CN106771320A - A kind of rocket sledge image speed measurement method - Google Patents

A kind of rocket sledge image speed measurement method Download PDF

Info

Publication number
CN106771320A
CN106771320A CN201611050474.2A CN201611050474A CN106771320A CN 106771320 A CN106771320 A CN 106771320A CN 201611050474 A CN201611050474 A CN 201611050474A CN 106771320 A CN106771320 A CN 106771320A
Authority
CN
China
Prior art keywords
prime
photo
pixel
point
photos
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.)
Granted
Application number
CN201611050474.2A
Other languages
Chinese (zh)
Other versions
CN106771320B (en
Inventor
魏宗康
彭智宏
江麒
吕滕
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
China Aerospace Times Electronics Corp
Beijing Aerospace Control Instrument Institute
Original Assignee
China Aerospace Times Electronics Corp
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by China Aerospace Times Electronics Corp filed Critical China Aerospace Times Electronics Corp
Priority to CN201611050474.2A priority Critical patent/CN106771320B/en
Publication of CN106771320A publication Critical patent/CN106771320A/en
Application granted granted Critical
Publication of CN106771320B publication Critical patent/CN106771320B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01PMEASURING LINEAR OR ANGULAR SPEED, ACCELERATION, DECELERATION, OR SHOCK; INDICATING PRESENCE, ABSENCE, OR DIRECTION, OF MOVEMENT
    • G01P3/00Measuring linear or angular speed; Measuring differences of linear or angular speeds
    • G01P3/36Devices characterised by the use of optical means, e.g. using infrared, visible, or ultraviolet light
    • G01P3/38Devices characterised by the use of optical means, e.g. using infrared, visible, or ultraviolet light using photographic means

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Electromagnetism (AREA)
  • Power Engineering (AREA)
  • General Physics & Mathematics (AREA)
  • Image Processing (AREA)
  • Image Analysis (AREA)

Abstract

A kind of rocket sledge image speed measurement method, the running of skid body is continuously shot using high speed camera, calculated by the pitching to high speed camera relative orbit and azimuth information amendment, skid body feature point extraction, the speed of service after the completion of experiment, complete the measurement to the skid body speed of service.The method has that rate accuracy is high, thinking is simple and clear, and the advantage that programming is easily realized improves the tachometric survey precision of key position in Rocket sled test.

