitude to the data after filtering in step (5).
The present invention's beneficial effect is compared with prior art:
The inventive method adopts the mode of floor treatment to process related data completely, to unmanned plane particularly Small and micro-satellite improve data processing precision have great improvement result.It carries out processing from image procossing and improves telemetry, improves the accuracy of observation of image and remote measurement synchronism and telemetry; Compare and synchronously process or adopt the simple synchronization processing mode quality of data to have obvious improvement with not adopting; Also characteristic and advantage of the present invention has been highlighted.
Accompanying drawing explanation
Fig. 1 is flow chart of data processing figure of the present invention;
Fig. 2 is adjustment processing schematic diagram.
Embodiment
The invention provides a kind of UAV Video data processing method, the data that unmanned plane passes down comprise video data and telemetry, telemetry comprises UAV Attitude data and load attitude data, and as shown in Figure 1, the step of UAV Video data processing method of the present invention is as follows:
(1) video data passed down unmanned plane carries out decompress(ion) and obtains image data, carries out feature point extraction to each frame image data; Digitized video adopts the depth of each pixel color on a series of numeral image.Light and shade on image, texture difference is all indicated by different digital values.And pixel value changes violent place shows as edge or angle point on image; These image features are determined by the geometry of target itself or color attribute.So same target is close closing on the performance in several video images, the geometric deformation of the target just caused because imaging angle is different is different.So edge or angle point describe the one of image, for describing the feature of image.
(2) characteristic point that front and back frame image data extracts is mated, be specially: by the expression same clarification of objective Point matching composition characteristic point pair in the frame image data of front and back, in the process that the process that composition characteristic point is right is feature point pair matching, compare between two according to characteristic point characteristic value, if the difference of characteristic value is less than given threshold value and can thinks characteristic point successful matching, otherwise it fails to match.In the process of coupling, there will be the coupling of mistake, the characteristic point rejecting matching error is elimination of rough difference process.
The feature point pairs of rejecting matching error can be realized by least squares estimate or stochastical sampling Uniform estimates method.
(3) relative attitude of front and back frame image is resolved according to the characteristic point after coupling, by relative orientation algorithm realization.Relative orientation algorithm be photogrammetric in rudimentary algorithm, its thought is: by the variable quantity of feature point pairs inverse video imaging moment attitude after coupling, that is to say relative attitudes vibration.Whole process is called relative orientation.
(4) UAV Attitude data in telemetry are passed down to unmanned plane and load attitude data carries out compound: the mode adopting three dimensional space coordinate conversion, UAV Attitude data are expressed as spin matrix, load attitude data is also expressed as spin matrix, then the matrix of consequence that unmanned plane spin matrix premultiplication load spin matrix obtains is the compound result of UAV Attitude data and load attitude data.Whole process that is to say the process of space coordinate conversion, and unmanned aerial vehicle platform attitude data and load attitude data are combined into an attitude data to imaging image, and the data in other words after compound are the attitude data in video imaging moment.
(5) filtering is carried out to the data after the compound obtained in step (4), adopt the random error filtering that Kalman filtering will occur.
(6) data after filtering in the relative attitude obtained in step (3) and step (5) are carried out correlation registration: first the data after filtering in step (5) are carried out difference processing, according to correlation coefficient process, correlation registration is carried out to the relative attitude in the result after described difference processing and step (3) afterwards.Correlation coefficient process is the measure of correlation or acquaintance property between judge two data, two data the most simply can be adopted to subtract each other, ask for subtract each other gained difference and, to difference with setting threshold value, can think that two data are relevant when being greater than given threshold value otherwise thinking that data are uncorrelated.
(7) according to the result after correlation registration, by the relative attitude obtained in step (3), update process is carried out to the data after filtering in step (5), be the result by correlation registration in step (6), according to relative attitude, adjustment processing carried out to the data after filtering in step (5).Adjustment processing is the process of systematic error equalization, in other words by process that the systematic error introduced in measuring process and accidental error are evenly shared by each measured value.This process is common in surveying, is improved the accuracy of observation of data entirety by adjustment processing.Concrete adjustment process is: complex data filter result is absolute data values, and in other words complex data illustrates concrete data value, as: complex data is 0.8 degree be attitude data is 0.8 degree.Relative orientation data are relative data value, that is to say the relative attitude value between image, are 0.2 degree if relative orientation is 0.2 degree of difference representing a rear frame image and former frame image.Adjustment result data are absolute data values, that is to say that attitude data is 0.8 degree as adjustment result is 0.9 degree.It is higher that relative orientation attitude data resolves acquisition precision by relative orientation; Complex data precision is lower, and average by the difference data between complex data and relative orientation data being carried out error, the complex data larger to error is modified, to obtain the difference between complex data and the height correlation between relative orientation data.Relative value so in other words between complex data and relative orientation data value have consistency, so exactly the error between complex data are averaged, improve the overall accuracy of observation of complex data.
