CN105528787A - Polarimetric SAR image bridge detection method and device based on level set segmentation - Google Patents

Abstract

The present invention provides a polarimetric SAR image bridge detection method and device based on level set segmentation. The method comprises a step of carrying out level set segmentation on a polarimetric SAR image according to a regional statistical characteristic and obtaining a land and a water area, a step of extracting the characteristic point of the contour of the water area and determining an area of interest in the land through the distance of the characteristic point to be a suspected bridge area, a step of removing the false alarm in the suspected bridge area to realize bridge detection, and a step of carrying out constant false alarm detection on the suspected bridge area with the removal of the false alarm and distinguishing a strong scatter bridge. According to the method of the embodiment of the present invention, the accuracy of bridge detection can be improved.

Description

Based on Polarimetric SAR Image Bridges Detection and the device of level-set segmentation

Technical field

The present invention relates to technical field of image processing, particularly a kind of Polarimetric SAR Image Bridges Detection based on level-set segmentation and device.

Background technology

Can detect that bridge is all extremely important for the formulation etc. of the renewal of geographical data bank, the assessment of disaster and military plan in the picture.In the related, by methods such as rim detection or Randon conversion, bridge machinery is carried out to polarization SAR (SyntheticApertureRadar, synthetic-aperture radar) image, but the accuracy detected is mostly lower.

Summary of the invention

The present invention is intended to solve one of technical matters in correlation technique at least to a certain extent.For this reason, one object of the present invention is to propose a kind of Polarimetric SAR Image Bridges Detection based on level-set segmentation, can improve the accuracy of bridge machinery.

Second object of the present invention is to propose a kind of Polarimetric SAR Image bridge detecting device based on level-set segmentation.

The Polarimetric SAR Image Bridges Detection based on level-set segmentation of embodiment, comprises the following steps: carry out level-set segmentation according to Region Statistical Features to Polarimetric SAR Image, obtains land and waters according to a first aspect of the present invention; Extract the unique point of the profile in described waters, and determine the area-of-interest in described land by the distance of described unique point, using as suspected bridge area; The false-alarm rejected in described suspected bridge area realizes bridge machinery; CFAR detection is carried out to the described suspected bridge area after rejecting false-alarm, distinguishes strong scatterer bridge.

According to the Polarimetric SAR Image Bridges Detection based on level-set segmentation of the embodiment of the present invention, by carrying out level-set segmentation to Polarimetric SAR Image to obtain land and waters, and extract the unique point of the profile in waters, according to unique point determination suspected bridge area, then CFAR detection is carried out to the suspected bridge area after rejecting false-alarm, thus distinguish strong scatterer bridge.Thus, by the method for level-set segmentation, land and waters can be obtained more accurately, then the process such as rejecting false-alarm and CFAR detection etc. that combines distinguishes strong scatterer bridge, greatly can improve the accuracy of bridge machinery.

In addition, the Polarimetric SAR Image Bridges Detection based on level-set segmentation according to the above embodiment of the present invention can also have following additional technical characteristic:

According to one embodiment of present invention, the unique point of the profile in described waters is extracted by digital curve division conflation algorithm.

According to one embodiment of present invention, using disconnected described area-of-interest as described false-alarm.

According to one embodiment of present invention, describedly carry out CFAR detection specifically comprise rejecting the described suspected bridge area after false-alarm: using the described suspected bridge area after described rejecting false-alarm as point target, and represent described point target by zone leveling coherence matrix, and by polarimetric whitening filter device, described point target is detected.

The Polarimetric SAR Image bridge detecting device based on level-set segmentation of embodiment according to a second aspect of the present invention, comprising: segmentation module, for carrying out level-set segmentation according to Region Statistical Features to Polarimetric SAR Image, obtains land and waters; Determination module, for extracting the unique point of the profile in described waters, and determines the area-of-interest in described land by the distance of described unique point, using as suspected bridge area; Reject module, realize bridge machinery for the false-alarm rejecting described suspected bridge area; Detection module, for carrying out CFAR detection to the described suspected bridge area after rejecting false-alarm, distinguishes strong scatterer bridge.

According to the Polarimetric SAR Image bridge detecting device based on level-set segmentation of the embodiment of the present invention, by carrying out level-set segmentation to Polarimetric SAR Image to obtain land and waters, and extract the unique point of the profile in waters, according to unique point determination suspected bridge area, then CFAR detection is carried out to the suspected bridge area after rejecting false-alarm, thus distinguish strong scatterer bridge.Thus, by level-set segmentation, land and waters can be obtained more accurately, then combination rejecting false-alarm and CFAR detection etc. distinguish strong scatterer bridge, greatly can improve the accuracy of bridge machinery.