Description

A kind of rocket sledge image speed measurement method
Technical field
The present invention relates to a kind of rocket sledge image speed measurement method, more particularly to it is a kind of using high-speed camera be continuously shot come The method for measuring the skid body speed of service, belongs to rocket sledge testing field.
Background technology
Rocket sled test has the ability for producing the integrated conditions such as big overload, high speed, strong vibration and impact, can most force True simulation guided missile Live Flying environment.Items of the inertial measurement system under the conditions of integrated environment can be examined by experiment Can index and precision, correctness of the checking inertial measuring system error model under high dynamic condition, particularly in big overload feelings Under condition, high-order term amplification can determine influence of the inertial measurement system higher order error item to navigation performance, be to realize inertia The optimal path of measuring system dynamic property checking.
It is main at present to be filled using radar ranging devices, shadow shield time space measure in inertial measurement system Rocket sled test Put the run location and speed of measurement skid body.
Radar ranging devices are launched the electromagnetic wave of continuous constant amplitude to rocket sledge from radar antenna and are received from skid body reflection The electromagnetic wave signal returned, using the Doppler frequency of measurement, skid body skid body is obtained relative to the radial direction that antenna is pointed to through treatment Speed and skid body along track the speed of service.But, there are following two shortcomings in radar velocity measurement:
(1) it is suitable for measurement at a high speed, low speed certainty of measurement is low.It is less than in speed in the practical application of Rocket sled test Not outputting measurement value during 10m/s;
(2) the wake flame effect of rocket sledge solid engines makes the precision of radar surveying relatively low, in rocket sledge first class engine The data actual effect of radar velocity measurement can be caused when being lighted a fire with two stage motors.
Shadow shield time space measure device is carried photoelectric subassembly by sledge and is installed at interval of certain distance along rocket sledge slide rail side Some shadow shields composition.Photoelectric subassembly is arranged in skid body, is made up of laser and laser pickoff, laser pickoff connection There is process circuit, realize shadow shield to cause the change transitions of luminous flux to be electric impulse signal, and export to data collecting system. Data collecting system then records the instantaneous moment of the electric pulse, can obtain the fortune of skid body along the distance of track by adjacent shadow shield Scanning frequency degree.But, the shortcoming using shadow shield photoelectric subassembly is:
(1) light after the laser of the photoelectric subassembly is lighted a fire to solid engines is more sensitive, is easily caused process circuit In saturation state, so as to cause photoelectric subassembly operation irregularity;
(2) adjacent shadow shield spacing is generally higher than 1 meter, and sledge is asked for divided by time of measuring interval using adjacent shadow shield distance The method of the speed of running body haves the shortcomings that measurement error is relatively large.
As seen through the above analysis, 16km is extended to by 9km as rocket sledge track is follow-up, using multi-stage ignition Demand is more and more urgent, and radar velocity measurement and shadow shield time space measure have not adapted to the requirement of multi-stage ignition section.Accordingly, it would be desirable to grind Study carefully a kind of new speed-measuring method to make up the defect of above two method.
In Rocket sled test, photograph at a high speed existing at present, but it is served only for the instantaneous multistage skid body point of observation engine ignition From whether function monitoring purposes is normally waited, patent research of the present invention is a kind of to be realized testing the speed on the basis of existing high speed photograph New method.
The content of the invention
Present invention solves the technical problem that being:A kind of rocket sledge image speed measurement method is have found, the method can be accurate The speed of measurement skid body operation, what radar, shadow shield were influenceed and failed by engine ignition when solving multi-stage engine separation asks Topic, to provide high-precision velocity information.
The present invention solve technical scheme be:
A kind of rocket sledge image speed measurement method, step is as follows:
(1) in Rocket sled test, high speed camera is set up by the track at multi-stage engine separating ignition, camera is set Filming frequency f, original photo in longitudinal length of side number of pixels M0With the number of pixels N of the horizontal length of side0;Measurement camera is relative The distance between adjacent fastener L in the geography information and track of track;M0、N0Positive integer is, f, L are arithmetic number;
(2) during Rocket sled test multi-stage engine separating ignition, K width is continuously shot by high speed camera and is shone Piece, K is positive integer;It is made up of the mode of bidimensional pixel additional color in every photos, position of the pixel in photo Information is the coordinate value of original photo pixel coordinate system;
(3) in each photos, the coordinate points (m, n) in original photo pixel coordinate system are transformed into original photo Point (x, y) in center point coordinate system, its conversion formula is as follows:
After converted, with (x, y, rx,y,gx,y,bx,y) description original photo in the pixel position and colouring information;
The original photo center point coordinate system is with photo center as origin, ox axles level right-handed coordinate system oxy to the right;
(4) according in the prefiring first photos middle orbit distortion information amendment photo of multi-stage engine so that photo In space length corresponding to each pixel it is equal;
(5) according to distortionless photo after the amendment obtained in step (4), rocket sledge skid body in each photos of extraction Position (I of the fisrt feature point after the amendment in photograph pixel point coordinates system om ' n 's,Js) and the fisrt feature point in this position Pixel valueBehalf photo width number;
(6) the fisrt feature point horizontal direction pixel value J obtained according to step (5)s, and the parameter in step (1), ask Take the velocity series of rocket sledge skid body;
(7) according to distortionless photo after the amendment obtained in step (4), rocket prizes skid body in extracting each photos Further feature point, repeat step (5) and (6) obtain the velocity series of the corresponding rocket sledge skid body of each characteristic point, afterwards with The corresponding velocity series of one characteristic point are merged, so as to obtain the overall process velocity series of rocket sledge skid body operation.
, in the photo upper left corner, down, on axles are towards the right side for om axles for the origin of original photo pixel coordinate system omn;Pixel Color value is made up of red, green ratio different with blueness, with five dimension group (m, n, rm,n,gm,n,bm,n) represent, wherein, (m, N) position of the pixel in photo is described, the span of m is from 1 to M0Positive integer, the span of n is from 1 To N0Positive integer;(rm,n,gm,n,bm,n) describe the colouring information of the pixel in original photo, red rm,n, green gm,n、 Blue bm,nSpan from 0 to 255.
The step (4) according in the prefiring first photos middle orbit distortion information amendment photo of multi-stage engine, Specially:
(3.1) in the photo of amendment before processing, the coordinate value of each fastener is from left to right extracted successively and is numbered, H >=4 fastener is provided with, with coordinate value (xp,yp) represent, wherein p=1,2 ..., h;