As seen in Figure 2:
Complex data filter result is data sequence:
(1)0.5，0.6，0.8，1.1，0.8，0.6，0.8，1.1，1.0
Relative orientation attitude data sequence is:
(2)0.1，0.1，0.2，0.1，0.2，-0.2，-0.1，0.1，0.2
Adjustment process is:
Complex data carries out difference processing, and to obtain result as follows:
(3)0.1，0.2，0.3，-0.3，-0.2，0.2，0.3，-0.1
It is as follows that relative orientation attitude data and complex data difference result carry out correlation registration
(4)0.1，0.1，0.2，0.1，0.2，-0.2，-0.1，0.1，0.2
0.1，0.2，0.3，-0.3，-0.2，0.2，0.3，-0.1
As follows with relative orientation Data Update complex data filter result after correlation registration:
(5)0.5，0.6，0.8，1.1，0.8，0.6，0.8，1.1
0.1，0.2，0.3，-0.3，-0.2，0.2，0.3，-0.1
(6)0.5，0.6，0.7，0.9，1.0，0.7，0.5，0.7。
Claims (1)
1. a UAV Video data processing method, is characterized in that: the data that unmanned plane passes down comprise video data and telemetry, and telemetry comprises UAV Attitude data and load attitude data, and the step of described UAV Video data processing method is as follows:
(1) video data passed down unmanned plane carries out decompress(ion) and obtains image data, carries out feature point extraction to each frame image data;
(2) characteristic point that front and back frame image data extracts is mated;
(3) relative attitude of front and back frame image is resolved according to the characteristic point after coupling;
(4) UAV Attitude data in telemetry are passed down to unmanned plane and load attitude data carries out compound;
(5) filtering is carried out to the data after the compound obtained in step (4), adopt the random error filtering that Kalman filtering will occur;
(6) data after filtering in the relative attitude obtained in step (3) and step (5) are carried out correlation registration;
(7) according to the result after correlation registration, by the relative attitude obtained in step (3), update process is carried out to the data after filtering in step (5);
The characteristic point extracted front and back frame image data in described step (2) is carried out coupling and is specially: by the expression same clarification of objective Point matching composition characteristic point pair in the frame image data of front and back, rejects the feature point pairs of matching error;
The feature point pairs of rejecting matching error can be realized by least squares estimate or stochastical sampling Uniform estimates method;
The relative attitude of front and back frame image is resolved by relative orientation algorithm realization according to the characteristic point after coupling in described step (3);
Pass down UAV Attitude data in telemetry and load attitude data to unmanned plane in described step (4) to carry out compound and be specially:
Adopt the mode of three dimensional space coordinate conversion, UAV Attitude data are expressed as spin matrix, load attitude data is also expressed as spin matrix, then the matrix of consequence that unmanned plane spin matrix premultiplication load spin matrix obtains is the compound result of UAV Attitude data and load attitude data;
In described step (6), the data after relative attitude and filtering are carried out correlation registration to be specially:
First the data after filtering in step (5) are carried out difference processing, according to correlation coefficient process, correlation registration is carried out to the relative attitude in the result after described difference processing and step (3) afterwards;
Described step (7) is carried out update process and is result by correlation registration in step (6), carries out adjustment processing according to relative attitude to the data after filtering in step (5).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210243293.7A CN102799665B (en) | 2012-07-13 | 2012-07-13 | Unmanned aerial vehicle video data processing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210243293.7A CN102799665B (en) | 2012-07-13 | 2012-07-13 | Unmanned aerial vehicle video data processing method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102799665A CN102799665A (en) | 2012-11-28 |
| CN102799665B true CN102799665B (en) | 2015-03-25 |
Family
ID=47198775
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210243293.7A Active CN102799665B (en) | 2012-07-13 | 2012-07-13 | Unmanned aerial vehicle video data processing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102799665B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104899831B (en) * | 2015-06-09 | 2019-05-24 | 天津航天中为数据系统科技有限公司 | A kind of unmanned plane image data real-time processing method and system |
| CN106326920B (en) * | 2016-08-16 | 2020-02-07 | 天津航天中为数据系统科技有限公司 | Off-line synchronization method and device for telemetering data and video image data |
