CN103955710B - Method for monocular vision space recognition in quasi-earth gravitational field environment - Google Patents
Method for monocular vision space recognition in quasi-earth gravitational field environment Download PDFInfo
- Publication number
- CN103955710B CN103955710B CN201410212438.6A CN201410212438A CN103955710B CN 103955710 B CN103955710 B CN 103955710B CN 201410212438 A CN201410212438 A CN 201410212438A CN 103955710 B CN103955710 B CN 103955710B
- Authority
- CN
- China
- Prior art keywords
- segment
- pixel
- sky
- image
- ground
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses a method for monocular vision space recognition in a quasi-earth gravitational field environment. The method for monocular vision space recognition in the quasi-earth gravitational field environment is characterized by comprising the following steps that firstly, superpixel segmentation is conducted on an image based on the CIELAB color space values L, a and b of a pixel, the x coordinate value and the y coordinate value, so that a superpixel image is generated; secondary, dimensionality reduction is conducted on the superpixel image obtained through segmentation and a large image block is formed according to a clustering algorithm based on the superpixel color characteristic, the textural feature vector distance and the adjacency relation; thirdly, the pixel of the obtained large image block is multiplied by a gravitational field fuzzy distribution density function which represents the sky, the ground and the stereo object, then the value of expectation of the large image block is obtained, and preliminary classification of the sky, the ground and a stereo object is completed; fourthly, a classification map of the sky, the ground and the stereo object is extracted through single-layer wavelet sampling and the Manhattan direction characteristic; finally, a space depth perception image is generated based on a small aperture imaging model and ground linear perspective information. The method for monocular vision space recognition in the quasi-earth gravitational field environment is easy to implement, the resolution ratio is high, and the application range is wide.
Description
Technical field
The present invention relates to a kind of image processing method, especially one kind can be widely applied to as robot visual guidance,
The image processing method that space identity can be improved in the fields such as the target measurement of large space, target tracking and positioning, specifically
It is monocular vision space recognition method under a species ground gravitational field environment.
Background technology
Understand that 3d space structure, as the basic problem of machine vision, is paid close attention to by people for a long time and studied, in early days
The research work clue paying close attention to stereoscopic vision or obtain 3D by the motion at visual angle.In recent years, many researchers will
Focus focuses on reconstruct 3d space structure from monocular vision image, and most of monocular vision 3d space recognition methods are many at present
Using supervision type machine learning method, such as：Markov random field (MRFs), conditional probability random field (CRFs) and dynamic
Bayesian network (DBN) etc..However, these methods frequently rely on its priori, that is, it is only capable of learning training concentration and is gathered
Image-context.Therefore, when sample devices or sampling environment change, the result of monocular vision 3d space identification will produce
Larger difference.In order to solve this problem, the present invention proposes gravitational field factor is added in graphical analysis, constructs a kind of new
Unsupervised learning monocular space recognition method.
Content of the invention
The purpose of the present invention is to be required for greatly by just completing to the study of image for existing image-recognizing method,
These methods have that data processing amount is big, speed slow, bad adaptability, more problem limited by range, and invention one kind need not
Study and can quickly identify, monocular vision space recognition method under the degree of accuracy simultaneously is high, applicability is wide class ground gravitational field environment.
The technical scheme is that：
Monocular vision space recognition method under one species ground gravitational field environment, is characterized in that it comprises the following steps：
First, image is carried out with CIELAB color space values L, a, b and the x based on pixel, the super-pixel of y-coordinate value is divided
Cut, to produce the super-pixel image with certain density；
Secondly, using the general clustering algorithm based on super-pixel color characteristics, texture feature vector distance and syntople, will
The super-pixel image dimensionality reduction of segmentation formation simultaneously generates big segment；
3rd, will be big with gained respectively for the gravitational field Fuzzy Distribution density function representing sky, ground and facade object
Segment pixel is multiplied, and obtains the desired value of big segment, thus completing the preliminary classification of sky, ground and facade object；
4th, the classification of sky, ground and facade object is gone out by the sampling of individual layer small echo and Manhattan Directional feature extraction
Figure；
Finally, based on the linear perspective information of national forest park in Xiaokeng and ground generate spatial depth perceptual map, thus complete by
The plane picture that picture pick-up device obtains, to the conversion of stereo-picture, realizes monocular vision space identity under class ground gravitational field environment.
The invention has the beneficial effects as follows：
Present invention firstly provides being added to gravitational field factor in graphical analysis, construct a kind of new unsupervised learning
Monocular space recognition method, simulates human visual system to ground continuous surface integrated processing method, builds one and have one
Determine monocular vision space identity pattern under universality class ground gravitational field environment, it changes traditional single camera vision system 3D reconstruct
And the algorithm pattern of depth perception.
1., invention emulates human visual system, build a class ground gravitational field environment with certain universality and place an order
Visually feel space recognition method it is pointed out that the method can apply to the classes such as martian surface and moonscape ground gravity
Visual space measurement under the environment of field, as shown in figure 15.
2., in cancelling to image during the constraints of sky brightness, the present invention can also identify urban landscape environment, such as
Shown in Figure 16.
3. the present invention need not carry out the study of priori to computer and training just can be under class ground gravitational field environment
Monocular vision image effectively identified and 3D reconstruct.
4. the present invention changes the algorithm pattern of traditional single camera vision system 3D reconstruct and depth perception, can extensively answer
For fields such as such as robot visual guidance, the target measurement of large space, target tracking and positioning.
Brief description
Fig. 1 is the schematic flow sheet of the present invention.