(3.2) to the fastener sequence (x in photop,yp), p=1,2 ..., h;Using straight line model yl=klx+bl, with most Small square law design factor klAnd blValue, specific formula for calculation is as follows:
Wherein,
(3.3) to the fastener sequence (x in photop,yp), p=1,2 ..., h;Calculate two neighboring fastener apart from lqWith Center position coordinates (x0,q,y0,q), q=1,2 ..., h-1;Specific formula for calculation is
(3.4) the secondary distortion model of cut-off trajectoryUsing least square method design factor α and β, specific formula is as follows:
Wherein
(3.5) according to factor alpha, klAnd blValue, set M1And N1, and original photo is corrected, specific formula is as follows:
(rm',n',gm',n',bm',n')=(rm,n,gm,n,bm,n)
Wherein, array (rm,n,gm,n,bm,n) subscript m, n illustrate original photo pixel coordinate system coordinate points position Put;Array (rm',n',gm',n',bm',n') subscript m ', n ' illustrate amendment after photograph pixel point coordinates system coordinate points position, After correction photograph pixel point coordinates system om ' n ' be original photo correction after, with the photo upper left corner as origin, om ' axles vertically downward, On ' axles level coordinate system to the right;M1And N1The vertical and horizontal number of pixels of photo after respectively correcting, makes m ' and n ' values More than 0;(x ', y ') represents the coordinate points of optical center coordinate system ox ' y ' after correction, optical center coordinate system after correction After ox ' y ' are original photo correction, with photo center as origin, ox ' axles level right-handed coordinate system to the right;
(3.6) there is the point not being colored, the photograph pixel after amendment in revised photo, in the region being stretched The region that is stretched is repaired in point coordinates system, specially:
Wherein, (rm″,n″,gm″,n″,bm″,n″) represent amendment after there is no the colouring information of upper color dot in photo, m " and n " is represented The coordinate position of color dot is not gone up by photograph pixel point coordinates system after amendment, and λ is represented around the pixel in eight pixels, red The number of color, green, blue first pixel value sum more than 0.
The fisrt feature point of rocket sledge skid body photograph pixel point after amendment in each photos of extraction in the step (5) Position (I in coordinate system om ' n 's,Js), and the fisrt feature point is in the pixel value of this positionSpecifically For:
(4.1) since the first photos, Z photos are chosen at equal intervals, find these photos photograph pixel after amendment The skid body feature point coordinates (I of point coordinates system om ' n 's,Js);The pixel value of the fisrt feature point in this position of skid body is extracted simultaneouslyWhereinRepresent coordinate points (Is,Js) red pixel value,Represent coordinate points (Is,Js) green picture Element value,Represent coordinate points (Is,Js) blue pixel value, behalf picture width number;
The fisrt feature point of these photos at equal intervals is found, five dimension groups of these fisrt feature points are readScope according to more than where five dimension groups determines skid body fisrt feature point colorThe domain value range of red, green and blueness is respectively (a1,a4),(a2,a5),(a3,a6), and sled Body characteristicses point coordinates (Is,Js) where scope Imin≤Is≤Imax,Jmin≤Js≤Jmax;a1、a2、a3、a4、a5、a6It is just whole Number;
(4.2) since the first photos, each secondary photographic intelligence is read, according to Imin≤Is≤Imax,Jmin≤Js≤Jmax In the range of coordinate points (Is,Js) color dataDetermine the position of characteristic point;If Meet simultaneouslyWithAnd JsValue for it is maximum when, then extract (Is, Js), and record width number s;IfCan not meetOrIn any one, then take Js=Js-1
The step (6) is according to horizontal direction pixel value Js, and the parameter in step (1), ask for rocket sledge skid body Velocity series, concretely comprise the following steps:
(5.1) x is given1And x (0)2(0) initial value 0;
(5.2) instantaneous velocity v in the change calculations s pair photos according to characteristic point position in photos, 1≤s≤K -1, tool Body formula is:
vs=kfc (Js+1-Js)
C is the corresponding length of single pixel point in formula, and f is the frequency that camera shoots, Js+1And JsS+1 is represented respectively The horizontal direction coordinate of prying body characteristic point in photo and s photo, k is real for proportionality coefficient k represents that single pixel point corresponds to Distance in the space of border is imaged the ratio of physical length on component, v with itsIt is instantaneous velocity;
(5.3) by the instantaneous velocity v of the characteristic point of all photossIt is arranged in order to form instantaneous velocity sequence sequentially in time Row, obtain the velocity series of skid body after being smoothed to instantaneous velocity sequenceWherein, speed's Computing formula is
Wherein, b0、A1、A2、B1、B2It is the parameter of discretization second-order low-pass filter, x1And x (0)2(0) it is discretization two Two initial parameters of rank low pass filter.
The step (7) obtains the velocity series of the corresponding rocket sledge skid body of each characteristic point, afterwards with fisrt feature point Corresponding velocity series are merged, specially:
(6.1) other characteristic points and fisrt feature point position relationship, i.e. other characteristic point water are determined according to the first photos Square to pixel value Js τWith the horizontal direction pixel value J of fisrt feature pointsBetween relation:
Js τ=Jsτ
Wherein ωτIt is the difference of other characteristic points and fisrt feature point horizontal direction pixel value, τ=2,3,4 ..., Js τRepresent τ characteristic point horizontal direction pixel values;
(6.2) repeat step (5) obtains the velocity series of other characteristic points, will described other characteristic point velocity series and the In chronological sequence order is arranged one characteristic point velocity series, if synchronization has multiple speed, takes these synchronization speed Arithmetic average, thus complete fusion, obtain rocket sledge skid body operation overall process velocity series.
Present invention advantage compared with prior art is as follows:
(1) the skid body speed of service when present invention can be accurately obtained rocket sledge multi-stage engine separating ignition, overcomes thunder Up to the problem for testing the speed, the measurement of shadow shield photoelectric subassembly is inaccurate, certainty of measurement is higher.
(2) this method can half fully automatic operation, it is not necessary to adds excessive artificial participation, can realize photo pitching Calculated with skid body feature point extraction, skid body speed before and after automatic amendment, the engine ignition of azimuth deviation information and noise eliminated, Theoretical clear simple, programming is easy.
(3) at present still without the method that can carry out accurate measurement to the skid body speed of service by high speed camera, this method It is a kind of innovation, for the design rationality for analyzing and evaluating Rocket sled test provides a kind for the treatment of method.
(4) photo is converted to numerical matrix by the present invention, is changed using numerical relation analysis characteristic point position, whole process It is scientific and reasonable, rigorous accurate, while more rapidly can accurately read picture data being analyzed.Photo eigen point is extracted simultaneously Method, the data volume very little of required access, even if the photo that treatment high-volume is continuously shot can also go out in very short time series analysis Characteristic point position changes
(5) at present the amendment photo method that uses all exist it is unsatisfactory local, it is necessary to certain working environment, speed Slowly, efficiency is low and expensive.The present invention sets up the actual change that distortion model directly asks for the direction of motion to original photo, has Its obvious advantage, greatly improves amendment efficiency.
