CN1030863C - Movement objective orbit image analytical method - Google Patents

Abstract

The movement characteristics of a movement target are analyzed by images of the movement target. For a group of continuous images of the movement target, when each frame graph comes along, the grey values of positions of the movement target on the frame graph are recorded by a grey comparison method so as to form a grey axial trace graph. When the grey values of the positions of the movement target on the frame graph are recorded, pixel values corresponding to the positions of the movement target in another frame graph are written into the number of the frame so as to form a trace graph with a frame number. Then, the grey axial trace graph or the trace graph with the frame number is input into a computer through an interface circuit for analysis.

Description

Movement objective orbit image analytical method

The present invention relates to moving target be carried out the method for specificity analysis with image processing techniques.Moving target floats on the buoy that is used to measure fluid flow characteristics on the liquid level just like the sperm that moves about in the seminal fluid, or the like.Kinetic characteristic is meant the form of its movement locus, the speed of motion, and acceleration, travel direction, or the like.

The method that adopts is with camera moving target to be taken a picture constantly in difference earlier at present, again the film of picked-up is input to computing machine through the projection digitizer with the coordinate position of moving target, so just can obtain this target in difference position constantly, thereby the analysis kinetic characteristic adopts 86 years products " nac film motion analysis system (model 200A) " of system such as Japanese Na Ku company of this method.This method must manually be found out each moving target at each coordinate position constantly.Can certainly adopt image processing techniques, find out each moving target automatically at each coordinate position constantly by computing machine.Live image can be got off with cinefilm or video recording earlier, send into computing machine more frame by frame and carry out target and locate automatically, because the data volume of every two field picture is very big, processing speed is very slow.Also can adopt high speed device to carry out real-time video and handle, finish the kinetic characteristic analysis with hardware, the problem of bringing thus is the hardware complexity, and cost is extremely expensive.

The purpose of this invention is to provide a kind of method of analyzing the moving target kinetic characteristic, this method needn't be absorbed film or video-tape, need not manually to seek the coordinate position of moving target, does not also need expensive hardware device, but automatically analyzes apace.

The present invention is recorded in Moving Target on the piece image in real time, automatically analyzes each bar track on this width of cloth image by computing machine then.Usually moving target can show as darker or brighter on image.The hypothesis motive target imaging is darker earlier, also is that the gray-scale value of its place pixel is less, and the image sequence that video camera obtains is S, asks the pointwise minimum value just can form trajectory diagram.If motive target imaging is brighter, then ask the pointwise maximal value just can form trajectory diagram.

Key of the present invention is to form a width of cloth trajectory diagram.Be provided with the motion that one group of consecutive image S has write down target, S={s ₀, s ₁, s ₂..., s _N-1, p _Ij ^kRepresent K width of cloth image s _kIn the gray-scale value of the capable pixel of i row j, establishing trajectory diagram is T, q _IjThe capable pixel of i row j among the expression trajectory diagram T.If moving target is darker than background, then get $q_{\mathrm{ij}}=\underset{k}{\mathrm{Min}}\left\{p_{\mathrm{ij}}^k\right\};----\left(1\right)$ At this moment T is the figure that contains dark track.Moving target is brighter than background else if, then gets $q_{\mathrm{ij}}=\underset{k}{M}\mathrm{ax}\left\{p_{\mathrm{ij}}^k\right\};---\left(2\right)$ At this moment T is the figure that contains bright track.

If the tonal range of moving target is [T ₁, T ₂], make function

If (1) formula and (2) formula change into respectively: $q_{\mathrm{ij}}=\underset{k}{\mathrm{Min}}\left\{f_1\left(p_{\mathrm{ij}}^k\right)\right\};---\left(5\right)$ $q_{\mathrm{ij}}=\underset{k}{M}\mathrm{ax}\left\{f_2\left(p_{\mathrm{ij}}^k\right)\right\};----\left(6\right)$ Then because shielded part background gray scale, background interference reduces on the trajectory diagram of formation.