Fig. 2 is the general cluster process based on super-pixel and the effect diagram of the present invention.In Fig. 2：A () is original image, (b)
For 951 super-pixel segmentation images, the image of (c) 145 spectral clusterings, 92 of (d) 3 iteration convergences clusters images.
Fig. 3 be the present invention utilization geometry inclusion relation eliminate segment in isolated island process schematic.In Fig. 3：A () is to build
Isolated island segment, (b) is stayed to be the knot eliminating segment isolated island based on geometry inclusion relation clustering algorithm after building thing window clustering algorithm
Really.
Fig. 4 is mankind's gravitational field visual cognition model schematic.
Fig. 5 is the determination schematic diagram of the eye-level display position of the present invention.
Fig. 6 is that the image eye-level display position of the present invention determines equivalent schematic diagram, in figure：HIFor the height of image, HI=HS+
HG.
Fig. 7 is to divide through the ground based on gravitational field Fuzzy Distribution density function gained for the present invention, sky, facade object
Class process schematic diagram.
Fig. 8 is facade object and the sky sorting algorithm process schematic of the present invention.
Fig. 9 be the present invention gravitational field ambiguity function judge in there is not meeting the result schematic diagram of gravitational field.In figure (a)
It is to distinguish the result after calculating through facade object and ground for artwork, (b).
Figure 10 is the ambiguity function and facade object and sky sorted result of calculation schematic diagram through the present invention.Wherein
A () is that gravitational field segment reclassifies, (b) is the cluster result after facade object and ground differentiation to not meeting.
Figure 11 is the result of output after facade object of the present invention is further discriminated between with ground.
Figure 12 is the Vision imaging system physical model schematic diagram of the present invention.
Figure 13 is the mapping schematic diagram in Lab space for the depth projection angle of the present invention.
Figure 14 is the depth perception figure of corresponding Figure 11.
Figure 15 is the result signal that using the method for the present invention, NASA Mars picture is carried out with space identity and depth recognition
Figure.
Figure 16 is to the space identity of urban landscape picture and 3D reconstruct image using the method for the present invention.
Specific embodiment
The present invention is further illustrated for constructive embodiment and accompanying drawing below.
As represented in figures 1 through 14.
Monocular vision space recognition method under one species ground gravitational field environment, it comprises the following steps：
(1) first image is carried out with the super-pixel image segmentation based on pixel color and locus, is formed and have necessarily
The super-pixel image of density；
(2) pass through with general based on super-pixel color space distance, texture feature vector distance and geometry syntople
Clustering algorithm is by the big segment dendrogram picture of super-pixel image dimensionality reduction to less than 10%；
(3) represent the gravitational field Fuzzy Distribution density function of sky, ground and facade object respectively with these big segment pictures
Element is multiplied, and obtains the desired value of these big segments, thus producing the preliminary classification of sky, ground and facade object, by entering
The property sort algorithms such as one layer of small echo sampling of one step, the extraction of Manhattan direction, extract accurate sky, ground and stand
Face object classification figure；
(4) last, spatial depth perceptual map is generated based on the linear perspective information of national forest park in Xiaokeng and ground.
Details are as follows：
1. super-pixel clustering algorithm.
The simple linear Iterative Clustering that can be proposed using Achanta R is SLIC (Simple Linear
Iterative Clustering), this algorithm is with L, a, b value of CIELAB color space of pixel and the x of pixel, y-axis coordinate
Build 5 dimension spaces, and define normalized apart from Ds, be defined as follows：
Wherein：Ck=[lk,ak,bk,xk,yk]TCenter for cluster；[li,ai,bi,xi,yi]T5 dimensions for image slices vegetarian refreshments
Space coordinates；N is the pixel count of image；K is the number of the super-pixel that expectation obtains；S is super-pixel center grates spacing；DsFor
Color lab is apart from dlabAnd dxyStandardization distance based on S；M is controlled super-pixel density factor.
2. the general clustering algorithm based on super-pixel.
(1) using n super-pixel produced by SLIC algorithm as undirected weights figure G summit V={ v1,v2,…,vn}；
(2) adjacency matrix builds, i=1,2 ... n；J=1,2 ... n, wherein, n is the number of super-pixel；
(3) structure of weights adjacency matrix, i=1,2 ... n；J=1,2 ... n；
Wherein weight w (i, j) is the standardization CIELAB color histogram calculating between two neighboring super-pixel
Bhattacharyya coefficient, concrete construction method is that the color space of image is transformed into CIELab space, and by L * channel
Span is divided into 8 grades of deciles, and the span of a passage is divided into 16 deciles, the span of b passage is divided into 16
Grade, the purpose wherein L * channel span being divided into 8 grades is the disturbance reducing chroma-luminance change to weights.Each surpasses
Pixel calculates value histogram in the space of 8 × 16 × 16=2048 dimensionWherein l
=2048, then work as Ei,jWhen=1
For the value of weight w (i, j), when being embodied as, color distance and texture energy can be based respectively on by increasing by two
Span from constraints, be described below：
1. it is based on color distance constraints：W (i, j)≤W at that timeT, then w (i, j)=0, wherein W are takenTSpan be
Between (0.7～1.0)；
2. the constraints of texture energy distance：Using l2The average energy of norm calculation each super-pixel segment is estimated, that is,
Wherein R (i, j) is the small echo sampled value at (i, j) point in image, calculates each super-pixel block according to formula (8)
Four-dimensional wavelet-based attribute vector, i.e. e (i)=(ei(LL), ei(LH), ei(HL), ei(HH)), and ask between its neighbouring super pixels
Bhattacharyya coefficient value Be(i,j).