(6) method that the present invention is smoothed using pie slice, the not only significantly influence of release rate noise, and cause Rate curve more rationally, clearly, while being changed using multiple characteristic point analysis rocket sledge skid body speeds, is met and is hidden in environment The requirement that the harshness external condition such as gear, light change, multi-stage ignition can also be accurately measured.
Brief description of the drawings
Fig. 1 realizes flow chart for the present invention;
Fig. 2 is rocket sledge track and the fastener position in original photo.
Fig. 3 is fastener sequence straight line model;
Fig. 4 is secondary distortion model schematic diagram;
Fig. 5 is the photo after correction;
Fig. 6 is positions of the skid body characteristic point I in photo;
Fig. 7 is positions of the skid body characteristic point II in photo;
Fig. 8 is pixel positions of the skid body characteristic point I in different photos;
Fig. 9 is pixel positions of the skid body characteristic point II in different photos;
Figure 10 is pixel positions of the skid body characteristic point I and characteristic point II after alignment in different photos;
Figure 11 is pixel positions of the fused rear skid body characteristic point I in different photos;
Figure 12 is the result of spectrum analysis to velocity noise;
Figure 13 is the calculated value and smooth value of the skid body speed of service.
Specific embodiment
Realization principle of the invention is:The running of skid body is continuously shot using high speed camera, is led to after the completion of experiment Pitching and azimuth information amendment, skid body feature point extraction, the speed of service calculating to high speed camera relative orbit are crossed, is completed to sledge The measurement of running body speed.The method has that rate accuracy is high, thinking is simple and clear, and the advantage that programming is easily realized improves The tachometric survey precision of key position in Rocket sled test.
A kind of rocket sledge image speed measurement method is given below, it is characterised in that step is as follows:
(1) high speed camera is set up by the track at Rocket sled test multi-stage engine separating ignition, the bat of camera is set Take the photograph the number of pixels M of longitudinal length of side in frequency f, original photo0With the number of pixels N of the horizontal length of side0Etc. parameter;Measurement camera phase To the distance between adjacent fastener in the geography information and track of track L;M0、N0Positive integer is, f, L are arithmetic number;Such as Fig. 2 It is shown, in the rocket sledge track of photo bottom mark for rectangular box part for fastener in the position of rocket sledge track, Numerical value represents two pixel distances of adjacent fastener in white box.
(2) during Rocket sled test multi-stage engine separating ignition, K width is continuously shot by high speed camera and is shone Piece;It is made up of the mode of bidimensional pixel additional color in every photos, positional information of the pixel in photo is original The coordinate value of photograph pixel point coordinates system;The origin of original photo pixel coordinate system (omn) in the upper left corner, om axles down, on Axle is towards the right side;The color value of pixel is made up of red, green ratio different with blueness;With five dimension group (m, n, rm,n,gm,n, bm,n) represent, wherein, (m, n) describes position of the pixel in photo, the span of m is from 1 to M0It is just whole Number, the span of n is from 1 to N0Positive integer;(rm,n,gm,n,bm,n) describe the pixel in original photo color letter Breath, red rm,n, green gm,n, blueness bm,nSpan from 0 to 255.
(3) in each photos, the coordinate points (m, n) in original photo pixel coordinate system are transformed into in photo The heart is the point (x, y) in origin, ox axles level right-handed coordinate system oxy (original photo center point coordinate system) to the right, its conversion Formula is as follows:
After converted, with (x, y, rx,y,gx,y,bx,y) describe position and the colouring information of the pixel in original photo; As shown in Figure 3 by original photo by after discoloration treatment, indicia grid graduation mark, original photo center point coordinate system's transverse axis and vertical Axle be figure in intersect heavy black line:
(4) according to the first photos middle orbit distortion information amendment photo before engine ignition so that every in photo Space length corresponding to individual pixel is equal;Correct concretely comprising the following steps for photo:
A. the coordinate value of each fastener is from left to right extracted successively and is numbered, h (h >=4) individual fastener is provided with, with coordinate Value (xp,yp) represent, wherein p=1,2 ..., h;As shown in Fig. 2 mark is rectangle side in the rocket sledge track of photo bottom The part of frame be fastener in the position of rocket sledge track, be marked its coordinate value (xp, yp) one by one to it;
B. to the fastener sequence (x in photop,yp), p=1,2 ..., h;Using straight line model yl=klx+bl,
With least square method design factor klAnd blValue, specific formula for calculation is as follows:
Wherein,
C. to the fastener sequence (x in photop,yp), p=1,2 ..., h;Calculate two neighboring fastener apart from lqAnd center Position coordinates (x0,q,y0,q), q=1,2 ..., h-1;Specific formula for calculation is
D. the secondary distortion model of cut-off trajectoryUsing least square method design factor α and β, tool Body formula is as follows:
Wherein
E. according to factor alpha, klAnd blValue, set M1And N1, and correction original photo,
Specific formula is as follows:
(rm',n',gm',n',bm',n')=(rm,n,gm,n,bm,n)
Wherein array (rm,n,gm,n,bm,n), subscript m, n illustrates the coordinate points position of original photo pixel coordinate system; Array (rm',n',gm',n',bm',n'), subscript m ', n ' illustrates the coordinate points position of photograph pixel point coordinates system after correction, correction Afterwards photograph pixel point coordinates system (om ' n ') be original photo correction after, with the photo upper left corner as origin, om ' axles vertically downward, On ' axles level coordinate system to the right;M1And N1The vertical and horizontal number of pixels of photo after respectively correcting, value should be guaranteed that M ' and n ' is more than 0;(x ', y ') illustrates the coordinate points of optical center coordinate system (ox ' y ') after correction, photo center after correction After point coordinates system (ox ' y ') is original photo correction, with photo center as origin, ox ' axles level right-handed coordinate system to the right.
F. there is the point not being colored in the region being stretched, after being corrected in photograph pixel point coordinates system after correction Photo method for repairing and mending is:
Wherein, (rm″,n″,gm″,n″,bm″,n″) represent amendment after there is no the colouring information of upper color dot in photo, m " and n " is represented The coordinate position of color dot is not gone up by photograph pixel point coordinates system after amendment, and λ is represented around the pixel in eight pixels, red The number of color, green, blue first pixel value sum more than 0.As shown in Figure 5 by revised photo, wherein each pixel Space actual range representated by point size is equal, and the repair part of photo is the outside black portions of photo;
(5) according to distortionless photo after the amendment obtained in step (4), rocket sledge skid body in each photos of extraction Position (I of the fisrt feature point after the amendment in photograph pixel point coordinates system om ' n 's,Js) and the fisrt feature point in this position Pixel valueBehalf photo width number;As shown in fig. 6, fisrt feature point is in the black center line portion of rocket sledge Divide low order end;
(6) parameter in the horizontal direction pixel value obtained according to step (5), and step (1), asks for the instantaneous of skid body Speed, concretely comprises the following steps:
A. x is given1And x (0)2(0) initial value 0;