In addition, the Polarimetric SAR Image bridge detecting device based on level-set segmentation according to the above embodiment of the present invention can also have following additional technical characteristic:

According to one embodiment of present invention, described determination module extracts the unique point of the profile in described waters by digital curve division conflation algorithm.

According to one embodiment of present invention, described rejecting module using disconnected described area-of-interest as described false-alarm.

According to one embodiment of present invention, described detection module specifically for: using the described suspected bridge area after described rejecting false-alarm as point target, and represent described point target by zone leveling coherence matrix, and by polarimetric whitening filter device, described point target is detected.

Accompanying drawing explanation

Fig. 1 is according to an embodiment of the invention based on the process flow diagram of the Polarimetric SAR Image Bridges Detection of level-set segmentation;

Fig. 2 is the unique point of the profile in waters according to an embodiment of the invention and the schematic diagram of suspected bridge area;

Fig. 3 is the pcolor of the Polarimetric SAR Image in the Singapore area comprising multiple bridge according to an embodiment of the invention;

Fig. 4 is the result schematic diagram of according to an embodiment of the invention image in Fig. 3 being carried out to bridge machinery;

Fig. 5 is the result schematic diagram distinguishing strong scatterer bridge according to an embodiment of the invention;

Fig. 6 is according to an embodiment of the invention based on the structured flowchart of the Polarimetric SAR Image bridge detecting device of level-set segmentation.

Embodiment

Be described below in detail embodiments of the invention, the example of described embodiment is shown in the drawings, and wherein same or similar label represents same or similar element or has element that is identical or similar functions from start to finish.Be exemplary below by the embodiment be described with reference to the drawings, be intended to for explaining the present invention, and can not limitation of the present invention be interpreted as.

Below with reference to the accompanying drawings Polarimetric SAR Image Bridges Detection based on level-set segmentation and the device of the embodiment of the present invention are described.

Fig. 1 is according to an embodiment of the invention based on the process flow diagram of the Polarimetric SAR Image Bridges Detection of level-set segmentation.

As shown in Figure 1, the Polarimetric SAR Image Bridges Detection based on level-set segmentation of the embodiment of the present invention, comprises the following steps:

S101, carries out level-set segmentation according to Region Statistical Features to Polarimetric SAR Image, obtains land and waters.

Be described to look Polarimetric SAR Image in an embodiment of the present invention more.Look Polarimetric SAR Image scattering matrix more and obey multiple Wishart (Vichy is special) distribution, for homogeneous region, the coherence matrix T obtained by scattering matrix vector quantization also obeys multiple Wishart and distributes.If the average of the coherence matrix of homogeneous region is Σ, be L depending on number, and POLARIZATION CHANNEL number is p, then can be designated as T ~ W (Σ, L, p).Thus, the probability density function of above-mentioned correlation matrix is:

$f\left(T\right|\mathrm{\Σ},L,p)=\frac{L^{pL}{\left(\right|T\left|\right)}^{L-p}\exp \{-Ltr\left({\mathrm{\Σ}}^{-1}T\right)\}}{K(L,p){\left(\right|\mathrm{\Σ}\left|\right)}^L}---\left(1\right)$

Wherein, $K(L,p)={\mathrm{\π}}^{\frac{p(p-1)}{2}}G\left(L\right)\mathrm{...}G(L-p+1),$ G () is gamma (Gamma) function.

In one embodiment of the invention, represent the plane of delineation with R, represent given Polarimetric SAR Image, represent segmentation with P with I, represent the boundary line in land and waters with Γ, wherein, Γ is closed curve, meanwhile, with R ₁and R ₂represent the land and waters split by curve Γ respectively, then, under splitting P, the posterior probability of Polarimetric SAR Image I is p (I| Ρ (R ₁, R ₂)).According to bayesian criterion, when posterior probability is maximum, be divided into optimal segmentation, namely optimal segmentation the condition that meets be:

$P=\underset{R_1,R_2}{\max }p\left(I\right|P\left(R_1,R_2\right))---\left(2\right)$

Again because p (I|P) ∝ p (P|I) p (P), if land R ₁with waters R ₂separate, area condition probability function is f (I _i| R _i), wherein i is 1 or 2, represents the probability density function of image I when being divided into land or waters respectively.Segmentation prior probability p (P) may be defined as outline line length function p (S) ∝ e ^{-ν | Γ |}, ν >0, then can split the energy function of curve and the equivalent form of value of parted pattern:

$E(\mathrm{\Γ},\{R_1,R_2\left\}\right)=v\left|\mathrm{\Γ}\right|-\underset{R_1}{\∫}\log f\left(x\right|R_1)dx-\underset{R_2}{\∫}\log f\left(x\right|R_2)dx$

$\widehat{\mathrm{\Γ}}=\underset{\mathrm{\Γ}}{\min }\left\{E(\mathrm{\Γ},\{R_1,R_2\left\}\right)\right\}---\left(3\right)$

Wherein ν is curve gauge parameter, | _Γ| represent length of curve, for optimal segmentation curve.

