CN105258673A - Target range finding method and apparatus based on binocular synthetic aperture focus image - Google Patents
Target range finding method and apparatus based on binocular synthetic aperture focus image Download PDFInfo
- Publication number
- CN105258673A CN105258673A CN201510732130.9A CN201510732130A CN105258673A CN 105258673 A CN105258673 A CN 105258673A CN 201510732130 A CN201510732130 A CN 201510732130A CN 105258673 A CN105258673 A CN 105258673A
- Authority
- CN
- China
- Prior art keywords
- image
- information
- synthetic aperture
- target
- width images
- 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
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C3/00—Measuring distances in line of sight; Optical rangefinders
- G01C3/32—Measuring distances in line of sight; Optical rangefinders by focusing the object, e.g. on a ground glass screen
Abstract
The invention discloses a target range finding method based on a binocular synthetic aperture focus image. The method comprises the following steps: first taking a symmetric center of two area arrays identical in structures as a benchmark position, respectively carrying out the optical synthetic aperture imaging for a scene in which a target stays, and respectively obtaining a group of synthetic aperture images corresponding to different focusing distances; then taking the difference between binocular matching position compatibility information amount and binocular mismatch position compatibility information amount as a detection amount, finding out a maximum focusing distance of the detection amount, if the binocular matching position compatibility information amount of the focusing distance is greater than a preset compatibility information amount threshold value, the focusing distance is a target distance. The invention also discloses a target range finding apparatus based on the binocular synthetic aperture focus image. Compared with the prior art, the common-mode error and misfocus imaging interference can be effectively inhibited, and the synthetic aperture focus image range finding performance is improved.
Description
Technical field
The present invention relates to a kind of distance-finding method, particularly relate to a kind of target ranging method based on binocular synthetic aperture focusing image, device.
Background technology
The method of measurement target distance has multiple, as the some telemetry based on Laser pulse time difference, the degree of depth based on Extent Modulate Wave phase differential obtain, geometrical optics focusing, Moire topologies, holographic interferometry, Fresnel diffractive technology and Structure light method etc.Ultrasonic being also commonly used to obtains depth information.Computer stereo vision technique is used widely in recent years in a lot of fields.Basic skills goes to observe Same Scene from two or more viewpoint, obtain one group of image under different visual angles, obtain the parallax between respective pixel in different images, then the depth information of target in scene is measured by triangulation calculation, it needs to determine the corresponding point in binocular or many orders image, and this is a very difficult problem.When space three-dimensional scene is projected as two dimensional image, some useful informations lost due to projection, same scenery can be very different in the image under different points of view, blocked or the impact of shade, some points of scenery likely do not appear in all images, and the many changing factors in scene, as illumination condition, noise, the distortion of scene geometry, surface physical characteristic and camera properties etc., all be aggregated in single image intensity value, only to determine that above factors is very difficult by this gray-scale value, so far this problem is not also well solved.Increase base length and can improve depth survey precision, but the difference between image can be increased simultaneously, increase the degree of difficulty of coupling.Multimachine position pinhole imaging system can synthesize large aperture image, consistent with single-lens imaging.There are document [Kusumoto, N. in Japan; Hiura, S.; Sato, K.UncalibratedSyntheticApertureforDefocusControlIEEEConf erenceonComputerVisionandPatternRecognition, 2009.CVPR2009.P:2252-2259] have studied utilize synthetic aperture method to pin-hole imaging photo carry out art processing, make it to produce non-main body defocus effect, with outstanding case of visual emphasis.Monocular gathers/range of defocusing method development comparatively maturation, use multiple video camera can obtain monocular large aperture effect according to synthetic aperture principle, therefore can use for reference monocular imaging abundant poly-/range of defocusing algorithm, synthetic aperture imaging is found range and has higher development foundation.Because the digital equipment cost meeting pin-hole imaging feature is very low, imaging results is convenient to digitized processing, and the development prospect of synthetic aperture imaging is wide.
One section of Chinese invention patent (CN103033166B) discloses a kind of target ranging method based on synthetic aperture focusing image, the method comprises the following steps: step 1, the seat in the plane of the linear array at equal intervals image sequence utilizing the acquisition of national forest park in Xiaokeng video camera vertical with line of sight, and the image of linear array and line of sight point of intersection seat in the plane is as benchmark image; Step 2, distance range can be surveyed be divided into multiple distance segment, for each distance segment, first calculate the aberration between each width image and benchmark image in described image sequence respectively, then superpose after width image each in image sequence being carried out aberration correction, obtain the aberration correction superimposed image corresponding to this distance segment; The corresponding width aberration correction superimposed image of each distance segment;
The similarity of respective regions in the neighborhood of each pixel and each width aberration correction superimposed image in step 3, Calculation Basis image, and choose similarity and be greater than the pixel of a predetermined threshold value as pixel of can finding range with the scope that aberration correction superimposed image changes; Step 4, for each the found range pixel in benchmark image, select the aberration correction superimposed image that the similarity of the neighborhood of respective regions and this pixel of can finding range is maximum, this distance segment corresponding to aberration correction superimposed image is the corresponding distance segment residing for impact point of this pixel of can finding range.The method utilizes the principle of synthetic aperture focusing imaging to carry out object ranging, has and realizes the advantages such as cost is low, antijamming capability is strong, algorithm is simple.But it similarly is difficult for obtaining accurate synthetic aperture focusing, especially large aperture, and the focusing of follow shot image sequence is more difficult; The interference of defocused image patch also can not be ignored; In addition, synthetic aperture superposition picture has super-resolution rebuilding potentiality, and the range finding of existing monocular is difficult to utilize to these potentiality.