B. instantaneous velocity v in the secondary photos of change calculations s according to characteristic point position in photos, 1≤s≤K -1, specifically Formula is:
vs=kfc (Js+1-Js)
C is the corresponding length of single pixel point in formula, and f is the frequency that camera shoots, Js+1And JsS+1 is represented respectively The horizontal direction coordinate of prying body characteristic point in photo and s photo, k is real for proportionality coefficient k represents that single pixel point corresponds to Distance in the space of border is imaged the ratio of physical length on component, v with itsIt is instantaneous velocity;
C. by the instantaneous velocity v of the characteristic point of all photossIt is arranged in order to form instantaneous velocity sequence sequentially in time, The velocity series of skid body are obtained after being smoothed to instantaneous velocity sequenceWherein, speed's Computing formula is
Wherein, b0、A1、A2、B1、B2It is the parameter of discretization second-order low-pass filter, x1And x (0)2(0) it is discretization two Two initial parameters of rank low pass filter.If Figure 12 is the result of spectrum analysis to velocity noise, it can be seen that noise is main Frequency range is the high band more than 10Hz;Therefore, second-order low-pass filter is used to be filtered to eliminate high-frequency noise to speed
(7) according to distortionless photo after the correction obtained in step (4), other spies of skid body in each photos are extracted Levy a little, repeat step (5) and (6) are merged, with the speed obtained using characteristic point I after obtaining speed so as to obtain skid body The overall process velocity series of operation.If Figure 13 is the calculated value and smooth value of the skid body speed of service, while the clear earth's surface of smooth value The overall process movement velocity change of bright rocket sledge.
The implementation method of characteristic point is extracted in the step (5):
(1) since the first photos, Z photos are chosen at equal intervals, find these photos photograph pixel point after correction The skid body feature point coordinates (I of coordinate system (om ' n ')s,Js);The pixel value of the characteristic point in this position of skid body is extracted simultaneouslyWhereinRepresent coordinate points (Is,Js) red pixel value,Represent coordinate points (Is,Js) green picture Element value,Represent coordinate points (Is,Js) blue pixel value, behalf picture width number.According to these at equal intervals photo determine prying body Characteristic point colorDomain value range is respectively (a1,a4),(a2,a5),(a3,a6), and prying body characteristic point Coordinate (Is,Js) where scope Imin≤Is≤Imax,Jmin≤Js≤Jmax
(2) since the first photos, each secondary photographic intelligence is read, according to Imin≤Is≤Imax,Jmin≤Js≤JmaxModel Enclose internal coordinate point (Is,Js) color dataDetermine the position of characteristic point;IfTogether When meet And JsValue for it is maximum when, then extract (Is,Js), and And record width number s;IfCan not meet Any one, then take Js=Js-1.As Fig. 8 is characterized pixel positions of the point I in each photos, wherein pixel position is Taken out with the coordinate in photograph pixel point coordinates system after correction, have rough part in curve mainly due to ambient noise and light Line influences, and takes Js=Js-1But, have no effect on curve general trend and precision;After being fused such as Figure 11, skid body characteristic point exists Pixel position in each photos, wherein pixel position is taken out with the coordinate in photograph pixel point coordinates system after correction, Characteristic point II by translation after, by the characteristic point II pixels position and characteristic point I pixels position in chronological sequence order Arrangement, if synchronization has multiple pixel positions, takes the arithmetic average of these synchronization pixel positions.
The measurement of the rocket sledge skid body speed of service can be completed by the above method.
Embodiment:
In a Rocket sled test, if filming frequency f=2000 width/second, the corresponding length c=22 μ of single pixel point M, proportionality coefficient k are=853.65.Fig. 2 be the position of rocket sledge track and fastener in original photo, and by image at Photo disposal is that, without ambient noise, mark is that the part of rectangular box is fastener in the rocket sledge track of photo bottom by reason In the position of rocket sledge track, numerical value represents two pixel distances of adjacent fastener in white box.;Fig. 3 is fastener sequence Original photo point position set up straight line model, coordinate system be with photo center as origin, the ox axles level right hand to the right Coordinate system oxy (original photo center point coordinate system), reference axis photo intermediate interdigitated heavy black line represents that the straight line of foundation is Y=-0.01057x-101.2;Fig. 4 is that secondary distortion model schematic diagram represents pixel position and adjacent fastener apart from derivative Relation, what empty circles were represented in figure is relation of the known adjacent fastener apart from derivative and pixel position, and black curve is According to the matched curve that empty circles position is formed, the distortion model of foundation isFig. 5 is correction Photo afterwards, the space actual range wherein corresponding to each pixel is equal, and has repaired in the region being stretched and exist The part of the point not being colored;Fig. 6 is characterized positions of the point I in skid body, and this characteristic point is middle black in rocket sledge skid body The low order end of color linear parts, has very big difference with surrounding pixel point color value, it is easy to determine that scope takes out its coordinate, meets The operation requirement of characteristic point is read, but, characteristic point I can be blocked in Fig. 7 photos, therefore, selected characteristic point II, this feature Point is black triangles tapered top portion shell white centers point in rocket sledge skid body, and Fig. 7 is seen in its position in skid body;Fig. 8 is spy Pixel positions of the point I in each photos is levied, wherein pixel position is with the seat in photograph pixel point coordinates system after correction Mark takes out, and has rough part to be influenceed mainly due to ambient noise and light in curve, but have no effect on curve and totally become Gesture and precision, Fig. 9 are characterized pixel positions of the point II in each photos, and wherein pixel position is with photo picture after correction Coordinate in vegetarian refreshments coordinate system takes out, and can be carried out by after translation with the position of characteristic point I from the 70th width to the 140th width Compare checking;Figure 10 is the pixel position after alignment in each photos to characteristic point I and characteristic point II pixels position; After being fused such as Figure 11, pixel position of the skid body characteristic point in each photos, wherein pixel position is with after correction Coordinate in photograph pixel point coordinates system takes out, after characteristic point II is by translating, by the characteristic point II pixels position and spy In chronological sequence order is arranged to levy point I pixels position, if synchronization has multiple pixel positions, takes these synchronizations The arithmetic average of pixel position;Figure 12 is the result of spectrum analysis to velocity noise, it can be seen that the main frequency range of noise is High band more than 10Hz;Therefore, second-order low-pass filter is used to be filtered speed to eliminate high-frequency noise, Figure 13 is sledge The calculated value and smooth value of running body speed, it can be seen that calculated value has the fluctuation of constant amplitude, keep stable tendency, while smooth value The movement velocity variation tendency of rocket sledge is clearly showed that, rocket sledge image speed measurement overall process is so far completed.
The non-detailed description of the present invention is known to the skilled person technology.