In one embodiment of the invention, if by level set function Φ (c (t), t) represent curve Γ, then zero level collection homologous thread Γ (t)=c (t) | Φ (c (t), t)=0}.Meanwhile, the energy definition of level set functional Φ is obtained according to curve energy definition:

$E\left(\mathrm{\Φ}\right)=-\underset{R}{\∫}\left(H\right(\mathrm{\Φ}\left)log\right(f\left(x|R_1\right))+(1-H\left(\mathrm{\Φ}\right)\left)log\right(f\left(x|R_2\right)\left)\right)dx+v\underset{R}{\∫}|\▿H\left(\mathrm{\Φ}\right)|dx---\left(4\right)$

Wherein, H (Φ) is step function, when Φ >=0, and H (Φ)=1, as Φ <0, H (Φ)=0.R ₁corresponding region, Φ>=0, R ₂corresponding Φ <0 region.

The partial differential equation of level set functional Φ is:

$\frac{\&part;\mathrm{\&Phi;}}{\&part;t}=-\mathrm{\&delta;}\left(\mathrm{\&Phi;}\right)(v\mathrm{\&kappa;}+log\frac{f\left(x\right|R_2)}{f\left(x\right|R_1)})---\left(5\right)$

Wherein δ (Φ) is impulse function, for curvature of curve.Then can according to formula (5), and by the variational method along level set energy function negative gradient direction Approach by inchmeal, solve the zero level set function under the minimum condition of energy.In one embodiment of the invention, if the mean value of the coherence matrix on waters and land is respectively Σ ₁, Σ ₂, then level set evolution function is:

$\frac{\&part;\mathrm{\&Phi;}}{\&part;t}=-\mathrm{\&delta;}\left(\mathrm{\&Phi;}\right)(v\mathrm{\&kappa;}+L(log\left|{\mathrm{\&Sigma;}}_1\right|+tr\left({{\mathrm{\&Sigma;}}_1}^{-1}T\right))-L(log\left|{\mathrm{\&Sigma;}}_2\right|+tr\left({{\mathrm{\&Sigma;}}_2}^{-1}T\right)\left)\right)---\left(6\right)$

Wherein, Σ ₁, Σ ₂obtain by possibility predication, under the initialization curve Γ and corresponding Φ and parameter value ν of setting, utilize formula (6) to carry out iteration, until when zero level set function no longer changes, can level-set segmentation be realized.

After realizing level-set segmentation, can judge the classification in two regions according to the average scattering watt level in two regions of segmentation, generally speaking, the average scattering power in waters is little.Because each region contour of level-set segmentation closes continuously, for land and water segmentation binary map, the region of all connections can be obtained by 8 connected domain evaluation algorithm.

S102, is extracted the unique point of profile in waters, and is determined the area-of-interest in land by the distance of unique point, using as suspected bridge area.

After above-mentioned primary segmentation process, the very little waters of some areas and land is there is in segmentation result, for the not intensive especially situation of bridge, these small size regions think irrelevant with bridge machinery, can reject under the elemental area threshold value of setting, threshold value can be determined according to image size and resolution.Must arrive waters thus still has some by the region of lower scattering strength split by mistake, if the coupling part in these regions is apart from little, can form some suspected bridge area.Consider that detecting bridge is positioned on main sea or divarication, carries out waters merging by the spacing of each waters profile, and according to the width of bridge setting reasonable threshold value, extract with main body waters and Qi Ge branch apart near waters part.

Particularly, the unique point of the profile in waters is extracted by digital curve division conflation algorithm.In one embodiment of the invention, digital curve division conflation algorithm can select DP (Douglas-Peucker) algorithm, and its idiographic flow is: the two ends selecting the contour curve section in waters are initial characteristics point; Obtain the air line distance of points all on segment of curve to initial characteristics point; If points all on segment of curve is all less than default Maximum tolerance to the air line distance of initial characteristics point, then determine that this initial characteristics point is the unique point of the profile in waters, otherwise, the point maximum with the air line distance of initial characteristics point is selected to be cut-point, and with this cut-point, segment of curve is divided into two sections, then respectively above-mentioned algorithm is adopted, until determine all unique points of the contour curve section in waters to the segment of curve recurrence after these two sections segmentations.