Summary of the invention
Technical matters to be solved by this invention is to overcome prior art deficiency, provides a kind of target ranging method based on binocular synthetic aperture focusing image, effectively can suppress common-mode error and defocused image interference, improve the performance of synthetic aperture focusing image range finding.
The present invention specifically solves the problems of the technologies described above by the following technical solutions:
Based on a target ranging method for binocular synthetic aperture focusing image, comprise the following steps:
Step 1, from the identical face battle array of two structures, using the symcenter of described two face battle arrays as benchmark seat in the plane, respectively optical synthesis aperture imaging is carried out to target place scene, obtain the synthetic aperture image of one group of corresponding different focusing distance respectively;
Step 2, to each focusing distance, using the two width synthetic aperture images obtained from two face battle arrays corresponding to it as a pair, then following process is carried out to each Technologies Against Synthetic Aperture image: the target image in wherein piece image is as with reference to image, using region corresponding with reference picture in another piece image as matched position image, and by the mould of the aberration vector in another piece image and between matched position image (0, d] the interior a series of regions changed of scope are as non-matching location drawing picture, and d is default maximum aberration; Compatible quantity of information R between computing reference image and matched position image
0and the compatible quantity of information between reference picture and each non-matching location drawing picture, and be greater than R by the compatible quantity of information between calculated reference picture and each non-matching location drawing picture
0primary system one be revised as R
0, then calculate the average of the compatible quantity of information between amended reference picture and each non-matching location drawing picture; Last with R
0with the difference of the described average detection limit as this focusing distance; Described compatible quantity of information for measure two width images information in the amount of same section;
Step 3, find out the maximum focusing distance of detection limit, and judge the R corresponding to this focusing distance
0whether be greater than default compatible quantity of information threshold value, in this way, then using this focusing distance as target range.
Following technical scheme can also be obtained according to identical invention thinking:
Based on an object ranging device for binocular synthetic aperture focusing image, comprising:
Binocular synthetic aperture imaging unit, for from the identical face battle array of two structures, using the symcenter of described two face battle arrays as benchmark seat in the plane, respectively optical synthesis aperture imaging is carried out to target place scene, obtain the synthetic aperture image of one group of corresponding different focusing distance respectively;
Distance detecting unit, the synthetic aperture image of two groups for obtaining according to binocular synthetic aperture imaging unit corresponding different focusing distances detects target range, its detection method is specific as follows: to each focusing distance, using the two width synthetic aperture images obtained from two face battle arrays corresponding to it as a pair, then following process is carried out to each Technologies Against Synthetic Aperture image: the target image in wherein piece image is as with reference to image, using region corresponding with reference picture in another piece image as matched position image, and by the mould of the aberration vector in another piece image and between matched position image (0, d] a series of regions of change in scope are as non-matching location drawing picture, d is default maximum aberration, compatible quantity of information R between computing reference image and matched position image
0and the compatible quantity of information between reference picture and each non-matching location drawing picture, and be greater than R by the compatible quantity of information between calculated reference picture and each non-matching location drawing picture
0primary system one be revised as R
0, then calculate the average of the compatible quantity of information between amended reference picture and each non-matching location drawing picture, last with R
0with the difference of the described average detection limit as this focusing distance, described compatible quantity of information for measure two width images information in the amount of same section, find out the focusing distance that detection limit is maximum, and judge the R corresponding to this focusing distance
0whether be greater than default compatible quantity of information threshold value, in this way, then using this focusing distance as target range.
Preferably, described battle array is rounded face battle array; Described optical synthesis aperture imaging takes realization by the national forest park in Xiaokeng video camera array gathered in rounded face battle array simultaneously, or is realized by a series of densely arranged seat in the plane timesharing shooting in rounded face battle array of single national forest park in Xiaokeng video camera.
Preferably, for gray level image, the metric form of described compatible quantity of information is the product of similarity between the information entropy of two width images and two width images, or be the product of similarity between the average information entropy of two width images and two width images, or be the acutance average of two width images or gradient mean value or the product of similarity between contrast average and two width images, or be the cross-entropy of two width images; For coloured image, the metric form of described compatible quantity of information be adopt the compatible informational content measure form of gray level image to obtain two width images in RGB color space between each passage subimage compatible quantity of information and or average.
Preferably, described aberration vector mould (0, d] change by pixel in scope.
Compared to existing technology, the present invention has following beneficial effect:
1) can be embodied as the flexible movement of image planes with signal processing method, lens opening (aperture) be hypertonic flexibly, can realize poly-/loose measuring distance information with signal processing method.
2) the method has abandoned the process of images match, and obtains target range according to all image informations, and result affects minimum by single camera, also substantially do not affect by picture material.