Claims (6)

1. a kind of rocket sledge image speed measurement method, it is characterised in that step is as follows:
(1) in Rocket sled test, high speed camera is set up by the track at multi-stage engine separating ignition, the bat of camera is set Take the photograph the number of pixels M of longitudinal length of side in frequency f, original photo0With the number of pixels N of the horizontal length of side0;Measurement camera relative orbit Geography information and track on the distance between adjacent fastener L;M0、N0Positive integer is, f, L are arithmetic number;
(2) during Rocket sled test multi-stage engine separating ignition, K photos are continuously shot by high speed camera, K is Positive integer;It is made up of the mode of bidimensional pixel additional color in every photos, positional information of the pixel in photo is The coordinate value of original photo pixel coordinate system;
(3) in each photos, the coordinate points (m, n) in original photo pixel coordinate system are transformed into original photo center Point (x, y) in point coordinates system, its conversion formula is as follows:
x = n - 1 2 N 0 y = - m + 1 2 M 0
After converted, with (x, y, rx,y,gx,y,bx,y) description original photo in the pixel position and colouring information;
The original photo center point coordinate system is with photo center as origin, ox axles level right-handed coordinate system oxy to the right;
(4) according in the prefiring first photos middle orbit distortion information amendment photo of multi-stage engine so that every in photo Space length corresponding to individual pixel is equal;
(5) according to distortionless photo after the amendment obtained in step (4), first of rocket sledge skid body in each photos is extracted Position (I of the characteristic point after the amendment in photograph pixel point coordinates system om ' n 's,Js) and the fisrt feature point this position pixel ValueBehalf photo width number;
(6) the fisrt feature point horizontal direction pixel value J obtained according to step (5)s, and the parameter in step (1), ask for fire The velocity series of arrow sledge skid body;
(7) according to distortionless photo after the amendment obtained in step (4), other of rocket sled skid body in each photos of extraction Characteristic point, repeat step (5) and (6) obtain the velocity series of the corresponding rocket sledge skid body of each characteristic point, special with first afterwards Levy a little corresponding velocity series to be merged, so as to obtain the overall process velocity series of rocket sledge skid body operation.
2. a kind of rocket sledge image speed measurement method according to claim 1, it is characterised in that:Original photo pixel point coordinates Be the origin of omn in the photo upper left corner, down, on axles are towards the right side for om axles;The color value of pixel is different with blueness by red, green Ratio is constituted, with five dimension group (m, n, rm,n,gm,n,bm,n) represent, wherein, (m, n) describes position of the pixel in photo Put, the span of m is from 1 to M0Positive integer, the span of n is from 1 to N0Positive integer;(rm,n,gm,n,bm,n) description The colouring information of the pixel, red r in original photom,n, green gm,n, blueness bm,nSpan from 0 to 255.
3. a kind of rocket sledge image speed measurement method according to claim 1, it is characterised in that:Step (4) basis exists The prefiring first photos middle orbit distortion information amendment photo of multi-stage engine, specially:
(3.1) in the photo of amendment before processing, the coordinate value of each fastener is from left to right extracted successively and is numbered, be provided with H >=4 fastener, with coordinate value (xp,yp) represent, wherein p=1,2 ..., h;
(3.2) to the fastener sequence (x in photop,yp), p=1,2 ..., h;Using straight line model yl=klx+bl, with a most young waiter in a wineshop or an inn Multiplication design factor klAnd blValue, specific formula for calculation is as follows:
k l b l = ( C l T C l ) - 1 C l T Y l
Wherein,
(3.3) to the fastener sequence (x in photop,yp), p=1,2 ..., h;Calculate two neighboring fastener apart from lqAnd center Position coordinates (x0,q,y0,q), q=1,2 ..., h-1;Specific formula for calculation is
l q = ( x q + 1 - x q ) 2 + ( y q + 1 - y q ) 2 ;
x 0 , q = x q + 1 + x q 2 ;
y 0 , q = y q + 1 + y q 2 ;
(3.4) the secondary distortion model of cut-off trajectoryUsing least square method design factor α and β, tool Body formula is as follows:
β α α 2 = ( A T A ) - 1 A T B
Wherein
(3.5) according to factor alpha, klAnd blValue, set M1And N1, and original photo is corrected, specific formula is as follows:
(rm',n',gm',n',bm',n')=(rm,n,gm,n,bm,n)
m ′ = M 1 2 - y ′ n ′ = x ′ + N 1 2 x ′ = s i g n ( n - N 0 2 ) | 1 α - α 2 ( n - N 0 2 ) - 1 α | y ′ = [ M 0 2 - m - k l · ( n - N 0 2 ) - b l ] x ′ n - N 0 2
Wherein, array (rm,n,gm,n,bm,n) subscript m, n illustrates the coordinate points position of original photo pixel coordinate system;Number Group (rm',n',gm',n',bm',n') subscript m ', n ' illustrate amendment after photograph pixel point coordinates system coordinate points position, after correction Photograph pixel point coordinates system om ' n ' be original photo correction after, with the photo upper left corner as origin, om ' axles vertically downward, on ' axles Level coordinate system to the right;M1And N1The vertical and horizontal number of pixels of photo after respectively correcting, makes m ' and n ' values be more than 0; (x ', y ') represents the coordinate points of optical center coordinate system ox ' y ' after correction, and optical center coordinate system ox ' y ' are after correction After original photo correction, with photo center as origin, ox ' axles level right-handed coordinate system to the right;
(3.6) there is the point not being colored in revised photo, in the region being stretched, photograph pixel point is sat after amendment The region that is stretched is repaired in mark system, specially:
( r m ′ ′ , n ′ ′ , g m ′ ′ , n ′ ′ , b m ′ ′ , n ′ ′ ) = 1 λ ( ( r m ′ ′ - 1 , n ′ ′ - 1 , g m ′ ′ - 1 , n ′ ′ - 1 , b m ′ ′ - 1 , n ′ ′ - 1 ) + ( r m ′ ′ - 1 , n ′ ′ , g m ′ ′ - 1 , n ′ ′ , b m ′ ′ - 1 , n ′ ′ ) + ( r m ′ ′ - 1 , n ′ ′ + 1 , g m ′ ′ - 1 , n ′ ′ + 1 , b m ′ ′ - 1 , n ′ ′ + 1 ) + ( r m ′ ′ , n ′ ′ - 1 , g m ′ ′ , n ′ ′ - 1 , b m ′ ′ , n ′ ′ - 1 ) + ( r m ′ ′ , n ′ ′ + 1 , g m ′ ′ , n ′ ′ + 1 , b m ′ ′ , n ′ ′ + 1 ) + ( r m ′ ′ + 1 , n ′ ′ - 1 , g m ′ ′ + 1 , n ′ ′ - 1 , b m ′ ′ + 1 , n ′ ′ - 1 ) + ( r m ′ ′ + 1 , n ′ ′ , g m ′ ′ + 1 , n ′ ′ , b m ′ ′ + 1 , n ′ ′ ) + ( r m ′ ′ + 1 , n ′ ′ + 1 , g m ′ ′ + 1 , n ′ ′ + 1 , b m ′ ′ + 1 , n ′ ′ + 1 ) ) λ ≠ 0 ( r m ′ ′ , n ′ ′ , g m ′ ′ , n ′ ′ , b m ′ ′ , n ′ ′ ) = 0 λ = 0
Wherein, (rm″,n″,gm″,n″,bm″,n″) represent amendment after there is no the colouring information of upper color dot in photo, m " and n " represents repairing The coordinate position of color dot is not gone up by photograph pixel point coordinates system after just, and λ is represented around the pixel in eight pixels, red, The number of green, blue first pixel value sum more than 0.
4. a kind of rocket sledge image speed measurement method according to claim 1, it is characterised in that:
Fisrt feature point photograph pixel point coordinates after amendment of rocket sledge skid body in each photos is extracted in the step (5) It is the position (I in om ' n 's,Js), and the fisrt feature point is in the pixel value of this positionSpecially:
(4.1) since the first photos, Z photos are chosen at equal intervals, find these photos photograph pixel point after amendment and sit The skid body feature point coordinates (I of mark system om ' n 's,Js);The pixel value of the fisrt feature point in this position of skid body is extracted simultaneouslyWhereinRepresent coordinate points (Is,Js) red pixel value,Represent coordinate points (Is,Js) green picture Element value,Represent coordinate points (Is,Js) blue pixel value, behalf picture width number;
The fisrt feature point of these photos at equal intervals is found, five dimension groups of these fisrt feature points are readScope according to more than where five dimension groups determines skid body fisrt feature point colorThe domain value range of red, green and blueness is respectively (a1,a4),(a2,a5),(a3,a6), and sled Body characteristicses point coordinates (Is,Js) where scope Imin≤Is≤Imax,Jmin≤Js≤Jmax;a1、a2、a3、a4、a5、a6It is just whole Number;
(4.2) since the first photos, each secondary photographic intelligence is read, according to Imin≤Is≤Imax,Jmin≤Js≤JmaxScope Internal coordinate point (Is,Js) color dataDetermine the position of characteristic point;IfSimultaneously MeetWithAnd JsValue for it is maximum when, then extract (Is,Js), and And record width number s;IfCan not meetOrIn Any one, then take Js=Js-1
5. a kind of rocket sledge image speed measurement method according to claim 1, it is characterised in that:The step (6) is according to water Square to pixel value Js, and the parameter in step (1), the velocity series of rocket sledge skid body are asked for, concretely comprise the following steps:
(5.1) x is given1And x (0)2(0) initial value 0;
(5.2) instantaneous velocity v in the change calculations s pair photos according to characteristic point position in photos, 1≤s≤K -1, specific public affairs Formula is:
vs=kfc (Js+1-Js)
C is the corresponding length of single pixel point in formula, and f is the frequency that camera shoots, Js+1And JsRepresent respectively s+1 photos and The horizontal direction coordinate of prying body characteristic point in s photo, k represents that single pixel point corresponds to real space for proportionality coefficient k In distance the ratio of physical length on component is imaged with it
, vsIt is instantaneous velocity;
(5.3) by the instantaneous velocity v of the characteristic point of all photossIt is arranged in order to form instantaneous velocity sequence sequentially in time, it is right Instantaneous velocity sequence obtains the velocity series of skid body after being smoothedWherein, speedCalculating Formula is
x 1 ( s ) = - A 1 x 1 ( s - 1 ) - A 2 x 2 ( s - 1 ) + b 0 v s x 2 ( s ) = x 1 ( s - 1 ) v ‾ s = ( B 1 - A 1 ) x 1 ( s - 1 ) + ( B 2 - A 2 ) x 2 ( s - 1 ) + b 0 v s
Wherein, b0、A1、A2、B1、B2It is the parameter of discretization second-order low-pass filter, x1And x (0)2(0) it is discretization step low-pass Two initial parameters of wave filter.
6. a kind of rocket sledge image speed measurement method according to claim 1, it is characterised in that:The step (7) obtains often The velocity series of the corresponding rocket sledge skid body of individual characteristic point, afterwards velocity series corresponding with fisrt feature point merged, have Body is:
(6.1) other characteristic points and fisrt feature point position relationship, i.e. other characteristic point level sides are determined according to the first photos To pixel value Js τWith the horizontal direction pixel value J of fisrt feature pointsBetween relation:
Js τ=Jsτ
Wherein ωτIt is the difference of other characteristic points and fisrt feature point horizontal direction pixel value, τ=2,3,4 ..., Js τRepresent τ Characteristic point horizontal direction pixel value;
(6.2) repeat step (5) obtains the velocity series of other characteristic points, will described other characteristic point velocity series and the first spy In chronological sequence order is arranged to levy spot speed degree series, if synchronization has multiple speed, takes the calculation of these synchronization speed Number average value, thus completes fusion, obtains the overall process velocity series of rocket sledge skid body operation.
CN201611050474.2A 2016-11-23 2016-11-23 A kind of rocket sledge image speed measurement method Active CN106771320B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201611050474.2A CN106771320B (en) 2016-11-23 2016-11-23 A kind of rocket sledge image speed measurement method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201611050474.2A CN106771320B (en) 2016-11-23 2016-11-23 A kind of rocket sledge image speed measurement method