Wherein,
Work as Be(i,j)≤BTWhen, then take w (i, j)=0, wherein BTSpan between (0.85～1.0)；
The purpose of two above constraints is to improve the color of neighbouring super pixels point and the similarity threshold of texture, to protect
Boundary characteristic between shield sky and facade object, between facade object and ground.
(4) structure of matrix, i=1,2 ... n are spent；J=1,2 ... n；
(5) structure of standardization Laplacian matrix
Normalized Laplacian matrix is calculated using Normalized-cut criterion：
Lsym=I-D-1/2WD-1/2(12)
Wherein：D is degree matrix, and W is weights adjacency matrix.
(6) calculate LsymCarry out Eigenvalues Decomposition, and take the characteristic vector corresponding to front K minimal eigenvalue, V1, V2...,
Vk；, wherein K=[0.1 × n], that is, take the 10% of n as image clustering characteristic vector dimension, to realize dimensionality reduction purpose；
(7) by V1, V2..., VkRearrange Rn×kEach element in matrix is simultaneously taken absolute value to obtain matrix U by matrix；
(8) for i=1,2 ... n, make yi∈RkThe i-th row vector for matrix U；
(9) y to non-zeroi∈RkVector is normalized, and is clustered with Bhattacharyya Y-factor method Y, wherein
Bhattacharyya is apart from BUThreshold value be (0.85～1.0) between, that is, work as BUDuring more than or equal to threshold value, gathered between super-pixel
Class；
(10) segment code requirement CIELAB color histogram is clustered to each, and formula (7) is adopted to adjacent segment class
Carry out Bhattacharyya apart from w (i, j), adopt formula (9) to calculate the B of adjacent segment simultaneouslye(i, j), as w (i, j) >=WTAnd
Be(i,j)≥BTShi Jinhang clusters；
(11) repeat (10) step, until convergence.
This algorithm is as shown in Figure 2 to the cluster process of Make3D Image data image library image and effect.
3. geometry inclusion relation clustering algorithm.
In order to improve the degree of accuracy that Fuzzy Distribution density function judges to sky, ground and facade object, need to segment
Carry out the cluster based on geometry inclusion relation, to eliminate isolated island segment, so-called isolated island segment refers to one or more segment quilts
The segment (as shown in Figure 3) of one big segment encirclement completely, isolated island segment can be clustered by the clustering algorithm of geometry inclusion relation
Become to surround the big segment of this isolated island completely, thus avoiding geometry contextual algorithms to strange produced by isolated island segment spatial classification
Different.
Specific algorithm is as follows：
(1) find hollow out segment, its criterion is to work as Nb-nb>When 0, then segment is hollow out segment, and wherein Nb is all sides of segment
The pixel value on boundary, nbFor the pixel value of segment external boundary, if Nb-nb>0 entrance next step, otherwise segment is not hollow out figure
Block；
(2) segment is filled for border with the mark value of artwork block with external boundary；
(3) replace former hollow out segment to fill segment.
4. the structure of human vision cognitive model and sorting algorithm in gravitational field.
Fig. 4 is mankind's gravitational field visual cognition model.
When the eye-level display of the mankind become level or during close to level its to inference pattern of sky, ground and facade object as schemed
Shown in 4, wherein stain represents the maximum probability point position sky, ground or facade object in human vision respectively.To sky
The distributed density values of the probability-distribution function of empty reasoning from human visual field angle the maximum prolonged be gradually lowered to human visual field
Angle is the most downward, and its probability density value on eye-level display is zero；To the distributed density values of the probability-distribution function of ground reasoning from
Human visual field angle maximum the most downward be gradually lowered to human visual field angle on prolong, its its probability density value on eye-level display is
Zero；Both direction is gradually lowered maximum from eye-level display for the probability distribution density value of opposite object reasoning up and down, directly
To human visual field angle go up most with the most downward, its distributed density values is close to zero.
Inference pattern below according to above-mentioned sky, ground and facade object combines the perspective projection characteristic of image, structure
Build following gravitational field Fuzzy Distribution density function：
(1) set the position of the eye-level display of image, as shown in figure 5, eye-level display was photocentre and the ground level of video camera
Or the straight line of plane-parallel, and the horizontal line in image be eye-level display with the intersection point of video camera imaging target surface and ground level or
The straight line of plane-parallel is as shown in Figure 6.
(2) gravity field on earth's surface dimness of vision distribution density function G：
Work as HG≥HSWhen：OrderAnd
Then
H in formulaGFor the distance away from image base for the eye-level display；HSFor the distance away from image top margin for the eye-level display；X is pixel in figure
Coordinate as short transverse；N is the exponent number of density function.
Work as HG＜ HSWhen：G (x)=- S (x)
I.e.
Wherein：N=1,2,3 ... N, N ∈ positive integer, generally take n=1.
(3) sky gravitational field dimness of vision distribution density function S：
Work as HG＜ HSWhen：OrderAnd
Then
Work as HG≥HSWhen：S (x)=- G (x)
I.e.
Wherein：N=1,2,3 ... N, N ∈ positive integer, generally take n=1.