3) same national forest park in Xiaokeng video camera timesharing can be utilized to take and obtain binocular synthetic aperture image sequence, be convenient to realize, the application especially for range finding of taking photo by plane has more significance.
4) the passive mapping do not transmitted earthward has hidden, without Contamination of Electromagnetic Wave, can effectively reduce mapping cost;
5) face, left and right battle array is gathered in two large disc battle arrays, and length of synthetic aperture is less, can avoid barrel shape distortion correction, to realize real-time measurement.
6) compare the existing target ranging method based on synthetic aperture focusing image, the present invention can suppress common-mode error and defocused image interference, improves the performance of synthetic aperture focusing image range finding.
7) because of poly-defocusing effect, be not attenuated, the easier target from same distance of image information entropy at the target information entropy of focusing distance, therefore in synthetic aperture superposition picture, target neighborhood can be relatively large, to improve ranging information source.
Accompanying drawing explanation
Fig. 1 is optical synthesis aperture image-forming principle schematic diagram;
Fig. 2 is the target ranging method schematic flow sheet in the specific embodiment of the invention;
Fig. 3 is the binocular synthetic aperture imaging cell schematics that the present invention preferably uses;
Fig. 4 is the image sequence that pinhole camera acquisition time arrives;
Fig. 5 is obtained part synthetic aperture image, wherein (a) focuses on vehicle forward position place for left order, b () be relative distant view place for left order focuses on, (c) focuses on vehicle forward position place for right order, and (d) be relative distant view place for right order focuses on;
Fig. 6 is the change profile schematic diagram of the compatible quantity of information of binocular target image with focusing distance and parallax of different target; Wherein, a the measuring distance of target of () is on building edge line, b () compatible quantity of information corresponding to (a) is with the change profile schematic diagram of focusing distance and parallax, c the measuring distance of target of () is building edge line with upper part, d () compatible quantity of information corresponding to (c) is with the change profile schematic diagram of focusing distance and parallax, (e) is compatible quantity of information after peak clipping process with the change profile schematic diagram of focusing distance and parallax;
Fig. 7 is that detection limit is with distance change curve, wherein, a () is for compatible quantity of information when measuring distance of target is on building edge line is with the change curve of focusing distance, b () holds quantity of information with the change curve of focusing distance at building edge line with upper part phase for measuring distance of target, (c) for measuring distance of target at building edge line with the change curve of the difference of the quantity of information average compatible with non-matching position of the compatible quantity of information in binocular ranging position during upper part with focusing distance;
Fig. 8 is the target depth image utilizing the inventive method to obtain.
Embodiment
Below in conjunction with accompanying drawing, technical scheme of the present invention is described in detail:
For ease of public understanding, first the ultimate principle of Technologies Against Synthetic Aperture focal imaging is introduced.During high-aperture lenses imaging, by the incident ray at each position of camera lens after camera lens refraction, target light is assembled again, becomes two-dimensional image at imaging surface.If incident ray to be pressed passed through sub-aperture separate imaging, then superpose with signal processing method, just obtain synthetic aperture picture, imaging results is consistent with high-aperture lenses imaging.As shown in Figure 1, during full aperture imaging, correct distance place point on imaging surface, is a bit through high-aperture optical lens focus; And the impact point of another distance is imaged as a circle, it is made up of each sub-aperture image, and therefore we can emulate high-aperture lenses imaging process, and the method can applying monocular range finding measures the distance of synthetic aperture target.Form synthetic aperture picture, do not require that in Fig. 1 macropore collar, all sub-aperture pictures all participate in calculating, and only need parton aperture image to participate in superposition calculation, just can form synthetic aperture effect.Synthetic aperture range finding is with the difference of general multi-vision visual the synthetic aperture gain that synthetic aperture focusing is formed, improve signal to noise ratio (S/N ratio), in addition while it focuses on specific range, target astigmatism Jiao of other distances is the image patch of uniform smooth, the interference of these image patches to range finding is relatively little, and that is synthetic aperture is by defocusing the interference of other distance target pictures of filtering.Synthetic aperture focusing telemetry only focuses on specific range target, and focus objects signal energy is concentrated, and little by other distance objective effect of signals, focus on as concentration of energy in zonule, be easy to distinguish with other distance objectives, algorithmic stability, is suitable for engineer applied.But when length of synthetic aperture is larger, what obtain focus similarly is difficult, and the factors such as barrel shape distortion make general convergence correction very complicated.