| CN106357994B (en) * | 2016-08-28 | 2019-09-24 | 国家海洋技术中心 | The synchronous method and device of telemetry and video image data |
| CN113364511B (en) * | 2021-05-11 | 2023-04-07 | 上海卫星工程研究所 | Equilibrium transfer computing method and system based on CCSDS packet telemetry |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1790052A (en) * | 2005-12-19 | 2006-06-21 | 武汉大学 | Area feature variation detection method based on remote sensing image and GIS data |
| CN101154313A (en) * | 2006-09-28 | 2008-04-02 | 长江航道规划设计研究院 | Three-dimensional simulation digital information navigation channel system and its implementing method |
| WO2011049834A2 (en) * | 2009-10-19 | 2011-04-28 | Intergraph Technologies Company | Data search, parser, and synchronization of video and telemetry data |
| CN102081159A (en) * | 2009-11-30 | 2011-06-01 | 北京天宇数字城市科技有限公司 | Special work station for receiving and processing aerial remote sensing underground data of portable unmanned aerial vehicle |
-
2012
- 2012-07-13 CN CN201210243293.7A patent/CN102799665B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1790052A (en) * | 2005-12-19 | 2006-06-21 | 武汉大学 | Area feature variation detection method based on remote sensing image and GIS data |
| CN101154313A (en) * | 2006-09-28 | 2008-04-02 | 长江航道规划设计研究院 | Three-dimensional simulation digital information navigation channel system and its implementing method |
| WO2011049834A2 (en) * | 2009-10-19 | 2011-04-28 | Intergraph Technologies Company | Data search, parser, and synchronization of video and telemetry data |
| CN102081159A (en) * | 2009-11-30 | 2011-06-01 | 北京天宇数字城市科技有限公司 | Special work station for receiving and processing aerial remote sensing underground data of portable unmanned aerial vehicle |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102799665A (en) | 2012-11-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109785379A (en) | The measurement method and measuring system of a kind of symmetric objects size and weight | |
| CN108230397B (en) | Multi-view camera calibration and correction method and apparatus, device, program and medium | |
| WO2019042232A1 (en) | Fast and robust multimodal remote sensing image matching method and system | |
| CN102799665B (en) | Unmanned aerial vehicle video data processing method | |
| CN106548462B (en) | Non-linear SAR image geometric correction method based on thin-plate spline interpolation | |
| CN105205858A (en) | Indoor scene three-dimensional reconstruction method based on single depth vision sensor | |
| CN108665499B (en) | Near distance airplane pose measuring method based on parallax method | |
| CN107560592B (en) | Precise distance measurement method for photoelectric tracker linkage target | |
| CN110456330B (en) | Method and system for automatically calibrating external parameter without target between camera and laser radar | |
| CN103278138A (en) | Method for measuring three-dimensional position and posture of thin component with complex structure | |
| CN106296698B (en) | A kind of lightning 3-D positioning method based on stereoscopic vision | |
| CN103971378A (en) | Three-dimensional reconstruction method of panoramic image in mixed vision system | |
| CN109724586B (en) | Spacecraft relative pose measurement method integrating depth map and point cloud | |
| CN104197933B (en) | High magnitude slides enhancing and the extracting method of fixed star in a kind of range of telescope | |
| IL221061A (en) | Method of modelling buildings on the basis of a georeferenced image | |
| CN104537646B (en) | The automatic MTF methods of estimation of multi-angle of remote sensing images | |
| Anderson et al. | RANSAC for motion-distorted 3D visual sensors | |
| CN111784778A (en) | Binocular camera external parameter calibration method and system based on linear solving and nonlinear optimization | |
| CN110969667A (en) | Multi-spectrum camera external parameter self-correction algorithm based on edge features | |
| CN104268853A (en) | Infrared image and visible image registering method | |
| Khoshelham et al. | Generation and weighting of 3D point correspondences for improved registration of RGB-D data | |
| CN104268880A (en) | Depth information obtaining method based on combination of features and region matching | |
| CN105551050A (en) | Optical field based image depth estimation method | |
| CN108230375A (en) | Visible images and SAR image registration method based on structural similarity fast robust | |
| CN111536981A (en) | Embedded binocular non-cooperative target relative pose measuring method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20170306 Address after: 065500 Guan Industrial Park, Langfang, Hebei Patentee after: Rainbow UAV Technology Co., Ltd. Address before: 100074 Beijing, Fengtai District Yungang West Road, No. 17 Patentee before: China Aerospace Aerodynamic Technology Institute |