Fig. 2 shows the unique point of waters profile, and wherein, black region is land, and white portion is waters, and the point of " * " place mark is the unique point of waters profile.Suppose all unique point { u defining the profile in A waters according to above-mentioned algorithm _iand all unique point { v of profile in B waters _i, can { u be calculated _iand { v _ibetween the distance of each unique point.In an embodiment of the present invention, can by two the unique point { ps of distance lower than predeterminable range _iand { q _icomposition characteristic point pair, and determine waters bridge end points by the coordinate relation of unique point each in multiple feature point pairs.Further, according to the size setting threshold value of bridge, feature point pairs can be merged, such as, the feature point pairs of all distances lower than setting threshold value can be merged, thus when there is multiple bridge between A and B waters, the bridge of different length can be distinguished.If the correspondence set of A and B two waters is respectively { { S after merging ₁..., { S _nand { { W ₁..., { W _n, then { S _iand { W _ia new feature point pairs can be formed.Thus, new feature point pairs { S can be determined further _iand { W _irepresentated by waters bridge end points, and according to end points determination area-of-interest.For example, as shown in Figure 2, if waters bridge end points is four points, then the quadrilateral area that four points form is area-of-interest, and this area-of-interest is suspected bridge area.In a particular embodiment of the present invention, Maximum tolerance, predeterminable range and the bridge length threshold value preset can need according to the engineer's scale of Polarimetric SAR Image and concrete detection and set.Thus, by by two the unique points composition feature point pairs of distance lower than predeterminable range, can prevent from calculating all unique points and the calculated amount that causes area-of-interest that is excessive and that determine is comparatively large, thus speed and the precision of bridge machinery can be improved to a certain extent.

S103, the false-alarm rejected in suspected bridge area realizes bridge machinery.

In the waters obtained in step S101, still existence part and main body waters and branch thereof are apart from inner waters, near land, and the land between these waters and other waters is divided into suspected bridge area by mistake, namely becomes false-alarm targets.Usually these false-alarm targets are compared with true bridge, and contour shape is irregular, and the pontic region determined by end points is not communicated with, and therefore, using disconnected area-of-interest as false-alarm, and can be rejected.

S104, carries out CFAR detection to the suspected bridge area after rejecting false-alarm, distinguishes strong scatterer bridge.

Usually, there is guardrail wires around important bridge, the scattering strength of guardrail wires is higher, and bridge floor, dihedral angle between guardrail and waters or trihedral angle scattering can make pontic regions scatter intensity higher than other Bridge object and false-alarm targets.Therefore, from suspected bridge area, bridge area is distinguished by scattering strength.In one embodiment of the invention, utilisation point target constant false alarm detector carries out the differentiation of strong scatterer bridge.Particularly, using the suspected bridge area after rejecting false-alarm as point target, and point target can be represented by zone leveling coherence matrix, and by polarimetric whitening filter device, point target be detected, thus can effectively detect by the bridge higher to scattering strength.

More specifically, if the average coherence matrix of rejecting the suspected bridge area after false-alarm is C, then polarimetric whitening filter device can be:

$\mathrm{\&Lambda;}={\&Sigma;}_N^{-1}C---\left(7\right)$

Wherein, Σ _nfor the average coherence matrix in other regions.For other regions, it is (L, σ that Λ obeys parameter ²) gamma distribution, that is:

$p\left(\mathrm{\&Lambda;}\right|H_0)=\frac{1}{\mathrm{\&Gamma;}\left(L\right)}{\left(\frac{L}{{\mathrm{\&sigma;}}^2}\right)}^L{\mathrm{\&Lambda;}}^{L-1}{e}^{-\frac{L\mathrm{\&Lambda;}}{{\mathrm{\&sigma;}}^2}},\mathrm{\&Lambda;}\&GreaterEqual;0---\left(8\right)$

Wherein, L is equivalent number, σ ²for average power, H ₀represent other regions of image.

In one embodiment of the invention, can given detection threshold be γ, if Λ > is γ, then decision-point target exists.

Fig. 3 is the pcolor of the Polarimetric SAR Image in the Singapore area comprising multiple bridge according to an embodiment of the invention, and wherein 1 ~ 14 labelled notation region is bridge, and the resolution of this image is 4.73 meters × 4.80 meters, and image size is 5491 × 2156 pixels.With above-mentioned steps, bridge machinery is carried out to this image, wherein, curve regularisation parameter can be set to 0.2, iterations can be set to 100, the elemental area threshold value of setting can be 1000 pixels, the Maximum tolerance preset can be set to 10 pixels, and the threshold value of the size of bridge can be wide 30 pixels, long 150 pixels.Fig. 4 is the result of according to an embodiment of the invention image in Fig. 3 being carried out to bridge machinery according to step S101-S103, for the Bridge object of 14 in Fig. 3, propose algorithm correctly to detect 13 (No. 1-10 and No. 12-14), false-alarm 3 (No. 15-17), undetected 1 (No. 11).Fig. 5 shows the strong scatterer bridge distinguished according to step S104, as shown in Figure 5, has finally distinguished 1,3,4,8 and No. 12 strong scatterer bridge.