Publications (2)

Publication Number Publication Date
CN106771320A true CN106771320A (en) 2017-05-31
CN106771320B CN106771320B (en) 2019-03-12

Family

ID=58911141

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201611050474.2A Active CN106771320B (en) 2016-11-23 2016-11-23 A kind of rocket sledge image speed measurement method

Country Status (1)

Country Link
CN (1) CN106771320B (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108716878A (en) * 2018-04-18 2018-10-30 西北机电工程研究所 A kind of dynamic flying target Damage proficiency testing system and its test method
CN108845341A (en) * 2018-07-04 2018-11-20 中国船舶工业系统工程研究院 A kind of dynamic locating accuracy acceptance test method of GPS positioning system
CN110618287A (en) * 2019-09-20 2019-12-27 西安航天动力测控技术研究所 High-speed photographic data processing method for measuring object space movement
CN111028210A (en) * 2019-11-25 2020-04-17 北京航天控制仪器研究所 Deep neural network glass tube end surface defect detection method based on sparse automatic encoder
CN115096196A (en) * 2022-08-25 2022-09-23 中国科学院西安光学精密机械研究所 Visual height and speed measuring method and system for rocket recovery and storage medium

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH11211738A (en) * 1998-01-27 1999-08-06 Omron Corp Speed measurement method of traveling body and speed measuring device using the method
CN1987484A (en) * 2005-12-20 2007-06-27 日产自动车株式会社 Image velocity computer and method for computing image velocity
CN103278051A (en) * 2013-05-30 2013-09-04 北京航天控制仪器研究所 Speed measuring method used in rocket sled test
CN104331882A (en) * 2014-10-24 2015-02-04 南京理工大学 Method for measuring speed of aircraft

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH11211738A (en) * 1998-01-27 1999-08-06 Omron Corp Speed measurement method of traveling body and speed measuring device using the method
CN1987484A (en) * 2005-12-20 2007-06-27 日产自动车株式会社 Image velocity computer and method for computing image velocity
CN103278051A (en) * 2013-05-30 2013-09-04 北京航天控制仪器研究所 Speed measuring method used in rocket sled test
CN104331882A (en) * 2014-10-24 2015-02-04 南京理工大学 Method for measuring speed of aircraft

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
刘泽庆 等: "基于数字摄影测量的靶场高速摄影测速方法", 《弹道学报》 *

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108716878A (en) * 2018-04-18 2018-10-30 西北机电工程研究所 A kind of dynamic flying target Damage proficiency testing system and its test method
CN108845341A (en) * 2018-07-04 2018-11-20 中国船舶工业系统工程研究院 A kind of dynamic locating accuracy acceptance test method of GPS positioning system
CN110618287A (en) * 2019-09-20 2019-12-27 西安航天动力测控技术研究所 High-speed photographic data processing method for measuring object space movement
CN111028210A (en) * 2019-11-25 2020-04-17 北京航天控制仪器研究所 Deep neural network glass tube end surface defect detection method based on sparse automatic encoder
CN111028210B (en) * 2019-11-25 2023-07-18 北京航天控制仪器研究所 Glass tube end face defect detection method based on deep neural network
CN115096196A (en) * 2022-08-25 2022-09-23 中国科学院西安光学精密机械研究所 Visual height and speed measuring method and system for rocket recovery and storage medium

Also Published As

Publication number Publication date
CN106771320B (en) 2019-03-12

Similar Documents

Publication Publication Date Title
CN106771320B (en) A kind of rocket sledge image speed measurement method
CN104166731B (en) A kind of overlapping community discovery system and method for social networks
CN104732553B (en) A kind of Feature Points Extraction based on many laser assisted targets
CN107424181A (en) A kind of improved image mosaic key frame rapid extracting method
CN106097348A (en) A kind of three-dimensional laser point cloud and the fusion method of two dimensional image
CN106408527A (en) Automatic target scoring method based on video analysis
CN109934811A (en) A kind of optical element surface defect inspection method based on deep learning
CN102540252B (en) High-precision median stacking method on basis of cross-correlation
CN106680798B (en) A kind of identification of airborne LIDAR air strips overlay region redundancy and removing method
CN103365998B (en) A kind of similar character string search method
CN108982901A (en) A kind of rotating speed measurement method of at the uniform velocity rotary body
CN105913435B (en) A kind of multiscale morphology image matching method and system suitable for big region
CN110414380A (en) A kind of students ' behavior detection method based on target detection
CN110191335A (en) Remote sensing camera evaluation of imaging quality and analogue system
CN101285666A (en) Laser targeting device based on four quadrant optical-electric detector
CN103942786B (en) The self adaptation block objects detection method of unmanned plane visible ray and infrared image
CN106970417B (en) Ellipse expansion converted-wave velocity analysis method and system
CN103792069B (en) A kind of optical imaging system focal length accurate measurement method based on moon imaging
CN109916494A (en) Weighing-appliance scaling method and device
CN110363706A (en) A kind of large area bridge floor image split-joint method
CN102768474B (en) Projection objective lens wave aberration detection method based on two-order aberration model
CN205192728U (en) Star sensor optical system multiplying power colour difference test equipment
CN105223630B (en) Omnibearing observation systematic parameter Demonstration Method based on geological model
CN105783785A (en) Wavelet-ridge phase extraction method
CN112053331B (en) Bridge crack detection method based on image superposition and crack information fusion

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