High-aperture lenses only focuses on specific range target, and other distance targets defocus.Target information is weighed with the form of the entropy of image, and focus on the information entropy that picture maintains target image, defocusing effect then makes target information decay, and therefore can judge target range by the reservation situation detecting target image information entropy.The reservation situation of target image information entropy can be weighed by the compatible quantity of information between left and right order synthetic aperture image in theory, the basic skills of this binocular synthetic aperture focusing range finding just.Compatible quantity of information between two width images refers to the amount of same section in the information of two width images, its concrete tolerance can adopt various rational form, preferably, for gray level image, the metric form of described compatible quantity of information is the product of similarity between the information entropy of two width images and two width images, or be the product of similarity between the average information entropy of two width images and two width images, or the cross-entropy of two width images; For coloured image, the metric form of described compatible quantity of information be adopt the compatible informational content measure form of gray level image to obtain two width images in RGB color space between each passage subimage compatible quantity of information and or average.Certainly, also can adopt other metric forms, such as, can utilize the information entropy that the acutance of two width images or gradient or contrast replace in above-mentioned metric form, the tolerance of compatible quantity of information can be realized equally; Also can adopt or construct the metric form with above-mentioned compatible quantity of information with Similar trend voluntarily.Studholme proposes normalized mutual information (NormalizedMutualInformation, NMI):
(x in formula, y) be pixel coordinate, H (R) and H (F) is not that benchmark image and composograph are at coordinate (x, y) information entropy of neighborhood image block, H (R, F) combination entropy, the most frequently used histogram method calculates, highlighting from angle intuitively replaces probability to carry out density Estimation with frequency, and namely this embodiment adopts histogram method to calculate Mutual information entropy to measure compatible quantity of information.Monocular synthetic aperture focusing image is utilized to carry out object ranging, when length of synthetic aperture is larger, what obtain focus similarly is difficult, the factors such as barrel shape distortion make general convergence correction very complicated, this just needs measurement target distance under the condition that synthetic aperture focusing is bad, the meaning of this binocular synthetic aperture range finding just.Face, binocular synthetic aperture left and right battle array has consistance in design, and make synthetic aperture error also have consistance, in error analysis, uniform error is also called common-mode error, and difference processing can suppress common-mode error.System makes binocular synthetic aperture focusing common-mode error occupy larger specific gravity, creates condition for suppressing error range finding.The detection of target range affects as Size of Neighborhood by target, because of poly-defocusing effect, be not attenuated at the target information entropy of focusing distance, the easier target from same distance of image information entropy, therefore in synthetic aperture superposition picture, target neighborhood can be relatively large, to improve ranging information source.
With a specific embodiment, technical solution of the present invention is further elaborated below.The present invention is based on the target ranging method of binocular synthetic aperture focusing image, as shown in Figure 2, specifically comprise the following steps:
Step 1, from the identical face battle array of two structures, using the symcenter of described two face battle arrays as benchmark seat in the plane, respectively optical synthesis aperture imaging is carried out to target place scene, obtain the synthetic aperture image of one group of corresponding different focusing distance respectively.
Carry out optical synthesis aperture imaging respectively from two face battle arrays, binocular synthetic aperture image can be obtained, be i.e. the synthetic aperture image of two groups of corresponding different focusing distances.Synthetic aperture focusing imaging requirements is to the clear focusing of specific range, and target astigmatism Jiao of other distances is the image patch of uniform smooth, and for avoiding barrel shape distortion correction, to realize real-time measurement, the present invention preferably adopts the rounded face battle array that two structures are identical; Described optical synthesis aperture imaging takes realization by the national forest park in Xiaokeng video camera array gathered in rounded face battle array simultaneously, or is realized by a series of densely arranged seat in the plane timesharing shooting in rounded face battle array of single national forest park in Xiaokeng video camera.As shown in Figure 3, face, left and right battle array is gathered in two great circles, makes pupil circular, to reduce length of synthetic aperture.When barrel shape distortion is inevitable, can adopt interpolation, split camera review picture to reduce the method correction barrel shape distortions such as visual field cone angle, this process synchronously can be carried out in conjunction with super-resolution reconstruction.Each video camera imaging of video camera battle array meets national forest park in Xiaokeng, and the two sides battle array symcenter of binocular synthetic aperture imaging is benchmark seat in the plane.Synthetic aperture focusing stereoscopic vision requires the multiple images obtaining the different seat in the plane of same target, has stronger adaptive faculty to illumination, video camera difference, noise etc.Image sequence can, from video camera array, also can be same camera shifting timesharing shooting.Moving video image sequence can be combined into binocular synthetic aperture focusing image sequence, in fact aerial image sequence process, and binocular synthetic aperture principle also can be adopted to obtain depth information.In addition, it is important to note that some researchers propose image super-resolution rebuilding technology to combine with optical synthesis aperture imaging technique, to break through the restriction of image capture device intrinsic resolution, more high resolution synthetic aperture image is obtained.Also can use in this step and existingly maybe the various optical synthesis aperture formation methods based on Super-Resolution had be carried out optical synthesis aperture imaging respectively to target place scene, the super-resolution synthetic aperture focusing image obtained can further improve the precision of object ranging.
The process obtaining binocular synthetic aperture image is described for the sequence of video images of vertical target at the uniform velocity follow shot below:
Step 1-1, collection binocular camera array image:
Fig. 4 is the 60 width image sequences at the uniform velocity moving seat in the plane shooting, and video camera for shooting adopts the aperture camera lens of the super depth of field, and simultaneously image sequence is obtained (certainly, camera array also can be adopted to take) by single camera timesharing mobile collection.Image sequence is divided into two groups by seat in the plane proximity, often organizes 30 width composition left and right arrays.
Step 1-2, calculate the parallaxial displacement of each focusing distance hypograph and norm force machine interdigit, superpose after each focusing distance hypograph of left and right array corrects parallax respectively.