Because the background gray scale is constant on each frame figure in S, only contains background if there is a width of cloth figure B not contain moving target, its pixel is b _Ij, after subtracting each other, trajectory diagram and Background can eliminate more background interference, and at this moment corresponding (1) and (2) formula is: $q_{\mathrm{ij}}=b_{\mathrm{ij}}-\underset{k}{M}\mathrm{in}\left\{p_{\mathrm{ij}}^k\right\};---\left(7\right)$ $q_{\mathrm{ij}}=\underset{k}{\mathrm{Max}}\left\{p_{\mathrm{ij}}^k\right\}-b_{\mathrm{ij}};---\left(8\right)$ Corresponding (5) and (6) are $q_{\mathrm{ij}}=f_1\left(b_{\mathrm{ij}}\right)-\underset{k}{\mathrm{Min}}\left\{f_1\left(P_{\mathrm{ij}}^k\right)\right\};---\left(9\right)$ $q_{\mathrm{ij}}=\underset{k}{\mathrm{Max}}\left\{f_2\left(P_{\mathrm{ij}}^k\right)\right\}-f_2\left(b_{\mathrm{ij}}\right);---\left(10\right)$

Because being order, each frame figure of consecutive image S obtains; form above-mentioned movement objective orbit figure; can be earlier depositing a two field picture of moving target in frame deposits; with the picture signal of follow-up input be retained in the picture signal of frame in depositing and compare; make frame deposit middle less gray-scale value of this two width of cloth figure same position or the bigger gray-scale value of keeping; thereby form dark track or bright track, promptly so-called gray scale trajectory diagram.Can be used to analyze kinetic characteristics such as each Moving Target form, movement velocity.

When forming the gray scale trajectory diagram,, claim this two field picture trajectory diagram to be had contribution at this pixel if certain frame has been rewritten content during frame is deposited at certain location of pixels.With the i frame si of sequence S to the pixel of another width of cloth figure of the contribution part opposite position of gray scale trajectory diagram with frame number i assignment, then obtain the position frame by frame of moving target, the trajectory diagram that this has just formed the band frame number can be used to calculate acceleration, instantaneous velocity and transient motion direction etc.

Fig. 1 is the synoptic diagram that the method according to this invention forms the gray scale trajectory diagram.Vision signal from video camera enters A-D converter (ADC) 1, converts digital signal to, enters the table (LUT1) 2 of noting again, if will form dark track, the table 2 of noting is finished (3) formula function; If will form bright track, the table 2 of noting is finished (4) formula function.The track that forms should be the result of (5) formula or (6) formula mutually, if do not need to shield the background gray scale, then can cancel the table 2 of noting, and the table 2 of maybe will noting is arranged to linear oblique ascension amount, i.e. f (x)=x.The table 2 of noting can be set by computer program.

Before track formed, control signal E got low level, and Sheffer stroke gate 4 output makes the output of either-or switch 5 equal to import A, i.e. D=A, and at this moment the gray scale frame is deposited the current frame image that 8 data are exactly video camera.If the formation track is then put E=1, current image in frame is deposited is as O frame figure s ₀At the 1st frame figure s ₁During arrival, A and B are meant s respectively ₁And s ₀In the pixel value of same position, A is from the table 2 of noting, and B deposits the data of reading 8 from gray scale track frame, if comparer output C is with A＞B end, D=B when D=A during A＞B then, A≤B, formation be bright track; If comparer output C is with A＜B end, D=B when D=A during A＜B then, A 〉=B, formation be dark track.After the first frame figure finishes, during depositing, frame just leaves s ₀To s ₁Trajectory diagram.Similarly, s ₂, s ₃..., s _N-1After all having imported, during depositing, frame just leaves from s ₀To s _N-1The track of motion during this period of time.

Gray scale track frame is deposited 8 output and is gone monitor to show through digital-to-analog converter (DAC) and synchronizing circuit.And directly send the computer interface circuit to read in computing machine, analyze by computing machine.

Fig. 2 is that the method according to this invention not only forms the gray scale trajectory diagram but also form the block diagram of being with the frame number trajectory diagram.When forming trajectory diagram with the method for Fig. 1, if either-or switch 5 outputs equal to import A, also just claim this moment that trajectory diagram is had contribution, deposit if establish a frame more in addition, make its corresponding i frame figure be changed to frame number i, record the position of moving target during then this frame is deposited at each frame to the contributive pixel value of trajectory diagram.This frame is deposited and is called the track frame of being with frame number and deposits, and the image during this frame is deposited is called the trajectory diagram of being with frame number.

But the same topic of bringing thus is that the trajectory diagram of band frame number is indeterminate in track crossover part value.So use the mode of Fig. 2 instead.The pixel value of gray scale trajectory diagram is added a side-play amount through the table (LUT2) 9 of noting, go comparer 3 and present frame input signal A to compare again.Comparer 3 output C through control circuit directly two frames of control deposit 8,10 write and allow end.At this moment gray scale track frame trajectory diagram and Fig. 1 mode of depositing in 8 is basic identical.

The adding of table 9 of noting makes the frame number trajectory diagram get less or bigger frame number in the pixel of track crossover.Noting table 9 can be by the computer program setting, and also curable, the also available totalizer of the table 9 of noting replaces.If do increasing or reduce computing, silent gray scale track, the table 9 of noting make the frame number trajectory diagram get bigger or less frame number in that the pixel of track crossover is corresponding; On the contrary, if bright gray scale track, look-up table 9 is done to increase or is reduced computing and makes the frame number trajectory diagram get less or bigger frame number in that the pixel of track crossover is corresponding.Thereby guarantee that the frame number trajectory diagram partly has clear and definite value at the track crossover.

Two track frames are deposited 8,10 output through either-or switch 11, go monitor to show through digital-to-analog converter (DAC) and synchronizing circuit again, and directly send the computer interface circuit to read in computing machine, are analyzed by computing machine.

In addition, can make comparer 3 output C ineffective by control circuit 7, signal A directly enters frame and deposits 8, thereby forward one frame only contains the image of background, is carried out by computer program and subtracts each other, and promptly does (7)-(10) formula computing, eliminates the interference of background.

As mentioned above, the device that the present invention constitutes can be analyzed kinetic characteristic automatically, and speed is fast, and hardware is very simple.And prior art, or can not analyze automatically, cost is high again simultaneously, or cost is extremely expensive, can't promote.Another advantage of the present invention is that the frame number of formation track can be very big, and the frame number during the gray scale trajectory analysis only depends on the length of frame counter 6, if use 16 bits, can reach 65535 frames, if video signal source was 25 frame/seconds, then is equivalent to 43.7 minutes.To band frame number trajectory analysis, except that the length of frame counter 6, frame number also depends on is with frame number track frame to deposit 10 bit number, if use 8 bits, then can analyze 255 frames, if use 16 bits, just can analyze 65535 frames.Therefore, the present invention can realize the good movement destination image analytic system of cost-performance ratio.

Fig. 1 is the block diagram that explanation forms gray scale trajectory diagram method.

Fig. 2 illustrates the block diagram that not only forms the gray scale trajectory diagram but also form the method for band frame number trajectory diagram.

Claims (3)

1. movement objective orbit image analytical method, under computer control, a frame video image signal of the moving object that video camera is obtained deposits frame memory in through analog to digital conversion; Import computing machine analysis at last, it is characterized in that again with the picture signal of follow-up input be retained in the picture signal of frame in depositing and compare, make frame deposit middle less gray-scale value of this two images same position or the bigger gray-scale value of keeping, thereby form dark trajectory diagram or bright trajectory diagram, it is the gray scale trajectory diagram, it is that gray scale track frame is deposited that this frame is deposited, and at last the gray scale trajectory diagram is analyzed by interface circuit input computing machine.

2. movement objective orbit image analytical method according to claim 1, when it is characterized in that the content in each rewriting gray scale track frame is deposited, the respective pixel position writes the present frame numbering as pixel value in another frame is deposited, thereby form band frame number trajectory diagram, will be with frame number trajectory diagram input computing machine to analyze again.

3. movement objective orbit image analytical method according to claim 2, it is characterized in that adding a side-play amount earlier after the picture signal of gray scale track frame in depositing read goes to compare with received image signal again, and directly control writing that gray scale track frame is deposited and frame number track frame is deposited according to comparative result and allow end.

Images