According to the Polarimetric SAR Image Bridges Detection based on level-set segmentation of the embodiment of the present invention, by carrying out level-set segmentation to Polarimetric SAR Image to obtain land and waters, and extract the unique point of the profile in waters, according to unique point determination suspected bridge area, then CFAR detection is carried out to the suspected bridge area after rejecting false-alarm, thus distinguish strong scatterer bridge.Thus, by the method for level-set segmentation, land and waters can be obtained more accurately, then the process such as rejecting false-alarm and CFAR detection etc. that combines distinguishes strong scatterer bridge, greatly can improve the accuracy of bridge machinery.

For realizing above-described embodiment, the present invention also proposes a kind of Polarimetric SAR Image bridge detecting device based on level-set segmentation.

Fig. 6 is according to an embodiment of the invention based on the structured flowchart of the Polarimetric SAR Image bridge detecting device of level-set segmentation.

As shown in Figure 6, the Polarimetric SAR Image bridge detecting device based on level-set segmentation of the embodiment of the present invention, comprising: segmentation module 10, determination module 20, rejecting module 30 and detection module 40.

Wherein, segmentation module 10, for carrying out level-set segmentation according to Region Statistical Features to Polarimetric SAR Image, obtains land and waters; Determination module 20 for extracting the unique point of the profile in waters, and determines the area-of-interest in land by the distance of unique point, using as suspected bridge area; Reject module 30 and realize bridge machinery for the false-alarm rejecting suspected bridge area; Detection module 40, for carrying out CFAR detection to the suspected bridge area after rejecting false-alarm, distinguishes strong scatterer bridge.

Be described to look Polarimetric SAR Image in an embodiment of the present invention more.Look Polarimetric SAR Image scattering matrix more and obey multiple Wishart (Vichy is special) distribution, for homogeneous region, the coherence matrix T obtained by scattering matrix vector quantization also obeys multiple Wishart and distributes.If the average of the coherence matrix of homogeneous region is Σ, be L depending on number, and POLARIZATION CHANNEL number is p, then can be designated as T ~ W (Σ, L, p).Thus, the probability density function of above-mentioned correlation matrix is:

$f\left(T\right|\&Sigma;,L,p)=\frac{L^{pL}{\left(\right|T\left|\right)}^{L-p}\exp \{-Ltr\left({\&Sigma;}^{-1}T\right)\}}{K(L,p){\left(\right|\&Sigma;\left|\right)}^L}---\left(1\right)$

Wherein, $K(L,p)={\mathrm{\&pi;}}^{\frac{p(p-1)}{2}}G\left(L\right)\mathrm{...}G(L-p+1),$ G () is gamma (Gamma) function.

In one embodiment of the invention, represent the plane of delineation with R, represent given Polarimetric SAR Image, represent segmentation with P with I, represent the boundary line in land and waters with Γ, wherein, Γ is closed curve, meanwhile, with R ₁and R ₂represent the land and waters split by curve Γ respectively, then, under splitting P, the posterior probability of Polarimetric SAR Image I is p (I| Ρ (R ₁, R ₂)).According to bayesian criterion, when posterior probability is maximum, be divided into optimal segmentation, namely optimal segmentation the condition that meets be:

$P=\underset{R_1,R_2}{\max }p\left(I\right|P\left(R_1,R_2\right))---\left(2\right)$

Again because p (I|P) ∝ p (P|I) p (P), if land R ₁with waters R ₂separate, area condition probability function is f (I _i| R _i), wherein i is 1 or 2, represents the probability density function of image I when being divided into land or waters respectively.Segmentation prior probability p (P) may be defined as outline line length function p (S) ∝ e ^{-ν | Γ |}, ν >0, then can split the energy function of curve and the equivalent form of value of parted pattern:

$E(\mathrm{\&Gamma;},\{R_1,R_2\left\}\right)=v\left|\mathrm{\&Gamma;}\right|-\underset{R_1}{\&Integral;}\log f\left(x\right|R_1)dx-\underset{R_2}{\&Integral;}\log f\left(x\right|R_2)dx$

$\widehat{\mathrm{\&Gamma;}}=\underset{\mathrm{\&Gamma;}}{\min }\left\{E(\mathrm{\&Gamma;},\{R_1,R_2\left\}\right)\right\}---\left(3\right)$

Wherein ν is curve gauge parameter, | Γ | represent length of curve, for optimal segmentation curve.

In one embodiment of the invention, if by level set function Φ (c (t), t) represent curve Γ, then zero level collection homologous thread Γ (t)=c (t) | Φ (c (t), t)=0}.Meanwhile, the energy definition of level set functional Φ is obtained according to curve energy definition:

$E\left(\mathrm{\&Phi;}\right)=-\underset{R}{\&Integral;}\left(H\right(\mathrm{\&Phi;}\left)log\right(f\left(x|R_1\right))+(1-H\left(\mathrm{\&Phi;}\right)\left)log\right(f\left(x|R_2\right)\left)\right)dx+v\underset{R}{\&Integral;}|\&dtri;H\left(\mathrm{\&Phi;}\right)|dx---\left(4\right)$

Wherein, H (Φ) is step function, when Φ >=0, and H (Φ)=1, as Φ <0, H (Φ)=0.R ₁corresponding region, Φ>=0, R ₂corresponding Φ <0 region.