Can draw the parallax of video camera seat in the plane and benchmark seat in the plane according to camera parameters, target range and Fig. 3 geometric relationship, benchmark seat in the plane is positioned at the symcenter of Fig. 3 picture, i.e. arrow instruction place in figure.Not necessarily will gather image in benchmark seat in the plane, this does not affect binocular system of battle formations picture with this seat in the plane for reference corrects parallax.The each imaging importing of left side battle array after high-ranking officers' positive parallax is left order synthetic aperture image, the each imaging importing of right side battle array after high-ranking officers' positive parallax is right order synthetic aperture image, correcting parallax displacement is sub-pixel, synthetic aperture image focu can be made in any distance, and namely imaging and focusing face can movement arbitrarily.Each group of video camera battle array longest diameter length is length of synthetic aperture.Like this, for left and right array, the synthetic aperture image of one group of corresponding different focusing distance both can be obtained respectively.
Fig. 5 shows obtained part synthetic aperture image, wherein (a) focuses on vehicle forward position place for left order, b () be relative distant view place for left order focuses on, (c) focuses on vehicle forward position place for right order, and (d) be relative distant view place for right order focuses on.Because synthetic aperture diameter is more much bigger than optical lens, so small target range difference, significantly poly-/defocusing effect also can be caused.As seen from Figure 5, about specified distance target, battle array focuses on image position coincidence, and other distances not only target become defocused image, and position offsets.Image sequence synthesized image has obvious rejection ability to differences such as single image illumination, DE Camera Shake, noises.Resultant error makes synthesized image compared with single width picture, and sharpness obviously declines.And the range finding of binocular synthetic aperture focusing can suppress common mode resultant error, therefore allow synthesized image to focus on image sharpness because error causes and decline to some extent.
Step 2, to each focusing distance, using the two width synthetic aperture images obtained from two face battle arrays corresponding to it as a pair, then following process is carried out to each Technologies Against Synthetic Aperture image: the target image in wherein piece image is as with reference to image, using region corresponding with reference picture in another piece image as matched position image, and by the mould of the aberration vector in another piece image and between matched position image (0, d] the interior a series of regions changed of scope are as non-matching location drawing picture, and d is default maximum aberration; Compatible quantity of information R between computing reference image and matched position image
0and the compatible quantity of information between reference picture and each non-matching location drawing picture, and be greater than R by the compatible quantity of information between calculated reference picture and each non-matching location drawing picture
0primary system one be revised as R
0, then calculate the average of the compatible quantity of information between amended reference picture and each non-matching location drawing picture; Last with R
0with the difference of the described average detection limit as this focusing distance.
For a certain focusing distance z, judge that target similarly is no focusing, whether the compatible quantity of information not only will investigated in left and right order synthetic aperture image between target picture reaches maximum, also to investigate left and right target image position whether to overlap at ideal position, the left and right target image position of target, setting focusing distance place overlaps, other focusing distance place targets become defocused image, and position offsets.Therefore to the left and right target picture of each focusing distance, need search for and calculate in certain aberration distance range (owing to adopting linear array synthetic aperture imaging in the present embodiment, only need at [-d, d] aberration within the scope of search, d is default maximum aberration, preferably changes by pixel within the scope of this during search; If adopt preferred rounded face battle array, then needing at the mould of aberration vector is search in [0, d] scope) compatible quantity of information between target picture, if compatible quantity of information maximal value appears at imperfect position, then show this region defocused image interference range finding, this need be suppressed to disturb.The compatible quantity of information of binocular target image that Fig. 6 shows different target is with the change profile situation of focusing distance and parallax, and the venter of relicle in (a) is for needing point distance measurement, and be on building edge line, its image entropy is larger; B () is the compatible quantity of information R between this vertex neighborhood left and right order synthetic aperture image
zdthe two-dimensional distribution that (x, y) changes with focusing distance z and coupling aberration, brightness and the positive correlation of compatible quantity of information size in figure; The level of figure (b) is to being focusing distance parameter z, and from left to right, focusing distance parameter changes to 5 meters from 15 meters, and figure (b) Vertical dimension is aberration, and aberration parameter changes to+20 pixels from-20 pixels from bottom to top.Aberration 0 is appeared at just, herein the actual distance of distance parameter target just, see (a) in Fig. 7 by the visible compatible quantity of information of (b) in Fig. 6 is maximum.Venter of relicle place in the figure (c) of Fig. 6 is changed to when needing point distance measurement, its image entropy is less, and neighbouring building edge image entropy is larger, it can disturb range finding, the two-dimensional distribution that figure (d) in Fig. 6 changes with focusing distance and coupling aberration for the compatible quantity of information between this vertex neighborhood left and right order synthetic aperture image, visible compatible quantity of information maximal value appears at aberration-13 pixel place, distance parameter is herein not the actual distance of target, see the figure (b) in Fig. 7.For this reason, the present invention has carried out peak clipping process to the compatible quantity of information between calculated reference picture and each non-matching location drawing picture, is about to be greater than R in the compatible quantity of information between calculated reference picture and each non-matching location drawing picture
z0the R of (x, y)
zd(x, y) is unified reduces to R
z0(x, y), inhibit other target defocused image to disturb to a certain extent, through this process, the compatible quantity of information between impact point neighborhood left and right order synthetic aperture image with focusing distance and coupling aberration change Two dimensional Distribution as shown in the figure (e) in Fig. 6.