(4) facade object gravitational field dimness of vision distribution density function V：
Wherein：
(5) each pixel in cluster segment is obscured with ground Fuzzy Distribution density function G, sky in image vertical direction
Distribution density function S and facade object Fuzzy Distribution density function V, in (- HG, HS) in the range of be multiplied and seek its desired value, public
Formula is as follows：
Wherein：niFor clustering the number of pixels in the i-th row for the segment, rbFor clustering the most downward, the r of segmenttGoing up most for segment
Prolong, i ∈ (0,1 ... ..., H), H=HG+HSThen being categorized as of segment：
Fig. 7 is that this model has carried out sky, ground and facade thing to the cluster segment generating through corresponding clustering algorithm
The classification results of body.As can be seen from the figure the method can effectively distinguish sky and ground, to vertical near eye-level display
It is more accurate that face object judges, but for higher facade object and sky segment, and relatively low facade object is deposited with ground
In a certain degree of erroneous judgement it is therefore desirable to carry out two choosings between opposite object and sky and facade object and ground further
One classification.
5. in gravitational field sky and facade object vision sorter algorithm.
As previously described, because the effect of gravitational field makes the material on earth surface carry out stratification by its density dividing
Cloth, i.e. the facade object on ground is stood in the high solid matter formation of density, and low-density gas is (such as：Air and cloud)
Material forms sky, and therefore under the irradiation of light, the facade object of solid-state and sky present diverse reflecting effect,
Show distinct textural characteristics in the picture.
In the research to sky feature, we are to the different objects of image (such as：Sky, roof, wall, ground meadow
Deng) carry out 1 layer of wavelet transformation sampling, and adopt l2The average energy of each segment of norm calculation is estimated, that is,
Wherein：NpFor segment number of pixels, rbThe most downward for segment, rtFor segment on prolong, clFor the i-th row segment
Far Left, crFor the i-th row segment rightmost, wherein R (i, j) is the small echo sampled value at (i, j) point in image, noticeable
It is to need removal to answer energy produced by segment edge when each segment average energy of calculating is estimated.
Calculate the four-dimensional wavelet-based attribute vector that can obtain segment, i.e. (e by energy normLL,eLH,eHL,eHH), wherein
eLLThat characterize is segment overall brightness characteristic, eLH,eHL,eHHCharacterize is segment high frequency texture feature, and daytime, outdoor sky existed
Characteristic common manifestation in image is on high brightness and low-yield high frequency texture feature.
According to above-mentioned analysis, the vision sorter algorithm of following sky and facade object is proposed：
(1) if eLL>mean(eLL1, eLL2... eLLn) it is then candidate's sky segment, wherein eLL1, eLL2... eLLn∈ sky
E with facade objectLLValue, wherein：Mean () is mean value function；
(2) under meeting above-mentioned condition, when the energy norm of one layer of un-downsampling wavelet transform of segment
When, then segment is candidate's sky segment, during without meeting this condition segment, then judges segment not as sky segment, Ec's
Span is between (0～7)；
(3) when presence meets above-mentioned segment, then determine whether the segment as border is prolonged with image, if there are then sentencing
Disconnected have sky segment, otherwise no sky in process decision chart picture；
(4) under meeting above-mentioned condition if there is candidate's sky segment not unique, then choose area maximum segment be
Sky segment, and with color distance value dabAnd brightness distance value dLFor criterion, sky is clustered, formula is as follows：
And
Wherein as、bsIt is respectively the average of sky segment CIELAB color space a, b color channel, ai、biIt is respectively candidate
The average of sky segment CIELAB color space a, b color channel, as candidate sky segment dab≤ C and dL≤ L is then sky, no
It is then facade object, wherein, the span (0～30) of C, the span (0～70) of L.
(5) if the number that the sky area that cluster is generated is calculated its pixel is less than the 2 ‰ of image pixel, will
It is classified as facade object, and its reason is that the sky segment of very little has little significance to image space identification；
(6) all non-sky segments are classified as facade object.
Through facade object with sky sorting algorithm acquired results as shown in figure 8, as can be seen from the figure this algorithm calibrated
True has judged that in image, sky whether there is (as shown in Fig. 8 (c)), and achieves the cluster of non-conterminous sky segment (as Fig. 8
Shown in (b)).
6. the partitioning algorithm of ground and facade object.
As shown in Figure 8 based on above-mentioned ambiguity function, can will extract in ground most in image, but can go out
Now part facade object segment and the misjudged situation of ground segment, in addition it is also possible that not meeting the situation of gravitational field, such as
Shown in Fig. 9, No. 27 and No. 34 segments, occur in that ground is suspended in the situation on facade object, accordingly, it would be desirable to ambiguity function
Judged result further revised.
Only need to when not meeting gravitational field space geometry logic to carry out geometrically logic below judges can
To revise.Larger ground and there is low coverage with the situation of facade object erroneous judgement mainly due in image in aforesaid algorithm
From building caused by, it is therefore desirable to carry out to whether there is closely heavy construction in image shown in such as Fig. 8 (c) (d)
Judge.Concrete grammar is as follows：
(1) continuity according to ground and its gravitational field space geometry context property, will be suspended in facade object
Ground segment is classified as facade object, as shown in Figure 10；
(2) by carrying out Hough transform to being identified as facade object segment in image, and by based on rectilinear direction angle
The statistic histogram of degree, by the intensity to its Manhattan directional information, to judge that in figure whether there is and large-scale closely to build
Thing, if there is no then terminating the correction to ground, if there is then entering next step；
(3) with its fillet with ground segment of the Manhattan directional information correction of building in facade object, Figure 10
Ground border correction result for Fig. 9.
7. depth perception model.
This model assumes initially that ground is continuously to extend and more smooth, and Vision imaging system has clear and definite directionality, that is,
Image upper limb is the surface of 3d space, and lower edge is the underface of 3d space, the vision system physics based on pinhole imaging system principle
Model is as shown in figure 12.
In ground depth information and image pixel location perspective projection relation as follows：
Wherein：H is the height away from ground for the video camera, and β is the angle of camera optical axis and eye-level display, and depth projection angle α is
The angle of eye-level display oo ' and straight line op, its span isP ' is that the p point on ground is being imaged target
Projection on face, f is lens focus, and h is the distance that the eye-level display in imaging target surface is put to p ', then the appreciable ground of video camera
Span apart from d is
8. the depth perception figure of image.
From the relational expression (22) of height H away from ground of ground depth and video camera and depth projection angle α, when H is normal
During number, each pixel depth that ground projects can be represented in video camera with the value of α, we will
Value be mapped to CIELAB color spaceColour circle on, and the color of sky is defined as colour circleThe face at place
Color, as shown in figure 13.Depth perception figure corresponding to Figure 11 is as shown in figure 14.
Part that the present invention does not relate to is same as the prior art or can be realized using prior art.
Claims (6)
1. monocular vision space recognition method under a species ground gravitational field environment, is characterized in that it comprises the following steps：
First, image is carried out with CIELAB color space values L, a, b and the x based on pixel, the super-pixel segmentation of y-coordinate value, with
Produce super-pixel image；
Secondly, using the general clustering algorithm based on super-pixel color characteristics, texture feature vector distance and syntople, will split
The super-pixel image dimensionality reduction of formation simultaneously generates big segment；
3rd, the gravitational field Fuzzy Distribution density function big segment with gained respectively of sky, ground and facade object will be represented
Pixel is multiplied, and obtains the desired value of big segment, thus completing the preliminary classification of sky, ground and facade object；The described phase
Prestige value be by each pixel in big segment in image vertical direction with ground Fuzzy Distribution density function G, sky obscure point
Cloth density function S and facade object Fuzzy Distribution density function V, in (- HG, HS) in the range of multiplication gained, its computing formula
For：
Wherein：GE、SE、VEIt is that gravity field on earth's surface Fuzzy Distribution density function G, sky gravitational field mould are based on to the segment in image
The ground (ground) of paste distribution density function S and facade object gravitational field Fuzzy Distribution density function V summation gained, sky
(sky), the mathematical expectation of facade (vertical face), niFor clustering the number of pixels in the i-th row for the segment, rbFor dendrogram
The most downward, the r of blocktFor segment on prolong, i ∈ (0,1 ... ..., HZ), HZ is image max pixel value in the height direction；
HI=HG+HS, HIFor the height of image, HGFor the distance of image eye-level display to image base, HSImage eye-level display is to image apex
Distance；Then being categorized as of segment：
4th, the classification chart of sky, ground and facade object is gone out by the sampling of individual layer small echo and Manhattan Directional feature extraction；
Finally, spatial depth perceptual map is generated based on the linear perspective information of national forest park in Xiaokeng and ground.
2. method according to claim 1, it is characterized in that described general clustering algorithm include super-pixel clustering method and
General clustering method on the basis of super-pixel, described super-pixel clustering method is changed using the simple linear that Achanta R proposes
It is SLIC (Simple Linear Iterative Clustering) for clustering algorithm, this algorithm is with the CIELAB color of pixel
L, a, b value in space and the x of pixel, y-axis coordinate builds 5 dimension spaces, is defined as follows：
Wherein：Ck=[lk,ak,bk,xk,yk]TCenter for cluster；[li,ai,bi,xi,yi]T5 dimension spaces for image slices vegetarian refreshments
Coordinate；N is the pixel count of image；K is the number of the super-pixel that expectation obtains；S is super-pixel center grates spacing；DsFor color
Lab is apart from dlabAnd dxyStandardization distance based on S；M is controlled super-pixel density factor；
Described based on the general clustering method of super-pixel is：
(1) using n super-pixel produced by SLIC algorithm as undirected weights figure G summit V={ v1,v2,…,vn}；
(2) adjacency matrix builds, i=1,2 ... n；J=1,2 ... n, wherein, n is the number of super-pixel；
(3) structure of weights adjacency matrix, i=1,2 ... n；J=1,2 ... n；
Wherein weight w (i, j) is the standardization CIELAB color histogram calculating between two neighboring super-pixel
Bhattacharyya coefficient, concrete construction method is that the color space of image is transformed into CIELAB space, and by L * channel
Span is divided into 8 grades of deciles, and the span of a passage is divided into 16 deciles, the span of b passage is divided into 16
Grade, the purpose wherein L * channel span being divided into 8 grades is the disturbance reducing chroma-luminance change to weights；Each surpasses
Pixel calculates histogram in the space of 8 × 16 × 16=2048 dimensionWherein hl
I standardization histogram that () calculates in the space of l=8 × 16 × 16=2048 dimension for each super-pixel, l=2048, then when
Ei,jWhen=1
Value for weight w (i, j) increases by two constraintss being based respectively on color distance and texture energy distance, division
As follows：
1. it is based on color distance constraints：As w (i, j)≤WTWhen, then take w (i, j)=0, wherein WTSpan be (0.7
～1.0) between；
2. the constraints of texture energy distance：Using l2The average energy of norm calculation each super-pixel segment is estimated, that is,
Wherein rbThe most downward for segment, rtFor segment on prolong, clFor the i-th row segment Far Left, crThe rightest for the i-th row segment
Side, R (i, j) is the small echo sampled value at (i, j) point in image, calculates the four-dimensional small echo of each super-pixel block according to formula (8)
Characteristic vector, i.e. e (i)=(ei(LL), ei(LH), ei(HL), ei(HH)), and ask for the Bhattacharyya between its neighbouring super pixels
Coefficient value Be(i,j)；
Wherein,
Work as Be(i,j)≤BTWhen, then take w (i, j)=0, wherein BTSpan between (0.85～1.0)；
The purpose of two above constraints is to improve the color of neighbouring super pixels point and the similarity threshold of texture, to protect sky
Empty boundary characteristic and between facade object, between facade object and ground；
(4) structure of matrix, i=1,2 ... n are spent；J=1,2 ... n；
(5) structure of standardization Laplacian matrix, to be calculated normalized using Normalized-cut criterion
Laplacian matrix：
Lsym=I-D-1/2WD-1/2(12)
Wherein：D is degree matrix, and W is weights adjacency matrix；
(6) calculate LsymCarry out Eigenvalues Decomposition, and take the characteristic vector corresponding to front K minimal eigenvalue, V1, V2..., Vk；
Wherein K=[0.1 × n], that is, take the 10% of n as image clustering characteristic vector dimension, to realize dimensionality reduction purpose；
(7) by V1, V2..., VkRearrange Rn×kEach element in matrix is simultaneously taken absolute value to obtain matrix U by matrix；
(8) for i=1,2 ... n, make yi∈RkFor the i-th row vector of matrix U, RkIt is the real vector of K dimension；
(9) y to non-zeroi∈RkVector is normalized, and is clustered with Bhattacharyya Y-factor method Y, wherein
The B of Bhattacharyya distanceUThreshold value is between (0.85～1.0), that is, work as BUDuring more than or equal to threshold value, gathered between super-pixel
Class；
(10) segment code requirement CIELAB color histogram is clustered to each, and adjacent segment class is carried out using formula (7)
Bhattacharyya calculates apart from w (i, j), adopts formula (9) to calculate the B of adjacent segment simultaneouslye(i, j), as w (i, j) ＞ WTAnd
Be(i, j) ＞ BTShi Jinhang clusters；
(11) repeat (10th) step, until convergence.
3. method according to claim 1, is characterized in that described big segment generates and adopts geometry inclusion relation cluster side
Method, to eliminate isolated island segment, so-called isolated island segment refers to one or more segments by the segment of a big segment encirclement completely,
Isolated island segment can be clustered into the big segment surrounding this isolated island completely by the clustering algorithm of geometry inclusion relation, thus avoiding several
What contextual algorithms is to unusual produced by isolated island segment spatial classification；Concrete grammar is：
(1) find hollow out segment, its criterion is to work as Nb-nb>When 0, then segment is hollow out segment, wherein NbFor all borders of segment
Pixel value, nbFor the pixel value of segment external boundary, if Nb-nb>0 entrance next step, otherwise segment is not hollow out segment；
(2) segment is filled for border with the mark value of artwork block with external boundary；
(3) replace former hollow out segment to fill segment.
4. method according to claim 1, is characterized in that sky and facade thing are extracted in described employing individual layer small echo sampling
L is adopted during the classification chart of body2The average energy of norm calculation each object segment is estimated, that is,
Wherein：NpFor segment number of pixels, rbThe most downward for segment, rtFor segment on prolong, clThe most left for the i-th row segment
Side, crFor the i-th row segment rightmost, wherein R (i, j) is the small echo sampled value at (i, j) point in image, puts down calculating each segment
Need to remove energy produced by each segment edge during equal energy norm；
It is calculated the four-dimensional wavelet-based attribute vector of segment, i.e. (e by energy normLL,eLH,eHL,eHH), wherein eLLCharacterize
It is segment overall brightness characteristic, eLH,eHL,eHHCharacterize is segment high frequency texture feature, and outdoor sky on daytime is in the picture
Characteristic common manifestation is on high brightness and low-yield high frequency texture feature；
(1) if eLL>mean(eLL1, eLL2... eLLn) it is then candidate's sky segment, wherein
eLL1, eLL2... eLLn∈ sky and the e of facade objectLLValue, wherein：Mean () is mean value function；
(2) under meeting above-mentioned condition, when the energy norm of one layer of un-downsampling wavelet transform of segment
When, then segment is candidate's sky segment, during without meeting this condition segment, then judges segment not as sky segment, Ec's
Between span (0～7)；
(3) when presence meets above-mentioned segment, then determine whether the segment as border is prolonged with image, deposit if there are then judgement
Segment on high, otherwise no sky in process decision chart picture；
(4) under meeting above-mentioned condition if there is candidate's sky segment not unique, then choose area maximum segment be sky
Segment, and with color distance value dabAnd brightness distance value dLFor criterion, sky is clustered, formula is as follows：
And
Wherein as、bsIt is respectively the average of sky segment CIELAB color space a, b color channel, ai、biIt is respectively candidate's sky
The average of segment CIELAB color space a, b color channel, as candidate sky segment dab≤ C and dL≤ L then be sky, otherwise for
Facade object, wherein, the span (0～30) of C, the span (0～70) of L；
(5) if the number that the sky area that cluster is generated is calculated its pixel is less than the 2 ‰ of image pixel, returned
For facade object；
(6) all non-sky segments are classified as facade object.
5. method according to claim 1, is characterized in that ground and facade thing are extracted in described employing individual layer small echo sampling
Following method of discrimination is adopted during the classification chart of body：
(1) continuity according to ground and its upper and lower property of gravitational field space geometry, will be suspended on the surface map in facade object
Block is classified as facade object；
(2) by carrying out Hough transform to being identified as facade object segment in image, and by based on rectilinear direction angle
Statistic histogram, by the intensity to its Manhattan directional information, to judge that in figure whether there is large-scale closely building, such as
Fruit does not exist, and terminates the correction to ground, if there is then entering next step；
(3) with its fillet with ground segment of the Manhattan directional information correction of building in facade object.
6. method according to claim 1, is characterized in that the gravitational field of described sky, ground and facade object obscures and divides
Cloth density function is respectively：
(1) gravity field on earth's surface Fuzzy Distribution density function G：
Work as HG≥HSWhen：OrderAnd
Then
H in formulaGFor the distance away from image base for the eye-level display；HSFor the distance away from image top margin for the eye-level display；X is that pixel is high in image
The coordinate in degree direction；N is the exponent number of density function；
Work as HG＜ HSWhen：G (x)=- S (x)
I.e.
Wherein：N=1,2,3 ... N, N ∈ positive integer；
(2) sky gravitational field Fuzzy Distribution density function S：
Work as HG＜ HSWhen：OrderAnd
Then
Work as HG≥HSWhen：S (x)=- G (x)
I.e.
Wherein：N=1,2,3 ... N, N ∈ positive integer；
(3) facade object gravitational field Fuzzy Distribution density function V：
Priority Applications (5)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410212438.6A CN103955710B (en) | 2013-11-29 | 2014-05-19 | Method for monocular vision space recognition in quasi-earth gravitational field environment |
| US14/684,431 US9390348B2 (en) | 2014-05-19 | 2015-04-12 | Method for categorizing objects in image |
| US14/684,428 US9471853B2 (en) | 2014-05-19 | 2015-04-12 | Method and apparatus for image processing |
| US14/684,433 US9805293B2 (en) | 2014-05-19 | 2015-04-13 | Method and apparatus for object recognition in image processing |
| US14/684,434 US9626598B2 (en) | 2014-05-19 | 2015-04-13 | Method and apparatus for image processing |
Applications Claiming Priority (4)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310626783.X | 2013-11-29 | ||
| CN201310626783 | 2013-11-29 | ||
| CN201310626783X | 2013-11-29 | ||
| CN201410212438.6A CN103955710B (en) | 2013-11-29 | 2014-05-19 | Method for monocular vision space recognition in quasi-earth gravitational field environment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103955710A CN103955710A (en) | 2014-07-30 |
| CN103955710B true CN103955710B (en) | 2017-02-15 |
Family
ID=50213194
Family Applications (2)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310652422.2A Active CN103632167B (en) | 2013-11-29 | 2013-12-05 | Monocular vision space recognition method under class ground gravitational field environment |
| CN201410212438.6A Active CN103955710B (en) | 2013-11-29 | 2014-05-19 | Method for monocular vision space recognition in quasi-earth gravitational field environment |
Family Applications Before (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310652422.2A Active CN103632167B (en) | 2013-11-29 | 2013-12-05 | Monocular vision space recognition method under class ground gravitational field environment |
Country Status (1)
| Country | Link |
|---|---|
| CN (2) | CN103632167B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103632167B (en) * | 2013-11-29 | 2016-10-12 | 金陵科技学院 | Monocular vision space recognition method under class ground gravitational field environment |
| CN104077603B (en) * | 2014-07-14 | 2017-04-19 | 南京原觉信息科技有限公司 | Outdoor scene monocular vision space recognition method in terrestrial gravity field environment |
| CN104063707B (en) * | 2014-07-14 | 2017-05-24 | 南京原觉信息科技有限公司 | Color image clustering segmentation method based on multi-scale perception characteristic of human vision |
| CN104091180B (en) * | 2014-07-14 | 2017-07-28 | 南京原觉信息科技有限公司 | The recognition methods of trees and building in outdoor scene image |
| CN104077611B (en) * | 2014-07-14 | 2017-06-09 | 南京原觉信息科技有限公司 | Indoor scene monocular vision space recognition method under class ground gravitational field environment |
| EP2966616B1 (en) | 2014-07-10 | 2018-06-13 | Thomson Licensing | Method and apparatus for tracking superpixels between related images |
| CN104794688B (en) * | 2015-03-12 | 2018-04-03 | 北京航空航天大学 | Single image to the fog method and device based on depth information separation sky areas |
| CN106097252B (en) * | 2016-06-23 | 2019-03-12 | 哈尔滨工业大学 | High spectrum image superpixel segmentation method based on figure Graph model |
| CN111238490B (en) * | 2018-11-29 | 2022-03-08 | 北京地平线机器人技术研发有限公司 | Visual positioning method and device and electronic equipment |
| CN112419392A (en) * | 2020-11-30 | 2021-02-26 | 广州博进信息技术有限公司 | Method, apparatus and medium for calculating actual size of moving object based on machine vision |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1275189A (en) * | 1998-06-08 | 2000-11-29 | 卡尔海因茨·斯特罗贝尔 | Efficient light engine systems, components and methods of manufacture |
| US6404920B1 (en) * | 1996-09-09 | 2002-06-11 | Hsu Shin-Yi | System for generalizing objects and features in an image |
| CN101126698A (en) * | 2006-06-01 | 2008-02-20 | Ana技术公司 | Object analysis method and apparatus |
| CN101371165A (en) * | 2006-01-25 | 2009-02-18 | 阿克斯有限责任公司 | Geophysical terrain survey correction method |
| CN103632167A (en) * | 2013-11-29 | 2014-03-12 | 金陵科技学院 | Method for identifying monocular visual spaces in terrestrial gravitational field environments |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6526379B1 (en) * | 1999-11-29 | 2003-02-25 | Matsushita Electric Industrial Co., Ltd. | Discriminative clustering methods for automatic speech recognition |
| CN101751666A (en) * | 2009-10-16 | 2010-06-23 | 西安电子科技大学 | Semi-supervised multi-spectral remote sensing image segmentation method based on spectral clustering |
| CN103353987B (en) * | 2013-06-14 | 2015-10-28 | 山东大学 | A kind of superpixel segmentation method based on fuzzy theory |
| CN103413316B (en) * | 2013-08-24 | 2016-03-02 | 西安电子科技大学 | Based on the SAR image segmentation method of super-pixel and optimisation strategy |
| CN103456013B (en) * | 2013-09-04 | 2016-01-20 | 天津大学 | A kind of method representing similarity between super-pixel and tolerance super-pixel |
-
2013
- 2013-12-05 CN CN201310652422.2A patent/CN103632167B/en active Active
-
2014
- 2014-05-19 CN CN201410212438.6A patent/CN103955710B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6404920B1 (en) * | 1996-09-09 | 2002-06-11 | Hsu Shin-Yi | System for generalizing objects and features in an image |
| CN1275189A (en) * | 1998-06-08 | 2000-11-29 | 卡尔海因茨·斯特罗贝尔 | Efficient light engine systems, components and methods of manufacture |
| CN101371165A (en) * | 2006-01-25 | 2009-02-18 | 阿克斯有限责任公司 | Geophysical terrain survey correction method |
| CN101126698A (en) * | 2006-06-01 | 2008-02-20 | Ana技术公司 | Object analysis method and apparatus |
| CN103632167A (en) * | 2013-11-29 | 2014-03-12 | 金陵科技学院 | Method for identifying monocular visual spaces in terrestrial gravitational field environments |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103632167A (en) | 2014-03-12 |
| CN103955710A (en) | 2014-07-30 |
| CN103632167B (en) | 2016-10-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103955710B (en) | Method for monocular vision space recognition in quasi-earth gravitational field environment | |
| CN111832655B (en) | Multi-scale three-dimensional target detection method based on characteristic pyramid network | |
| CN106650640B (en) | Negative obstacle detection method based on laser radar point cloud local structure characteristics | |
| CN111145174B (en) | 3D target detection method for point cloud screening based on image semantic features | |
| CN106909902B (en) | Remote sensing target detection method based on improved hierarchical significant model | |
| CN106022381B (en) | Automatic extraction method of street lamp pole based on vehicle-mounted laser scanning point cloud | |
| CN109903331B (en) | Convolutional neural network target detection method based on RGB-D camera | |
| CN105631892B (en) | It is a kind of that detection method is damaged based on the aviation image building of shade and textural characteristics | |
| CN111626128A (en) | Improved YOLOv 3-based pedestrian detection method in orchard environment | |
| CN112766184B (en) | Remote sensing target detection method based on multi-level feature selection convolutional neural network | |
| CN106127791A (en) | A kind of contour of building line drawing method of aviation remote sensing image | |
| CN111753682B (en) | Hoisting area dynamic monitoring method based on target detection algorithm | |
| CN108804992B (en) | Crowd counting method based on deep learning | |
| CN107369158A (en) | The estimation of indoor scene layout and target area extracting method based on RGB D images | |
| CN106503170B (en) | It is a kind of based on the image base construction method for blocking dimension | |
| CN113129449A (en) | Vehicle pavement feature recognition and three-dimensional reconstruction method based on binocular vision | |
| CN114332921A (en) | Pedestrian detection method based on improved clustering algorithm for Faster R-CNN network | |
| CN104077603B (en) | Outdoor scene monocular vision space recognition method in terrestrial gravity field environment | |
| Babahajiani et al. | Comprehensive automated 3D urban environment modelling using terrestrial laser scanning point cloud | |
| CN104077611B (en) | Indoor scene monocular vision space recognition method under class ground gravitational field environment | |
| CN104008374B (en) | Miner's detection method based on condition random field in a kind of mine image | |
| CN110276270B (en) | High-resolution remote sensing image building area extraction method | |
| He et al. | Building extraction based on U-net and conditional random fields | |
| CN116052099A (en) | Small target detection method for unstructured road | |
| CN109522813B (en) | Improved random walk algorithm based on pedestrian salient features |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20190718 Address after: 211100 Jiangsu province Nanjing city Jiangning high tech park, Tianyuan Road No. 1009 Patentee after: NANJING YUANJUE INFORMATION AND TECHNOLOGY Co. Address before: No. 99 Jiangning Road, Nanjing District hirokage 211169 cities in Jiangsu Province Patentee before: Jinling Institute of Technology |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20221221 Address after: 271100 No. 001, Huiyuan Street, Laiwu District, Jinan, Shandong Patentee after: SHANDONG TAIJIN PRECISION FORGING CO.,LTD. Address before: 211100 Tianyuan East Road 1009, Jiangning High-tech Park, Nanjing, Jiangsu Province Patentee before: NANJING YUANJUE INFORMATION AND TECHNOLOGY Co. |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20230306 Address after: Room 907-910, Building 8, Phase II, Fortune Plaza, 228 Tianyuan East Road, Jiangning District, Nanjing, Jiangsu Province, 211100 Patentee after: NANJING YUANJUE INFORMATION AND TECHNOLOGY Co. Address before: 271100 No. 001, Huiyuan Street, Laiwu District, Jinan, Shandong Patentee before: SHANDONG TAIJIN PRECISION FORGING CO.,LTD. |
|
| TR01 | Transfer of patent right |