The partial differential equation of level set functional Φ is:

$\frac{\&part;\mathrm{\&Phi;}}{\&part;t}=-\mathrm{\&delta;}\left(\mathrm{\&Phi;}\right)(v\mathrm{\&kappa;}+log\frac{f\left(x\right|R_2)}{f\left(x\right|R_1)})---\left(5\right)$

Wherein δ (Φ) is impulse function, for curvature of curve.Then can according to formula (5), and by the variational method along level set energy function negative gradient direction Approach by inchmeal, solve the zero level set function under the minimum condition of energy.In one embodiment of the invention, if the mean value of the coherence matrix on waters and land is respectively Σ ₁, Σ ₂, then level set evolution function is:

$\frac{\&part;\mathrm{\&Phi;}}{\&part;t}=-\mathrm{\&delta;}\left(\mathrm{\&Phi;}\right)(v\mathrm{\&kappa;}+L(log\left|{\mathrm{\&Sigma;}}_1\right|+tr\left({{\mathrm{\&Sigma;}}_1}^{-1}T\right))-L(log\left|{\mathrm{\&Sigma;}}_2\right|+tr\left({{\mathrm{\&Sigma;}}_2}^{-1}T\right)\left)\right)---\left(6\right)$

Wherein, Σ ₁, Σ ₂obtain by possibility predication, under the initialization curve Γ and corresponding Φ and parameter value ν of setting, utilize formula (6) to carry out iteration, until when zero level set function no longer changes, can level-set segmentation be realized.

After realizing level-set segmentation, can judge the classification in two regions according to the average scattering watt level in two regions of segmentation, generally speaking, the average scattering power in waters is little.Because each region contour of level-set segmentation closes continuously, for land and water segmentation binary map, the region of all connections can be obtained by 8 connected domain evaluation algorithm.

After above-mentioned primary segmentation, there is the very little waters of some areas and land in segmentation result, for the not intensive especially situation of bridge, these small size regions think irrelevant with bridge machinery, can reject under the elemental area threshold value of setting, threshold value can be determined according to image size and resolution.Must arrive waters thus still has some by the region of lower scattering strength split by mistake, if the coupling part in these regions is apart from little, can form some suspected bridge area.Consider that detecting bridge is positioned on main sea or divarication, carries out waters merging by the spacing of each waters profile, and according to the width of bridge setting reasonable threshold value, extract with main body waters and Qi Ge branch apart near waters part.

Particularly, determination module 20 extracts the unique point of the profile in waters by digital curve division conflation algorithm.In one embodiment of the invention, digital curve division conflation algorithm can select DP (Douglas-Peucker) algorithm, and its idiographic flow is: the two ends selecting the contour curve section in waters are initial characteristics point; Obtain the air line distance of points all on segment of curve to initial characteristics point; If points all on segment of curve is all less than default Maximum tolerance to the air line distance of initial characteristics point, then determine that this initial characteristics point is the unique point of the profile in waters, otherwise, the point maximum with the air line distance of initial characteristics point is selected to be cut-point, and with this cut-point, segment of curve is divided into two sections, then respectively above-mentioned algorithm is adopted, until determine all unique points of the contour curve section in waters to the segment of curve recurrence after these two sections segmentations.

Fig. 2 shows the unique point of waters profile, and wherein, black region is land, and white portion is waters, and the point of " * " place mark is the unique point of waters profile.Suppose all unique point { u defining the profile in A waters according to above-mentioned algorithm _iand all unique point { v of profile in B waters _i, can { u be calculated _iand { v _ibetween the distance of each unique point.In an embodiment of the present invention, can by two the unique point { ps of distance lower than predeterminable range _iand { q _icomposition characteristic point pair, and determine waters bridge end points by the coordinate relation of unique point each in multiple feature point pairs.Further, according to the size setting threshold value of bridge, feature point pairs can be merged, such as, the feature point pairs of all distances lower than setting threshold value can be merged, thus when there is multiple bridge between A and B waters, the bridge of different length can be distinguished.If the correspondence set of A and B two waters is respectively { { S after merging ₁..., { S _nand { { W ₁..., { W _n, then { S _iand { W _ia new feature point pairs can be formed.Thus, new feature point pairs { S can be determined further _iand { W _irepresentated by waters bridge end points, and according to end points determination area-of-interest.For example, as shown in Figure 2, if waters bridge end points is four points, then the quadrilateral area that four points form is area-of-interest, and this area-of-interest is suspected bridge area.In a particular embodiment of the present invention, Maximum tolerance, predeterminable range and the bridge length threshold value preset can need according to the engineer's scale of Polarimetric SAR Image and concrete detection and set.Thus, by by two the unique points composition feature point pairs of distance lower than predeterminable range, can prevent from calculating all unique points and the calculated amount that causes area-of-interest that is excessive and that determine is comparatively large, thus speed and the precision of bridge machinery can be improved to a certain extent.

By still existence part in the waters that obtains of segmentation module 10 and main body waters and branch thereof apart from inner waters, near land, the land between these waters and other waters is divided into suspected bridge area by mistake, namely becomes false-alarm targets.Usually these false-alarm targets are compared with true bridge, and contour shape is irregular, and the pontic region determined by end points is not communicated with, and therefore, rejecting module 30 using disconnected area-of-interest as false-alarm, and can be rejected.

Usually, there is guardrail wires around important bridge, the scattering strength of guardrail wires is higher, and bridge floor, dihedral angle between guardrail and waters or trihedral angle scattering can make pontic regions scatter intensity higher than other Bridge object and false-alarm targets.Therefore, from suspected bridge area, bridge area is distinguished by scattering strength.In one embodiment of the invention, detection module 40 can be point target constant false alarm detector.Particularly, using the suspected bridge area after rejecting false-alarm as point target, and point target can be represented by zone leveling coherence matrix, and by polarimetric whitening filter device, point target be detected, thus can effectively detect by the bridge higher to scattering strength.

More specifically, if the average coherence matrix of rejecting the suspected bridge area after false-alarm is C, then polarimetric whitening filter device can be:

$\mathrm{\&Lambda;}={\&Sigma;}_N^{-1}C---\left(7\right)$

Wherein, Σ _nfor the average coherence matrix in other regions.For other regions, it is (L, σ that Λ obeys parameter ²) gamma distribution, that is:

$p\left(\mathrm{\&Lambda;}\right|H_0)=\frac{1}{\mathrm{\&Gamma;}\left(L\right)}{\left(\frac{L}{{\mathrm{\&sigma;}}^2}\right)}^L{\mathrm{\&Lambda;}}^{L-1}{e}^{-\frac{L\mathrm{\&Lambda;}}{{\mathrm{\&sigma;}}^2}},\mathrm{\&Lambda;}\&GreaterEqual;0---\left(8\right)$

Wherein, L is equivalent number, σ ²for average power, H ₀represent other regions of image.

In one embodiment of the invention, can given detection threshold be γ, if Λ > is γ, then decision-point target exists.

Fig. 3 is the pcolor of the Polarimetric SAR Image in the Singapore area comprising multiple bridge according to an embodiment of the invention, and wherein 1 ~ 14 labelled notation region is bridge, and the resolution of this image is 4.73 meters × 4.80 meters, and image size is 5491 × 2156 pixels.By said apparatus, bridge machinery is carried out to this image, wherein, curve regularisation parameter can be set to 0.2, iterations can be set to 100, the elemental area threshold value of setting can be 1000 pixels, the Maximum tolerance preset can be set to 10 pixels, and the threshold value of the size of bridge can be wide 30 pixels, long 150 pixels.Fig. 4 is the result of according to an embodiment of the invention image in Fig. 3 being carried out to bridge machinery, for the Bridge object of 14 in Fig. 3, propose algorithm correctly to detect 13 (No. 1-10 and No. 12-14), false-alarm 3 (No. 15-17), undetected 1 (No. 11).Fig. 5 shows the strong scatterer bridge that detection module 40 is distinguished, and as shown in Figure 5, has finally distinguished 1,3,4,8 and No. 12 strong scatterer bridge.

According to the Polarimetric SAR Image bridge detecting device based on level-set segmentation of the embodiment of the present invention, by carrying out level-set segmentation to Polarimetric SAR Image to obtain land and waters, and extract the unique point of the profile in waters, according to unique point determination suspected bridge area, then CFAR detection is carried out to the suspected bridge area after rejecting false-alarm, thus distinguish strong scatterer bridge.Thus, by level-set segmentation, land and waters can be obtained more accurately, then combination rejecting false-alarm and CFAR detection etc. distinguish strong scatterer bridge, greatly can improve the accuracy of bridge machinery.

In describing the invention, it will be appreciated that, term " " center ", " longitudinal direction ", " transverse direction ", " length ", " width ", " thickness ", " on ", D score, " front ", " afterwards ", " left side ", " right side ", " vertically ", " level ", " top ", " end ", " interior ", " outward ", " clockwise ", " counterclockwise ", " axis ", " radial direction ", orientation or the position relationship of the instruction such as " circumference " are based on orientation shown in the drawings or position relationship, only the present invention for convenience of description and simplified characterization, instead of indicate or imply that the device of indication or element must have specific orientation, with specific azimuth configuration and operation, therefore limitation of the present invention can not be interpreted as.