Suppressing in defocused image basis of signals through peak clipping process, calculate the average of the compatible quantity of information between the reference picture after peak clipping process and each non-matching location drawing picture
obtain a stable compatible quantity of information reference point, with reference to the difference of the compatible quantity of information between image and matched position image and this reference point
as detection limit, show two order visual image couplings and non-matching diversity factor, large as this diversity factor to focusing, little to this diversity factor of defocused image.This detection limit does not affect by left and right synthetic aperture focusing common-mode error, and this detection limit maximal value correspond to target range.For the target in the figure (a) of Fig. 6, detection limit is with distance change curve as shown in the figure (a) in Fig. 7, and visible detection amount maximal value is in 13 meters of.For the target in the figure (c) of Fig. 6, compatible quantity of information between its reference picture and matched position image with distance change curve as shown in the figure (b) in Fig. 7, the corresponding wrong distance parameter of its maximal value, and with the difference of the compatible quantity of information in binocular ranging position quantity of information average compatible with non-matching position
as detection limit, correctly can detect that target is in correct distance 13 meters of, as shown in the figure (c) in Fig. 7.Therefore detection limit proposed by the invention can be used as the accurate measurement index of object ranging.
In addition, in order to the range finding mistake preventing target from causing as information deficiency, the present invention establishes a threshold value to maximal compatible quantity of information, is less than thresholding and shows that this target picture lacks the information such as gray scale texture, can not find range; Be greater than the range finding of thresholding for effectively range finding.Specific as follows:
Step 3, find out the maximum focusing distance of detection limit, and judge the R corresponding to this focusing distance
0whether be greater than default compatible quantity of information threshold value, in this way, then using this focusing distance as target range.
Fig. 8 shows the final target depth figure obtained, and in figure, left side is for brightness distance is according to bar, and pitch black place is can not point distance measurement.Adding street model and auto model prior imformation, can analyze each track on road and take situation, cook up automatic driving vehicle and can use track, is automatic driving vehicle navigation.Complete stereoscopic vision also comprises the level and smooth of three-dimensional modeling and the work that becomes more meticulous, but can omit in automatic driving vehicle navigation.
Claims (10)
1. based on a target ranging method for binocular synthetic aperture focusing image, it is characterized in that, comprise the following steps:
Step 1, from the identical face battle array of two structures, using the symcenter of described two face battle arrays as benchmark seat in the plane, respectively optical synthesis aperture imaging is carried out to target place scene, obtain the synthetic aperture image of one group of corresponding different focusing distance respectively;
Step 2, to each focusing distance, using the two width synthetic aperture images obtained from two face battle arrays corresponding to it as a pair, then following process is carried out to each Technologies Against Synthetic Aperture image: the target image in wherein piece image is as with reference to image, using region corresponding with reference picture in another piece image as matched position image, and by the mould of the aberration vector in another piece image and between matched position image (0, d] the interior a series of regions changed of scope are as non-matching location drawing picture, and d is default maximum aberration; Compatible quantity of information R between computing reference image and matched position image
0and the compatible quantity of information between reference picture and each non-matching location drawing picture, and be greater than R by the compatible quantity of information between calculated reference picture and each non-matching location drawing picture
0primary system one be revised as R
0, then calculate the average of the compatible quantity of information between amended reference picture and each non-matching location drawing picture; Last with R
0with the difference of the described average detection limit as this focusing distance; Described compatible quantity of information for measure two width images information in the amount of same section;
Step 3, find out the maximum focusing distance of detection limit, and judge the R corresponding to this focusing distance
0whether be greater than default compatible quantity of information threshold value, in this way, then using this focusing distance as target range.
2. target ranging method as claimed in claim 1, it is characterized in that, described battle array is rounded face battle array; Described optical synthesis aperture imaging takes realization by the national forest park in Xiaokeng video camera array gathered in rounded face battle array simultaneously, or is realized by a series of densely arranged seat in the plane timesharing shooting in rounded face battle array of single national forest park in Xiaokeng video camera.
3. target ranging method as claimed in claim 1, it is characterized in that, for gray level image, the metric form of described compatible quantity of information is the product of similarity between the information entropy of two width images and two width images, or be the product of similarity between the average information entropy of two width images and two width images, or be the acutance average of two width images or gradient mean value or the product of similarity between contrast average and two width images, or be the cross-entropy of two width images; For coloured image, the metric form of described compatible quantity of information be adopt the compatible informational content measure form of gray level image to obtain two width images in RGB color space between each passage subimage compatible quantity of information and or average.
4. target ranging method as claimed in claim 1, is characterized in that, use the optical synthesis aperture formation method based on Super-Resolution to carry out optical synthesis aperture imaging respectively to target place scene.
5. target ranging method as claimed in claim 1, is characterized in that, the mould of described aberration vector (0, d] change by pixel in scope.