In addition, term " first ", " second " only for describing object, and can not be interpreted as instruction or hint relative importance or imply the quantity indicating indicated technical characteristic.Thus, be limited with " first ", the feature of " second " can express or impliedly comprise one or more these features.In describing the invention, the implication of " multiple " is two or more, unless otherwise expressly limited specifically.

In the present invention, unless otherwise clearly defined and limited, the term such as term " installation ", " being connected ", " connection ", " fixing " should be interpreted broadly, and such as, can be fixedly connected with, also can be removably connect, or integral; Can be mechanical connection, also can be electrical connection; Can be directly be connected, also indirectly can be connected by intermediary, can be the connection of two element internals or the interaction relationship of two elements.For the ordinary skill in the art, above-mentioned term concrete meaning in the present invention can be understood as the case may be.

In the present invention, unless otherwise clearly defined and limited, fisrt feature second feature " on " or D score can be that the first and second features directly contact, or the first and second features are by intermediary indirect contact.And, fisrt feature second feature " on ", " top " and " above " but fisrt feature directly over second feature or oblique upper, or only represent that fisrt feature level height is higher than second feature.Fisrt feature second feature " under ", " below " and " below " can be fisrt feature immediately below second feature or tiltedly below, or only represent that fisrt feature level height is less than second feature.

In the description of this instructions, specific features, structure, material or feature that the description of reference term " embodiment ", " some embodiments ", " example ", " concrete example " or " some examples " etc. means to describe in conjunction with this embodiment or example are contained at least one embodiment of the present invention or example.In this manual, to the schematic representation of above-mentioned term not must for be identical embodiment or example.And the specific features of description, structure, material or feature can combine in one or more embodiment in office or example in an appropriate manner.In addition, when not conflicting, the feature of the different embodiment described in this instructions or example and different embodiment or example can carry out combining and combining by those skilled in the art.

Although illustrate and describe embodiments of the invention above, be understandable that, above-described embodiment is exemplary, can not be interpreted as limitation of the present invention, and those of ordinary skill in the art can change above-described embodiment within the scope of the invention, revises, replace and modification.

Claims (8)

1., based on a Polarimetric SAR Image Bridges Detection for level-set segmentation, it is characterized in that, comprise the following steps:

According to Region Statistical Features, level-set segmentation is carried out to Polarimetric SAR Image, obtain land and waters;

Extract the unique point of the profile in described waters, and determine the area-of-interest in described land by the distance of described unique point, using as suspected bridge area;

The false-alarm rejected in described suspected bridge area realizes bridge machinery;

CFAR detection is carried out to the described suspected bridge area after rejecting false-alarm, distinguishes strong scatterer bridge.

2. the Polarimetric SAR Image Bridges Detection based on level-set segmentation according to claim 1, is characterized in that, is extracted the unique point of the profile in described waters by digital curve division conflation algorithm.

3. the Polarimetric SAR Image Bridges Detection based on level-set segmentation according to claim 1, is characterized in that, using disconnected described area-of-interest as described false-alarm.

4. the Polarimetric SAR Image Bridges Detection based on level-set segmentation according to claim 1, is characterized in that, describedly carries out CFAR detection specifically comprise rejecting the described suspected bridge area after false-alarm:

Using the described suspected bridge area after described rejecting false-alarm as point target, and represent described point target by zone leveling coherence matrix, and by polarimetric whitening filter device, described point target is detected.

5., based on a Polarimetric SAR Image bridge detecting device for level-set segmentation, it is characterized in that, comprising:

Segmentation module, for carrying out level-set segmentation according to Region Statistical Features to Polarimetric SAR Image, obtains land and waters;

Determination module, for extracting the unique point of the profile in described waters, and determines the area-of-interest in described land by the distance of described unique point, using as suspected bridge area;

Reject module, realize bridge machinery for the false-alarm rejecting described suspected bridge area;

Detection module, for carrying out CFAR detection to the described suspected bridge area after rejecting false-alarm, distinguishes strong scatterer bridge.

6. the Polarimetric SAR Image bridge detecting device based on level-set segmentation according to claim 5, is characterized in that, described determination module extracts the unique point of the profile in described waters by digital curve division conflation algorithm.

7. the Polarimetric SAR Image bridge detecting device based on level-set segmentation according to claim 5, is characterized in that, described rejecting module using disconnected described area-of-interest as described false-alarm.

8. the Polarimetric SAR Image bridge detecting device based on level-set segmentation according to claim 5, is characterized in that, described detection module specifically for:

Using the described suspected bridge area after described rejecting false-alarm as point target, and represent described point target by zone leveling coherence matrix, and by polarimetric whitening filter device, described point target is detected.