6., based on an object ranging device for binocular synthetic aperture focusing image, it is characterized in that, comprising:
Binocular synthetic aperture imaging unit, for from identical two the face battle arrays of structure, using the symcenter of described two face battle arrays as benchmark seat in the plane, respectively optical synthesis aperture imaging is carried out to target place scene, obtain the synthetic aperture image of one group of corresponding different focusing distance respectively;
Distance detecting unit, the synthetic aperture image of two groups for obtaining according to binocular synthetic aperture imaging unit corresponding different focusing distances detects target range, its detection method is specific as follows: to each focusing distance, using the two width synthetic aperture images obtained from two face battle arrays corresponding to it as a pair, then following process is carried out to each Technologies Against Synthetic Aperture image: the target image in wherein piece image is as with reference to image, using region corresponding with reference picture in another piece image as matched position image, and by the mould of the aberration vector in another piece image and between matched position image (0, d] a series of regions of change in scope are as non-matching location drawing picture, d is default maximum aberration, compatible quantity of information R between computing reference image and matched position image
0and the compatible quantity of information between reference picture and each non-matching location drawing picture, and be greater than R by the compatible quantity of information between calculated reference picture and each non-matching location drawing picture
0primary system one be revised as R
0, then calculate the average of the compatible quantity of information between amended reference picture and each non-matching location drawing picture, last with R
0with the difference of the described average detection limit as this focusing distance, described compatible quantity of information for measure two width images information in the amount of same section, find out the focusing distance that detection limit is maximum, and judge the R corresponding to this focusing distance
0whether be greater than default compatible quantity of information threshold value, in this way, then using this focusing distance as target range.
7. object ranging device as claimed in claim 6, it is characterized in that, described battle array is rounded face battle array; Described optical synthesis aperture imaging takes realization by the national forest park in Xiaokeng video camera array gathered in rounded face battle array simultaneously, or is realized by a series of densely arranged seat in the plane timesharing shooting in rounded face battle array of single national forest park in Xiaokeng video camera.
8. object ranging device as claimed in claim 6, it is characterized in that, for gray level image, the metric form of described compatible quantity of information is the product of similarity between the information entropy of two width images and two width images, or be the product of similarity between the average information entropy of two width images and two width images, or be the acutance average of two width images or gradient mean value or the product of similarity between contrast average and two width images, or be the cross-entropy of two width images; For coloured image, the metric form of described compatible quantity of information be adopt the compatible informational content measure form of gray level image to obtain two width images in RGB color space between each passage subimage compatible quantity of information and or average.
9. object ranging device as claimed in claim 6, is characterized in that, use the optical synthesis aperture formation method based on Super-Resolution to carry out optical synthesis aperture imaging respectively to target place scene.
10., as claim 6 object ranging device, it is characterized in that, the mould of described aberration vector (0, d] change by pixel in scope.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510732130.9A CN105258673B (en) | 2015-11-02 | 2015-11-02 | A kind of target ranging method based on binocular synthetic aperture focusing image, device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510732130.9A CN105258673B (en) | 2015-11-02 | 2015-11-02 | A kind of target ranging method based on binocular synthetic aperture focusing image, device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105258673A true CN105258673A (en) | 2016-01-20 |
| CN105258673B CN105258673B (en) | 2017-05-31 |
Family
ID=55098458
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510732130.9A Active CN105258673B (en) | 2015-11-02 | 2015-11-02 | A kind of target ranging method based on binocular synthetic aperture focusing image, device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105258673B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107716468A (en) * | 2017-10-31 | 2018-02-23 | 广东工业大学 | A kind of laser auto focusing method, system, device and readable storage medium storing program for executing |
| CN107820093A (en) * | 2017-11-15 | 2018-03-20 | 深圳大学 | Information detecting method, device and receiving device based on packet energy differences |
| CN108226901A (en) * | 2018-02-06 | 2018-06-29 | 北京万集科技股份有限公司 | Laser radar optical system |
| CN110749897A (en) * | 2018-07-23 | 2020-02-04 | Ibeo汽车系统有限公司 | Method and device for optical distance measurement |
| CN110910436A (en) * | 2019-10-30 | 2020-03-24 | 深圳供电局有限公司 | Distance measuring method, device, equipment and medium based on image information enhancement technology |
| CN113391644A (en) * | 2021-06-11 | 2021-09-14 | 中新国际联合研究院 | Unmanned aerial vehicle shooting distance semi-automatic optimization method based on image information entropy |
| CN116228547A (en) * | 2023-05-08 | 2023-06-06 | 天府兴隆湖实验室 | Simulation algorithm of synthetic aperture imaging process |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102867304A (en) * | 2012-09-04 | 2013-01-09 | 南京航空航天大学 | Method for establishing relation between scene stereoscopic depth and vision difference in binocular stereoscopic vision system |
| CN103026171A (en) * | 2011-05-27 | 2013-04-03 | 松下电器产业株式会社 | Image processing apparatus and image processing method |
| CN103033166A (en) * | 2012-12-13 | 2013-04-10 | 南京航空航天大学 | Target ranging method based on synthetic aperture focused images |
| US20140071427A1 (en) * | 2012-09-07 | 2014-03-13 | Apple Inc. | Imaging range finder fabrication |
-
2015
- 2015-11-02 CN CN201510732130.9A patent/CN105258673B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103026171A (en) * | 2011-05-27 | 2013-04-03 | 松下电器产业株式会社 | Image processing apparatus and image processing method |
| CN102867304A (en) * | 2012-09-04 | 2013-01-09 | 南京航空航天大学 | Method for establishing relation between scene stereoscopic depth and vision difference in binocular stereoscopic vision system |
| US20140071427A1 (en) * | 2012-09-07 | 2014-03-13 | Apple Inc. | Imaging range finder fabrication |
| CN103033166A (en) * | 2012-12-13 | 2013-04-10 | 南京航空航天大学 | Target ranging method based on synthetic aperture focused images |
Non-Patent Citations (1)
| Title |
|---|
| 周致远: "接近段目标双目视觉测量中的精度分析与像机标定方法研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107716468A (en) * | 2017-10-31 | 2018-02-23 | 广东工业大学 | A kind of laser auto focusing method, system, device and readable storage medium storing program for executing |
| CN107820093A (en) * | 2017-11-15 | 2018-03-20 | 深圳大学 | Information detecting method, device and receiving device based on packet energy differences |
| CN107820093B (en) * | 2017-11-15 | 2019-09-03 | 深圳大学 | Information detecting method, device and receiving device based on grouping energy differences |
| CN108226901A (en) * | 2018-02-06 | 2018-06-29 | 北京万集科技股份有限公司 | Laser radar optical system |
| CN108226901B (en) * | 2018-02-06 | 2024-03-01 | 武汉万集光电技术有限公司 | Laser radar optical system |
| CN110749897B (en) * | 2018-07-23 | 2023-06-23 | 微视公司 | Method and device for optical distance measurement |
| CN110749897A (en) * | 2018-07-23 | 2020-02-04 | Ibeo汽车系统有限公司 | Method and device for optical distance measurement |
| CN110910436A (en) * | 2019-10-30 | 2020-03-24 | 深圳供电局有限公司 | Distance measuring method, device, equipment and medium based on image information enhancement technology |
| CN110910436B (en) * | 2019-10-30 | 2022-10-28 | 深圳供电局有限公司 | Distance measuring method, device, equipment and medium based on image information enhancement technology |
| CN113391644A (en) * | 2021-06-11 | 2021-09-14 | 中新国际联合研究院 | Unmanned aerial vehicle shooting distance semi-automatic optimization method based on image information entropy |
| CN113391644B (en) * | 2021-06-11 | 2022-06-17 | 中新国际联合研究院 | Unmanned aerial vehicle shooting distance semi-automatic optimization method based on image information entropy |
| CN116228547A (en) * | 2023-05-08 | 2023-06-06 | 天府兴隆湖实验室 | Simulation algorithm of synthetic aperture imaging process |
| CN116228547B (en) * | 2023-05-08 | 2023-07-07 | 天府兴隆湖实验室 | Simulation method of synthetic aperture imaging process |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105258673B (en) | 2017-05-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105258673B (en) | A kind of target ranging method based on binocular synthetic aperture focusing image, device | |
| CN106091984B (en) | A kind of three dimensional point cloud acquisition methods based on line laser | |
| CN103776419B (en) | A kind of binocular distance measurement method improving measurement range | |
| CN103279982B (en) | The speckle three-dimensional rebuilding method of the quick high depth resolution of robust | |
| CN103150715B (en) | Image mosaic processing method and processing device | |
| CN102314683B (en) | Computational imaging method and imaging system based on nonplanar image sensor | |
| CN102831601A (en) | Three-dimensional matching method based on union similarity measure and self-adaptive support weighting | |
| CN107635129B (en) | Three-dimensional trinocular camera device and depth fusion method | |
| CN103868460A (en) | Parallax optimization algorithm-based binocular stereo vision automatic measurement method | |
| CN107167093B (en) | A kind of the combined type measuring system and measurement method of laser line scanning and shadow Moire | |
| CN104165626B (en) | Bionic compound eyes imageable target positioning system | |
| CN106296811A (en) | A kind of object three-dimensional reconstruction method based on single light-field camera | |
| CN109089025A (en) | A kind of image instrument digital focus method based on optical field imaging technology | |
| CN109712232B (en) | Object surface contour three-dimensional imaging method based on light field | |
| CN105282443A (en) | Method for imaging full-field-depth panoramic image | |
| CN107154014A (en) | A kind of real-time color and depth Panorama Mosaic method | |
| CN106534670B (en) | It is a kind of based on the panoramic video generation method for connecting firmly fish eye lens video camera group | |
| CN102665086A (en) | Method for obtaining parallax by using region-based local stereo matching | |
| CN111563952B (en) | Method and system for realizing stereo matching based on phase information and spatial texture characteristics | |
| CN103299343A (en) | Range image pixel matching method | |
| CN106170086B (en) | Method and device thereof, the system of drawing three-dimensional image | |
| CN110288659A (en) | A kind of Depth Imaging and information acquisition method based on binocular vision | |
| CN103033166B (en) | Target ranging method based on synthetic aperture focused images | |
| CN109883391A (en) | Monocular distance measuring method based on microlens array digital imagery | |
| CN106524995B (en) | Detect the localization method of target object space length in real time based on visible images |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |