CN105528787A

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

Info

Publication number: CN105528787A
Application number: CN201510887805.7A
Authority: CN
Inventors: 刘春�; 殷君君; 杨健
Original assignee: Tsinghua University
Current assignee: Tsinghua University
Priority date: 2015-12-07
Filing date: 2015-12-07
Publication date: 2016-04-27

Abstract

The present invention proposes a bridge detection method and device for polarimetric SAR images based on level set segmentation, wherein the method includes the following steps: performing level set segmentation on polarimetric SAR images according to regional statistical characteristics to obtain land and water areas; extracting water areas The feature points of the outline, and determine the area of interest in the land through the distance of the feature points, as the suspected bridge area; remove the false alarm in the suspected bridge area to realize bridge detection; perform constant false alarm on the suspected bridge area after removing the false alarm Detect, differentiate strong scatterer bridges. According to the method of the embodiment of the present invention, the accuracy of bridge detection can be improved.

Description

Method and device for bridge detection in polarimetric SAR images based on level set segmentation

技术领域technical field

本发明涉及图像处理技术领域，特别涉及一种基于水平集分割的极化SAR图像桥梁检测方法和装置。The present invention relates to the technical field of image processing, in particular to a level set segmentation-based polarization SAR image bridge detection method and device.

背景技术Background technique

能够在图像中检测出桥梁对于地理数据库的更新、自然灾害的评估及军事计划的制定等都具有极其重要的意义。在相关技术中，可通过边缘检测或Randon变换等方法对极化SAR(SyntheticApertureRadar，合成孔径雷达)图像进行桥梁检测，但检测的准确度大多较低。Being able to detect bridges in images is extremely important for updating geographic databases, assessing natural disasters, and formulating military plans. In related technologies, bridge detection can be performed on polarimetric SAR (Synthetic Aperture Radar, Synthetic Aperture Radar) images by methods such as edge detection or Randon transform, but the detection accuracy is mostly low.

发明内容Contents of the invention

本发明旨在至少在一定程度上解决相关技术中的技术问题之一。为此，本发明的一个目的在于提出一种基于水平集分割的极化SAR图像桥梁检测方法，能够提高桥梁检测的准确度。The present invention aims to solve one of the technical problems in the related art at least to a certain extent. Therefore, an object of the present invention is to propose a level set segmentation-based bridge detection method for polarimetric SAR images, which can improve the accuracy of bridge detection.

本发明的第二个目的在于提出一种基于水平集分割的极化SAR图像桥梁检测装置。The second object of the present invention is to propose a bridge detection device for polarization SAR images based on level set segmentation.

根据本发明第一方面实施例的基于水平集分割的极化SAR图像桥梁检测方法，包括以下步骤：根据区域统计特性对极化SAR图像进行水平集分割，得到陆地和水域；提取所述水域的轮廓的特征点，并通过所述特征点的距离确定所述陆地中的感兴趣区域，以作为疑似桥梁区域；剔除所述疑似桥梁区域中的虚警实现桥梁检测；对剔除虚警后的所述疑似桥梁区域进行恒虚警检测，区分强散射体桥梁。The level set segmentation based polarimetric SAR image bridge detection method according to the embodiment of the first aspect of the present invention includes the following steps: performing level set segmentation on the polarimetric SAR image according to regional statistical characteristics to obtain land and water areas; extracting the water area The feature points of the outline, and determine the region of interest in the land by the distance of the feature points, as the suspected bridge area; remove the false alarm in the suspected bridge area to achieve bridge detection; remove the false alarm for all Constant false alarm detection is performed on the suspected bridge area to distinguish bridges with strong scatterers.

根据本发明实施例的基于水平集分割的极化SAR图像桥梁检测方法，通过对极化SAR图像进行水平集分割以得到陆地和水域，并提取水域的轮廓的特征点，根据特征点确定疑似桥梁区域，然后对剔除虚警后的疑似桥梁区域进行恒虚警检测，从而区分出强散射体桥梁。由此，通过水平集分割的方法，能够更精确地得到陆地和水域，再结合剔除虚警和恒虚警检测等过程来区分出强散射体桥梁，能够大大提高桥梁检测的准确度。According to the level set segmentation based polarimetric SAR image bridge detection method of the embodiment of the present invention, the polarimetric SAR image is segmented by level set to obtain land and water, and feature points of the outline of the water are extracted, and the suspected bridge is determined according to the feature points area, and then perform constant false alarm detection on the suspected bridge area after removing false alarms, so as to distinguish strong scatterer bridges. Therefore, through the method of level set segmentation, the land and water areas can be obtained more accurately, and combined with the process of eliminating false alarms and constant false alarm detection to distinguish bridges with strong scatterers, the accuracy of bridge detection can be greatly improved.

另外，根据本发明上述实施例的基于水平集分割的极化SAR图像桥梁检测方法还可以具有如下附加的技术特征：In addition, the method for detecting bridges in polarized SAR images based on level set segmentation according to the above-mentioned embodiments of the present invention may also have the following additional technical features:

根据本发明的一个实施例，通过数字曲线分裂归并算法提取所述水域的轮廓的特征点。According to an embodiment of the present invention, the feature points of the outline of the water area are extracted through a digital curve splitting and merging algorithm.

根据本发明的一个实施例，将不连通的所述感兴趣区域作为所述虚警。According to an embodiment of the present invention, the disconnected ROI is used as the false alarm.

根据本发明的一个实施例，所述对剔除虚警后的所述疑似桥梁区域进行恒虚警检测具体包括：将所述剔除虚警后的所述疑似桥梁区域作为点目标，并通过区域平均相干矩阵表示所述点目标，并通过极化白化滤波器对所述点目标进行检测。According to an embodiment of the present invention, the performing constant false alarm detection on the suspected bridge area after removing false alarms specifically includes: taking the suspected bridge area after removing false alarms as a point target, and The coherence matrix represents the point target, and the point target is detected through a polarization whitening filter.

根据本发明第二方面实施例的基于水平集分割的极化SAR图像桥梁检测装置，包括：分割模块，用于根据区域统计特性对极化SAR图像进行水平集分割，得到陆地和水域；确定模块，用于提取所述水域的轮廓的特征点，并通过所述特征点的距离确定所述陆地中的感兴趣区域，以作为疑似桥梁区域；剔除模块，用于剔除所述疑似桥梁区域的虚警实现桥梁检测；检测模块，用于对剔除虚警后的所述疑似桥梁区域进行恒虚警检测，区分强散射体桥梁。According to the embodiment of the second aspect of the present invention, the polarimetric SAR image bridge detection device based on level set segmentation includes: a segmentation module, which is used to perform level set segmentation on the polarimetric SAR image according to regional statistical characteristics, to obtain land and water; a determination module , for extracting the feature points of the outline of the water area, and determining the region of interest in the land through the distance of the feature points as the suspected bridge area; The bridge detection is realized by alarming; the detection module is used for performing constant false alarm detection on the suspected bridge area after eliminating false alarms, and distinguishing strong scatterer bridges.

根据本发明实施例的基于水平集分割的极化SAR图像桥梁检测装置，通过对极化SAR图像进行水平集分割以得到陆地和水域，并提取水域的轮廓的特征点，根据特征点确定疑似桥梁区域，然后对剔除虚警后的疑似桥梁区域进行恒虚警检测，从而区分出强散射体桥梁。由此，通过水平集分割，能够更精确地得到陆地和水域，再结合剔除虚警和恒虚警检测等来区分出强散射体桥梁，能够大大提高桥梁检测的准确度。According to the level set segmentation based polarimetric SAR image bridge detection device of the embodiment of the present invention, the polarimetric SAR image is level set segmented to obtain land and water areas, and feature points of the outline of the water area are extracted, and suspected bridges are determined according to the feature points area, and then perform constant false alarm detection on the suspected bridge area after removing false alarms, so as to distinguish strong scatterer bridges. Therefore, through level set segmentation, the land and water areas can be obtained more accurately, and combined with false alarm detection and constant false alarm detection to distinguish strong scatterer bridges, the accuracy of bridge detection can be greatly improved.

另外，根据本发明上述实施例的基于水平集分割的极化SAR图像桥梁检测装置还可以具有如下附加的技术特征：In addition, the level set segmentation based polarimetric SAR image bridge detection device according to the above embodiments of the present invention may also have the following additional technical features:

根据本发明的一个实施例，所述确定模块通过数字曲线分裂归并算法提取所述水域的轮廓的特征点。According to an embodiment of the present invention, the determination module extracts the feature points of the outline of the water area through a digital curve splitting and merging algorithm.

根据本发明的一个实施例，所述剔除模块将不连通的所述感兴趣区域作为所述虚警。According to an embodiment of the present invention, the elimination module uses the disconnected ROI as the false alarm.

根据本发明的一个实施例，所述检测模块具体用于：将所述剔除虚警后的所述疑似桥梁区域作为点目标，并通过区域平均相干矩阵表示所述点目标，并通过极化白化滤波器对所述点目标进行检测。According to an embodiment of the present invention, the detection module is specifically configured to: use the suspected bridge area after removing false alarms as a point target, and represent the point target through the area average coherence matrix, and use polarization whitening A filter detects the point objects.

附图说明Description of drawings

图1为根据本发明一个实施例的基于水平集分割的极化SAR图像桥梁检测方法的流程图；Fig. 1 is the flow chart of the polarized SAR image bridge detection method based on level set segmentation according to one embodiment of the present invention;

图2为根据本发明一个实施例的水域的轮廓的特征点和疑似桥梁区域的示意图；Fig. 2 is a schematic diagram of feature points and suspected bridge areas of the outline of a water area according to an embodiment of the present invention;

图3为根据本发明一个实施例的包括多个桥梁的新加坡地区的极化SAR图像的伪彩图；FIG. 3 is a pseudo-color map of a polarimetric SAR image of the Singapore region including multiple bridges according to one embodiment of the present invention;

图4为根据本发明一个实施例的对图3中的图像进行桥梁检测的结果示意图；Fig. 4 is a schematic diagram of the results of bridge detection on the image in Fig. 3 according to an embodiment of the present invention;

图5为根据本发明一个实施例的区分出强散射体桥梁的结果示意图；Fig. 5 is a schematic diagram of the results of distinguishing strong scatterer bridges according to an embodiment of the present invention;

图6为根据本发明一个实施例的基于水平集分割的极化SAR图像桥梁检测装置的结构框图。Fig. 6 is a structural block diagram of a bridge detection device for polarimetric SAR images based on level set segmentation according to an embodiment of the present invention.

具体实施方式detailed description

下面详细描述本发明的实施例，所述实施例的示例在附图中示出，其中自始至终相同或类似的标号表示相同或类似的元件或具有相同或类似功能的元件。下面通过参考附图描述的实施例是示例性的，旨在用于解释本发明，而不能理解为对本发明的限制。Embodiments of the present invention are described in detail below, examples of which are shown in the drawings, wherein the same or similar reference numerals designate the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the figures are exemplary and are intended to explain the present invention and should not be construed as limiting the present invention.

下面参考附图描述本发明实施例的基于水平集分割的极化SAR图像桥梁检测方法和装置。The method and device for detecting bridges in polarized SAR images based on level set segmentation according to the embodiments of the present invention will be described below with reference to the accompanying drawings.

图1为根据本发明一个实施例的基于水平集分割的极化SAR图像桥梁检测方法的流程图。FIG. 1 is a flow chart of a method for detecting bridges in polarimetric SAR images based on level set segmentation according to an embodiment of the present invention.

如图1所示，本发明实施例的基于水平集分割的极化SAR图像桥梁检测方法，包括以下步骤：As shown in Figure 1, the polarized SAR image bridge detection method based on level set segmentation of the embodiment of the present invention comprises the following steps:

S101，根据区域统计特性对极化SAR图像进行水平集分割，得到陆地和水域。S101. Perform level set segmentation on the polarimetric SAR image according to regional statistical characteristics to obtain land and water areas.

在本发明的实施例中以多视极化SAR图像为例进行说明。多视极化SAR图像散射矩阵服从复Wishart(维希特)分布，对于同质区域，由散射矩阵矢量化而得到的相干矩阵T也服从复Wishart分布。如果同质区域的相干矩阵的均值为Σ，视数为L，且极化通道数为p，则可记为T～W(Σ,L,p)。由此，上述相关矩阵的概率密度函数为：In the embodiments of the present invention, a multi-view polarization SAR image is taken as an example for illustration. The scattering matrix of the multi-view polarization SAR image obeys the complex Wishart distribution, and for the homogeneous area, the coherence matrix T obtained by the vectorization of the scattering matrix also obeys the complex Wishart distribution. If the mean of the coherence matrix in the homogeneous area is Σ, the view number is L, and the number of polarization channels is p, it can be recorded as T~W(Σ,L,p). Therefore, the probability density function of the above correlation matrix is:

$f f ((T T | | Σ Σ,, L L,, p p)) = = \frac{{L L}^{p p L L} {((| | T T | |))}^{L L - - p p} exp exp {{- - L L t t r r (({Σ Σ}^{- - 11} T T))}}}{K K ((L L,, p p)) {((| | Σ Σ | |))}^{L L}} - - - - - - ((11))$

其中， $K (L, p) = π^{\frac{p (p - 1)}{2}} G (L) ... G (L - p + 1),$ G(·)为伽马(Gamma)函数。in, $K (L, p) = π^{\frac{p (p - 1)}{2}} G (L) ... G (L - p + 1),$ G(·) is a Gamma function.

在本发明的一个实施例中，以R表示图像平面，以I表示给定的极化SAR图像，以P表示分割，以Γ表示陆地和水域的交界线，其中，Γ为闭合的曲线，同时，以R₁和R₂分别表示由曲线Γ分割的陆地和水域，则分割P下，极化SAR图像I的后验概率为p(I|Ρ(R₁,R₂))。根据贝叶斯准则，后验概率最大时的分割为最佳分割，即最佳分割所满足的条件为：In one embodiment of the present invention, R represents the image plane, I represents a given polarized SAR image, P represents segmentation, and Γ represents the boundary line between land and water, where Γ is a closed curve, and , let R ₁ and R ₂ respectively denote the land and water divided by the curve Γ, then under the division P, the posterior probability of the polarimetric SAR image I is p(I|Ρ(R ₁ ,R ₂ )). According to the Bayesian criterion, the segmentation when the posterior probability is the largest is the optimal segmentation, that is, the condition that the optimal segmentation satisfies is:

$P P = = \underset{{R R}_{11},, {R R}_{22}}{max max} p p ((I I | | P P (({R R}_{11},, {R R}_{22})))) - - - - - - ((22))$

又因为p(I|P)∝p(P|I)p(P)，若陆地R₁与水域R₂相互独立，区域条件概率函数为f(I_i|R_i)，其中i为1或2，分别表示被划分为陆地或水域时图像I的概率密度函数。分割先验概率p(P)可定义为轮廓线长度函数p(S)∝e^-ν|Γ|,ν>0，则可得分割曲线的能量函数及分割模型的等价形式：And because p(I|P)∝p(P|I)p(P), if the land R ₁ and the water R ₂ are independent of each other, the regional conditional probability function is f(I _i |R _i ), where i is 1 or 2, represent the probability density function of image I when it is divided into land or water, respectively. The segmentation prior probability p(P) can be defined as the contour length function p(S)∝e ^-ν|Γ| , ν>0, then the energy function of the segmentation curve and the equivalent form of the segmentation model can be obtained:

$E E. ((Γ Γ,, {{{R R}_{11},, {R R}_{22}}})) = = v v | | Γ Γ | | - - \underset{{R R}_{11}}{&Integral; &Integral;} log log f f ((x x | | {R R}_{11})) d d x x - - \underset{{R R}_{22}}{&Integral; &Integral;} log log f f ((x x | | {R R}_{22})) d d x x$

$\overset{^^}{Γ Γ} = = \underset{Γ Γ}{min min} {{E E. ((Γ Γ,, {{{R R}_{11},, {R R}_{22}}}))}} - - - - - - ((33))$

其中ν为曲线规化参数，|_Γ|表示曲线长度，为最佳分割曲线。where ν is the curve normalization parameter, | _Γ | is the length of the curve, for the best segmented curve.

在本发明的一个实施例中，若通过水平集函数Φ(c(t),t)表示曲线Γ，则零水平集对应曲线Γ(t)＝{c(t)|Φ(c(t),t)＝0}。同时，根据曲线能量定义得到水平集泛函数Φ的能量定义：In one embodiment of the present invention, if the curve Γ is represented by the level set function Φ(c(t), t), then the zero level set corresponds to the curve Γ(t)={c(t)|Φ(c(t) ,t)=0}. At the same time, according to the definition of curve energy, the energy definition of the level set functional function Φ is obtained:

$E E. ((Φ Φ)) = = - - \underset{R R}{&Integral; &Integral;} ((H h ((Φ Φ)) l l o o g g ((f f ((x x | | {R R}_{11})))) + + ((11 - - H h ((Φ Φ)))) l l o o g g ((f f ((x x | | {R R}_{22})))))) d d x x + + v v \underset{R R}{&Integral; &Integral;} | | &dtri; &dtri; H h ((Φ Φ)) | | d d x x - - - - - - ((44))$

其中，H(Φ)为阶跃函数，当Φ≥0时，H(Φ)＝1，当Φ<0时，H(Φ)＝0。R₁对应Φ≥0区域，R₂对应Φ<0区域。Wherein, H(Φ) is a step function, when Φ≥0, H(Φ)=1, and when Φ<0, H(Φ)=0. R ₁ corresponds to the Φ≥0 region, and R ₂ corresponds to the Φ<0 region.

水平集泛函数Φ的偏微分方程为：The partial differential equation of the level set functional function Φ is:

$\frac{\partial \partial Φ Φ}{\partial \partial t t} = = - - δ δ ((Φ Φ)) ((v v κ κ + + l l o o g g \frac{f f ((x x | | {R R}_{22}))}{f f ((x x | | {R R}_{11}))})) - - - - - - ((55))$

其中δ(Φ)为冲激函数，为曲线曲率。则可根据式(5)，并通过变分法沿水平集能量函数负梯度方向逐次逼近，求解在能量最小的条件下的零水平集函数。在本发明的一个实施例中，若水域和陆地的相干矩阵的平均值分别为Σ₁,Σ₂，则水平集演进函数为：Where δ(Φ) is the impulse function, is the curvature of the curve. Then, according to formula (5), and through the variational method successively approaching along the negative gradient direction of the level set energy function, the zero level set function under the condition of minimum energy can be solved. In one embodiment of the present invention, if the average values of the coherence matrices of water area and land are Σ ₁ , Σ ₂ respectively, then the level set evolution function is:

$\frac{\partial \partial Φ Φ}{\partial \partial t t} = = - - δ δ ((Φ Φ)) ((v v κ κ + + L L ((l l o o g g | | {Σ Σ}_{11} | | + + t t r r (({Σ Σ}_{11}^{- - 11} T T)))) - - L L ((l l o o g g | | {Σ Σ}_{22} | | + + t t r r (({Σ Σ}_{22}^{- - 11} T T)))))) - - - - - - ((66))$

其中，Σ₁,Σ₂可通过似然估计获得，在设定的初始化曲线Γ及对应Φ和参数值ν下，利用式(6)进行迭代，直至零水平集函数不再变化时，即可实现水平集分割。Among them, Σ ₁ , Σ ₂ can be obtained by likelihood estimation, under the set initialization curve Γ and the corresponding Φ and parameter value ν, use formula (6) to iterate until the zero level set function no longer changes, then Implement level set partitioning.

在实现水平集分割后，可根据分割的两区域的平均散射功率大小判断出两区域的类别，一般而言，水域的平均散射功率小。由于水平集分割各区域轮廓连续闭合，对于水陆分割二值图，通过8连通域判断算法可得到所有连通的区域。After the level set segmentation is realized, the category of the two regions can be judged according to the average scattering power of the two regions. Generally speaking, the average scattering power of the water area is small. Since the contours of each area in level set segmentation are continuously closed, for the binary image of land and water segmentation, all connected areas can be obtained through the 8-connected area judgment algorithm.

S102，提取水域的轮廓的特征点，并通过特征点的距离确定陆地中的感兴趣区域，以作为疑似桥梁区域。S102 , extracting feature points of the outline of the water area, and determining an area of interest in the land based on the distance of the feature points as a suspected bridge area.

在经过上述的初步分割过程后，分割结果中存在一些面积很小的水域和陆地，对于桥梁不特别密集情况，这些小面积区域认为与桥梁检测无关，可在设定的像素面积阈值下剔除，阈值可根据图像大小及分辨率确定。由此得到水域仍有一些被误分割的较低散射强度的区域，这些区域的连接部分若距离小，会形成一些疑似桥梁区域。考虑到检测桥梁位于主要海面或河流分支之上，通过各水域轮廓之间距离进行水域合并，并根据桥梁的宽度设定合理阈值，提取与主体水域和其各分支距离近的水域部分。After the above preliminary segmentation process, there are some small areas of water and land in the segmentation results. For the case where bridges are not particularly dense, these small areas are considered irrelevant to bridge detection and can be eliminated under the set pixel area threshold. The threshold can be determined according to the image size and resolution. From this, it can be obtained that there are still some mis-segmented low-scattering intensity areas in the water area. If the distance between the connecting parts of these areas is small, some suspected bridge areas will be formed. Considering that the detected bridge is located on the main sea or river branch, the waters are merged according to the distance between the contours of the waters, and a reasonable threshold is set according to the width of the bridge to extract the part of the waters that is close to the main waters and its branches.

具体地，可通过数字曲线分裂归并算法提取水域的轮廓的特征点。在本发明的一个实施例中，数字曲线分裂归并算法可选择DP(Douglas-Peucker)算法，其具体流程为：选择水域的轮廓曲线段的首尾端为初始特征点；获取曲线段上所有的点到初始特征点的直线距离；如果曲线段上所有的点到初始特征点的直线距离均小于预设的最大容限，则确定该初始特征点为水域的轮廓的特征点，否则，选择与初始特征点的直线距离最大的点为分割点，并以该分割点将曲线段分成两段，然后分别对该两段分割后的曲线段递归采用上述算法，直至确定出水域的轮廓曲线段的所有特征点。Specifically, the feature points of the outline of the water area can be extracted through a digital curve splitting and merging algorithm. In one embodiment of the present invention, the digital curve splitting and merging algorithm can select DP (Douglas-Peucker) algorithm, and its specific process is: select the head and tail of the contour curve segment of the water area as initial feature points; obtain all points on the curve segment The straight-line distance to the initial feature point; if the straight-line distance from all points on the curve segment to the initial feature point is less than the preset maximum tolerance, then determine that the initial feature point is the feature point of the contour of the water area, otherwise, select the same as the initial feature point The point with the largest straight-line distance of the feature points is the split point, and the curve segment is divided into two segments by this split point, and then the above algorithm is recursively used for the two segmented curve segments until all the contour curve segments of the water area are determined. Feature points.

图2示出了水域轮廓的特征点，其中，黑色区域为陆地，白色区域为水域，“*”处所标记的点即为水域轮廓的特征点。假设依照上述算法确定出了A水域的轮廓的所有特征点{u_i}和B水域的轮廓的所有特征点{v_i}，可计算{u_i}和{v_i}之间各特征点的距离。在本发明的实施例中，可将距离低于预设距离的两个特征点{p_i}和{q_i}组成特征点对，并由多个特征点对中各特征点的坐标关系确定水域桥梁端点。进一步地，可根据桥梁的大小设定阈值将特征点对合并，例如，可将所有距离低于设定阈值的特征点对进行合并，从而在A和B水域间存在多个桥梁时，能够区分出不同长度的桥梁。若合并后A和B两水域对应集合分别为{{S₁},...,{S_N}}和{{W₁},...,{W_N}}，则{S_i}与{W_i}可组成一个新的特征点对。由此，可进一步确定新的特征点对{S_i}和{W_i}所代表的水域桥梁端点，并根据端点确定感兴趣区域。举例而言，如图2所示，若水域桥梁端点为四个点，则四个点所组成的四边形区域即为感兴趣区域，而该感兴趣区域即为疑似桥梁区域。在本发明的具体实施例中，预设的最大容限、预设距离和桥梁长度阈值可根据极化SAR图像的比例尺和具体的检测需要而设定。由此，通过将距离低于预设距离的两个特征点组成特征点对，能够防止对所有特征点进行计算而导致的计算量过大和确定的感兴趣区域较大，从而能够在一定程度上提高桥梁检测的速度和精度。Figure 2 shows the feature points of the water area outline, where the black area is the land, the white area is the water area, and the points marked with "*" are the feature points of the water area outline. Assuming that all the feature points {u _i } of the outline of the water area of A and all the feature points {v _i } of the outline of the water area of B are determined according to the above algorithm, the value of each feature point between {u _i } and {v _i } can be calculated distance. In an embodiment of the present invention, two feature points {p _i } and {q _i } whose distance is lower than a preset distance can be combined into a feature point pair, and determined by the coordinate relationship of each feature point in a plurality of feature point pairs Water bridge endpoint. Further, a threshold can be set according to the size of the bridge to merge feature point pairs. For example, all feature point pairs whose distance is lower than the set threshold can be merged, so that when there are multiple bridges between A and B waters, it is possible to distinguish Bridges of different lengths. If the corresponding sets of water areas A and B are {{S ₁ },...,{S _N }} and {{W ₁ },...,{W _N }} respectively after merging, then {S _i } and {W _i } can form a new pair of feature points. In this way, the end points of the water bridge represented by the new pair of feature points {S _i } and {W _i } can be further determined, and the region of interest can be determined according to the end points. For example, as shown in FIG. 2 , if the end points of the bridge in the water area are four points, the quadrilateral area formed by the four points is the region of interest, and the region of interest is the suspected bridge area. In a specific embodiment of the present invention, the preset maximum tolerance, preset distance and bridge length threshold can be set according to the scale of the polarimetric SAR image and specific detection requirements. Therefore, by combining two feature points whose distance is lower than the preset distance into a feature point pair, it is possible to prevent the calculation of all the feature points from being too large and determine a large region of interest, so that to a certain extent Improve the speed and accuracy of bridge inspections.

S103，剔除疑似桥梁区域中的虚警实现桥梁检测。S103, eliminating false alarms in the suspected bridge region to implement bridge detection.

在步骤S101中所得到的水域中仍然存在部分与主体水域及其分支距离近的陆地内部水域，这些水域与其他水域之间的陆地被错误划分为疑似桥梁区域，即成为虚警目标。通常这些虚警目标与真实桥梁相比，轮廓形状不规则，通过端点确定的桥体区域不连通，因此，可将不连通的感兴趣区域作为虚警，并予以剔除。In the waters obtained in step S101, there are still some land internal waters that are close to the main waters and its branches, and the land between these waters and other waters is mistakenly classified as a suspected bridge area, that is, it becomes a false alarm target. Usually, compared with the real bridge, these false alarm targets have irregular contours, and the bridge body area determined by the endpoints is not connected. Therefore, the disconnected regions of interest can be regarded as false alarms and eliminated.

S104，对剔除虚警后的疑似桥梁区域进行恒虚警检测，区分强散射体桥梁。S104, perform constant false alarm detection on the suspected bridge area after removing false alarms, and distinguish bridges with strong scatterers.

一般地，重要桥梁周围存在金属护栏，金属护栏的散射强度较高，并且桥面、护栏与水域之间的二面角或三面角散射会使得桥体区域散射强度高于其它桥梁目标和虚警目标。因此，可通过散射强度从疑似桥梁区域中区分出桥梁区域。在本发明的一个实施例中，可利用点目标恒虚警检测器进行强散射体桥梁的区分。具体地，可将剔除虚警后的疑似桥梁区域作为点目标，并通过区域平均相干矩阵表示点目标，并通过极化白化滤波器对点目标进行检测，从而能够对散射强度较高的桥梁进行有效检测。Generally, there are metal guardrails around important bridges, and the scattering intensity of metal guardrails is high, and the dihedral or trihedral angle scattering between the bridge deck, guardrails and water will make the scattering intensity of the bridge body area higher than that of other bridge targets and false alarms Target. Therefore, the bridge region can be distinguished from the suspected bridge region by the scattering intensity. In one embodiment of the present invention, the discrimination of strong scatterer bridges can be performed using a point target constant false alarm detector. Specifically, the suspected bridge area after removing false alarms can be used as a point target, and the point target is represented by the area average coherence matrix, and the point target is detected by a polarization whitening filter, so that the bridge with high scattering intensity can be detected effective detection.

更具体地，若剔除虚警后的疑似桥梁区域的平均相干矩阵为C，则极化白化滤波器可为：More specifically, if the average coherence matrix of the suspected bridge area after removing false alarms is C, then the polarization whitening filter can be:

$Λ Λ = = {Σ Σ}_{N N}^{- - 11} C C - - - - - - ((77))$

其中，Σ_N为其他区域的平均相干矩阵。对于其他区域，Λ服从参数为(L,σ²)的伽马分布，即：Among them, Σ _N is the average coherence matrix of other regions. For other regions, Λ obeys the gamma distribution with parameter (L,σ ² ), namely:

$p p ((Λ Λ | | {H h}_{00})) = = \frac{11}{Γ Γ ((L L))} {((\frac{L L}{{σ σ}^{22}}))}^{L L} {Λ Λ}^{L L - - 11} {e e}^{- - \frac{L L Λ Λ}{{σ σ}^{22}}},, Λ Λ &GreaterEqual; &Greater Equal; 00 - - - - - - ((88))$

其中，L为等效视数，σ²为平均功率，H₀表示图像的其他区域。Among them, L is the equivalent visual number, ^σ2 is the average power, and _H0 represents other regions of the image.

在本发明的一个实施例中，可给定检测阈值为γ，若Λ>γ，则判定点目标存在。In an embodiment of the present invention, the detection threshold may be given as γ, and if Λ>γ, it is determined that the point target exists.

图3为根据本发明一个实施例的包括多个桥梁的新加坡地区的极化SAR图像的伪彩图，其中1～14号标记区域为桥梁，该图像的分辨率为4.73米×4.80米，图像大小为5491×2156像素。对该图像以上述步骤进行桥梁检测，其中，曲线规则化参数可设置为0.2，迭代次数可设置为100，设定的像素面积阈值可为1000像素，预设的最大容限可设置为10像素，桥梁的大小的阈值可为宽30像素、长150像素。图4为根据本发明一个实施例的对图3中的图像依照步骤S101-S103进行桥梁检测的结果，对于图3中的14个桥梁目标，提出算法正确检测13个(1-10号和12-14号)，虚警3个(15-17号)，漏检1个(11号)。图5示出了依照步骤S104区分出的强散射体桥梁，如图5所示，最终区分出了1、3、4、8和12号强散射体桥梁。Fig. 3 is the pseudo-color map of the polarized SAR image of the Singapore region including multiple bridges according to an embodiment of the present invention, wherein No. 1 to No. 14 marked areas are bridges, and the resolution of the image is 4.73 meters * 4.80 meters, the image The size is 5491×2156 pixels. Perform bridge detection on the image with the above steps, where the curve regularization parameter can be set to 0.2, the number of iterations can be set to 100, the set pixel area threshold can be 1000 pixels, and the preset maximum tolerance can be set to 10 pixels , the threshold of the size of the bridge may be 30 pixels in width and 150 pixels in length. Fig. 4 is the result of carrying out bridge detection according to steps S101-S103 to the image in Fig. 3 according to an embodiment of the present invention. For the 14 bridge targets in Fig. 3, the proposed algorithm correctly detects 13 (No. 1-10 and No. 12 -No. 14), 3 false alarms (No. 15-17), and 1 missed detection (No. 11). FIG. 5 shows the strong scatterer bridges distinguished according to step S104 . As shown in FIG. 5 , No. 1, 3, 4, 8 and 12 strong scatterer bridges are finally distinguished.

为实现上述实施例，本发明还提出一种基于水平集分割的极化SAR图像桥梁检测装置。In order to realize the above-mentioned embodiments, the present invention also proposes a bridge detection device for polarimetric SAR images based on level set segmentation.

如图6所示，本发明实施例的基于水平集分割的极化SAR图像桥梁检测装置，包括：分割模块10、确定模块20、剔除模块30和检测模块40。As shown in FIG. 6 , the bridge detection device for polarized SAR images based on level set segmentation according to the embodiment of the present invention includes: a segmentation module 10 , a determination module 20 , an elimination module 30 and a detection module 40 .

其中，分割模块10用于根据区域统计特性对极化SAR图像进行水平集分割，得到陆地和水域；确定模块20用于提取水域的轮廓的特征点，并通过特征点的距离确定陆地中的感兴趣区域，以作为疑似桥梁区域；剔除模块30用于剔除疑似桥梁区域的虚警实现桥梁检测；检测模块40用于对剔除虚警后的疑似桥梁区域进行恒虚警检测，区分强散射体桥梁。Among them, the segmentation module 10 is used to perform level set segmentation on the polarimetric SAR image according to the regional statistical characteristics to obtain land and water; the determination module 20 is used to extract the feature points of the outline of the water area, and determine the sense of land in the land through the distance of the feature points. The area of interest is used as a suspected bridge area; the elimination module 30 is used to eliminate false alarms in the suspected bridge area to achieve bridge detection; the detection module 40 is used to perform constant false alarm detection on the suspected bridge area after removing false alarms, and distinguish strong scatterer bridges .

又因为p(I|P)∝p(P|I)p(P)，若陆地R₁与水域R₂相互独立，区域条件概率函数为f(I_i|R_i)，其中i为1或2，分别表示被划分为陆地或水域时图像I的概率密度函数。分割先验概率p(P)可定义为轮廓线长度函数p(S)∝e^-ν|Γ|,ν>0，则可得分割曲线的能量函数及分割模型的等价形式：And because p(I|P)∝p(P|I)p(P), if the land R ₁ and the water R ₂ are independent of each other, the regional conditional probability function is f(I _i |R _i ), where i is 1 or 2, represent the probability density function of image I when it is divided into land or water, respectively. The prior probability of segmentation p(P) can be defined as the contour length function p(S)∝e ^-ν|Γ| , ν>0, then the energy function of the segmentation curve and the equivalent form of the segmentation model can be obtained:

其中ν为曲线规化参数，|Γ|表示曲线长度，为最佳分割曲线。where ν is the curve normalization parameter, |Γ| is the length of the curve, for the best segmented curve.

在实现水平集分割后，可根据分割的两区域的平均散射功率大小判断出两区域的类别，一般而言，水域的平均散射功率小。由于水平集分割各区域轮廓连续闭合，对于水陆分割二值图，通过8连通域判断算法可得到所有连通的区域。After the level set segmentation is realized, the category of the two regions can be judged according to the average scattering power of the two regions. Generally speaking, the average scattering power of the water area is small. Since the contours of each region in level set segmentation are continuously closed, for the binary image of water and land segmentation, all connected regions can be obtained by the 8-connected region judgment algorithm.

在经过上述的初步分割后，分割结果中存在一些面积很小的水域和陆地，对于桥梁不特别密集情况，这些小面积区域认为与桥梁检测无关，可在设定的像素面积阈值下剔除，阈值可根据图像大小及分辨率确定。由此得到水域仍有一些被误分割的较低散射强度的区域，这些区域的连接部分若距离小，会形成一些疑似桥梁区域。考虑到检测桥梁位于主要海面或河流分支之上，通过各水域轮廓之间距离进行水域合并，并根据桥梁的宽度设定合理阈值，提取与主体水域和其各分支距离近的水域部分。After the above preliminary segmentation, there are some small areas of water and land in the segmentation results. For the case where bridges are not particularly dense, these small areas are considered irrelevant to bridge detection and can be eliminated under the set pixel area threshold. Threshold It can be determined according to the image size and resolution. From this, it can be obtained that there are still some mis-segmented low-scattering intensity areas in the water area. If the distance between the connecting parts of these areas is small, some suspected bridge areas will be formed. Considering that the detected bridge is located on the main sea or river branch, the waters are merged according to the distance between the contours of the waters, and a reasonable threshold is set according to the width of the bridge to extract the part of the waters that is close to the main waters and its branches.

具体地，确定模块20可通过数字曲线分裂归并算法提取水域的轮廓的特征点。在本发明的一个实施例中，数字曲线分裂归并算法可选择DP(Douglas-Peucker)算法，其具体流程为：选择水域的轮廓曲线段的首尾端为初始特征点；获取曲线段上所有的点到初始特征点的直线距离；如果曲线段上所有的点到初始特征点的直线距离均小于预设的最大容限，则确定该初始特征点为水域的轮廓的特征点，否则，选择与初始特征点的直线距离最大的点为分割点，并以该分割点将曲线段分成两段，然后分别对该两段分割后的曲线段递归采用上述算法，直至确定出水域的轮廓曲线段的所有特征点。Specifically, the determination module 20 may extract feature points of the outline of the water area through a digital curve splitting and merging algorithm. In one embodiment of the present invention, the digital curve splitting and merging algorithm can select DP (Douglas-Peucker) algorithm, and its specific process is: select the head and tail of the contour curve segment of the water area as initial feature points; obtain all points on the curve segment The straight-line distance to the initial feature point; if the straight-line distance from all points on the curve segment to the initial feature point is less than the preset maximum tolerance, then determine that the initial feature point is the feature point of the contour of the water area, otherwise, select the same as the initial feature point The point with the largest straight-line distance of the feature points is the split point, and the curve segment is divided into two segments by this split point, and then the above algorithm is recursively used for the two segmented curve segments until all the contour curve segments of the water area are determined. Feature points.

图2示出了水域轮廓的特征点，其中，黑色区域为陆地，白色区域为水域，“*”处所标记的点即为水域轮廓的特征点。假设依照上述算法确定出了A水域的轮廓的所有特征点{u_i}和B水域的轮廓的所有特征点{v_i}，可计算{u_i}和{v_i}之间各特征点的距离。在本发明的实施例中，可将距离低于预设距离的两个特征点{p_i}和{q_i}组成特征点对，并由多个特征点对中各特征点的坐标关系确定水域桥梁端点。进一步地，可根据桥梁的大小设定阈值将特征点对合并，例如，可将所有距离低于设定阈值的特征点对进行合并，从而在A和B水域间存在多个桥梁时，能够区分出不同长度的桥梁。若合并后A和B两水域对应集合分别为{{S₁},...,{S_N}}和{{W₁},...,{W_N}}，则{S_i}与{W_i}可组成一个新的特征点对。由此，可进一步确定新的特征点对{S_i}和{W_i}所代表的水域桥梁端点，并根据端点确定感兴趣区域。举例而言，如图2所示，若水域桥梁端点为四个点，则四个点所组成的四边形区域即为感兴趣区域，而该感兴趣区域即为疑似桥梁区域。在本发明的具体实施例中，预设的最大容限、预设距离和桥梁长度阈值可根据极化SAR图像的比例尺和具体的检测需要而设定。由此，通过将距离低于预设距离的两个特征点组成特征点对，能够防止对所有特征点进行计算而导致的计算量过大和确定的感兴趣区域较大，从而能够在一定程度上提高桥梁检测的速度和精度。Figure 2 shows the feature points of the water area outline, where the black area is the land, the white area is the water area, and the points marked with "*" are the feature points of the water area outline. Assuming that all the feature points {u _i } of the outline of the water area of A and all the feature points {v _i } of the outline of the water area of B are determined according to the above algorithm, the value of each feature point between {u _i } and {v _i } can be calculated distance. In an embodiment of the present invention, two feature points {p _i } and {q _i } whose distance is lower than a preset distance can be combined into a feature point pair, and determined by the coordinate relationship of each feature point in a plurality of feature point pairs Water bridge endpoint. Further, a threshold can be set according to the size of the bridge to merge the feature point pairs. For example, all feature point pairs whose distance is lower than the set threshold can be merged, so that when there are multiple bridges between A and B waters, it is possible to distinguish Bridges of different lengths. If the corresponding sets of water areas A and B are {{S ₁ },...,{S _N }} and {{W ₁ },...,{W _N }} respectively after merging, then {S _i } and {W _i } can form a new pair of feature points. In this way, the end points of the water bridge represented by the new pair of feature points {S _i } and {W _i } can be further determined, and the region of interest can be determined according to the end points. For example, as shown in FIG. 2 , if the end points of the bridge in the water area are four points, the quadrilateral area formed by the four points is the region of interest, and the region of interest is the suspected bridge area. In a specific embodiment of the present invention, the preset maximum tolerance, preset distance and bridge length threshold can be set according to the scale of the polarimetric SAR image and specific detection requirements. Therefore, by combining two feature points whose distance is lower than the preset distance into a feature point pair, it is possible to prevent the calculation of all the feature points from being too large and determine a large region of interest, so that to a certain extent Improve the speed and accuracy of bridge inspections.

由分割模块10所得到的水域中仍然存在部分与主体水域及其分支距离近的陆地内部水域，这些水域与其他水域之间的陆地被错误划分为疑似桥梁区域，即成为虚警目标。通常这些虚警目标与真实桥梁相比，轮廓形状不规则，通过端点确定的桥体区域不连通，因此，剔除模块30可将不连通的感兴趣区域作为虚警，并予以剔除。In the waters obtained by the segmentation module 10, there are still some land internal waters that are close to the main waters and its branches. The land between these waters and other waters is mistakenly classified as a suspected bridge area, that is, it becomes a false alarm target. Usually, compared with the real bridge, these false alarm targets have irregular contours, and the bridge body area determined by the endpoints is disconnected. Therefore, the elimination module 30 can regard disconnected regions of interest as false alarms and eliminate them.

一般地，重要桥梁周围存在金属护栏，金属护栏的散射强度较高，并且桥面、护栏与水域之间的二面角或三面角散射会使得桥体区域散射强度高于其它桥梁目标和虚警目标。因此，可通过散射强度从疑似桥梁区域中区分出桥梁区域。在本发明的一个实施例中，检测模块40可为点目标恒虚警检测器。具体地，可将剔除虚警后的疑似桥梁区域作为点目标，并通过区域平均相干矩阵表示点目标，并通过极化白化滤波器对点目标进行检测，从而能够对散射强度较高的桥梁进行有效检测。Generally, there are metal guardrails around important bridges, and the scattering intensity of metal guardrails is high, and the dihedral or trihedral angle scattering between the bridge deck, guardrails and water will make the scattering intensity of the bridge body area higher than that of other bridge targets and false alarms Target. Therefore, the bridge region can be distinguished from the suspected bridge region by the scattering intensity. In one embodiment of the present invention, the detection module 40 may be a point target constant false alarm detector. Specifically, the suspected bridge area after removing false alarms can be used as a point target, and the point target is represented by the area average coherence matrix, and the point target is detected by a polarization whitening filter, so that the bridge with high scattering intensity can be detected effective detection.

$Λ Λ = = {Σ Σ}_{N N}^{- - 11} C C - - - - - - ((77))$

图3为根据本发明一个实施例的包括多个桥梁的新加坡地区的极化SAR图像的伪彩图，其中1～14号标记区域为桥梁，该图像的分辨率为4.73米×4.80米，图像大小为5491×2156像素。对该图像通过上述装置进行桥梁检测，其中，曲线规则化参数可设置为0.2，迭代次数可设置为100，设定的像素面积阈值可为1000像素，预设的最大容限可设置为10像素，桥梁的大小的阈值可为宽30像素、长150像素。图4为根据本发明一个实施例的对图3中的图像进行桥梁检测的结果，对于图3中的14个桥梁目标，提出算法正确检测13个(1-10号和12-14号)，虚警3个(15-17号)，漏检1个(11号)。图5示出了检测模块40区分出的强散射体桥梁，如图5所示，最终区分出了1、3、4、8和12号强散射体桥梁。Fig. 3 is the pseudo-color map of the polarized SAR image of the Singapore region including multiple bridges according to an embodiment of the present invention, wherein No. 1 to No. 14 marked areas are bridges, and the resolution of the image is 4.73 meters * 4.80 meters, the image The size is 5491×2156 pixels. Perform bridge detection on the image through the above device, wherein the curve regularization parameter can be set to 0.2, the number of iterations can be set to 100, the set pixel area threshold can be 1000 pixels, and the preset maximum tolerance can be set to 10 pixels , the threshold of the size of the bridge may be 30 pixels in width and 150 pixels in length. Fig. 4 is the result of carrying out bridge detection to the image in Fig. 3 according to one embodiment of the present invention, for 14 bridge targets in Fig. 3, proposed algorithm correctly detects 13 (No. 1-10 and No. 12-14), There were 3 false alarms (No. 15-17), and 1 missed detection (No. 11). FIG. 5 shows the strong scatterer bridges identified by the detection module 40 . As shown in FIG. 5 , No. 1, 3, 4, 8 and 12 strong scatterer bridges are finally distinguished.

在本发明的描述中，需要理解的是，术语“中心”、“纵向”、“横向”、“长度”、“宽度”、“厚度”、“上”、“下”、“前”、“后”、“左”、“右”、“竖直”、“水平”、“顶”、“底”、“内”、“外”、“顺时针”、“逆时针”、“轴向”、“径向”、“周向”等指示的方位或位置关系为基于附图所示的方位或位置关系，仅是为了便于描述本发明和简化描述，而不是指示或暗示所指的装置或元件必须具有特定的方位、以特定的方位构造和操作，因此不能理解为对本发明的限制。In describing the present invention, it is to be understood that the terms "center", "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "front", " Back", "Left", "Right", "Vertical", "Horizontal", "Top", "Bottom", "Inner", "Outer", "Clockwise", "Counterclockwise", "Axial" , "radial", "circumferential" and other indicated orientations or positional relationships are based on the orientations or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying the referred device or Elements must have certain orientations, be constructed and operate in certain orientations, and therefore should not be construed as limitations on the invention.

此外，术语“第一”、“第二”仅用于描述目的，而不能理解为指示或暗示相对重要性或者隐含指明所指示的技术特征的数量。由此，限定有“第一”、“第二”的特征可以明示或者隐含地包括一个或者更多个该特征。在本发明的描述中，“多个”的含义是两个或两个以上，除非另有明确具体的限定。In addition, the terms "first" and "second" are used for descriptive purposes only, and cannot be interpreted as indicating or implying relative importance or implicitly specifying the quantity of indicated technical features. Thus, a feature defined as "first" and "second" may explicitly or implicitly include one or more of these features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

在本发明中，除非另有明确的规定和限定，术语“安装”、“相连”、“连接”、“固定”等术语应做广义理解，例如，可以是固定连接，也可以是可拆卸连接，或成一体；可以是机械连接，也可以是电连接；可以是直接相连，也可以通过中间媒介间接相连，可以是两个元件内部的连通或两个元件的相互作用关系。对于本领域的普通技术人员而言，可以根据具体情况理解上述术语在本发明中的具体含义。In the present invention, unless otherwise clearly specified and limited, terms such as "installation", "connection", "connection" and "fixation" should be understood in a broad sense, for example, it can be a fixed connection or a detachable connection , or integrated; it can be mechanically connected or electrically connected; it can be directly connected or indirectly connected through an intermediary, and it can be the internal communication of two components or the interaction relationship between two components. Those of ordinary skill in the art can understand the specific meanings of the above terms in the present invention according to specific situations.

在本发明中，除非另有明确的规定和限定，第一特征在第二特征“上”或“下”可以是第一和第二特征直接接触，或第一和第二特征通过中间媒介间接接触。而且，第一特征在第二特征“之上”、“上方”和“上面”可是第一特征在第二特征正上方或斜上方，或仅仅表示第一特征水平高度高于第二特征。第一特征在第二特征“之下”、“下方”和“下面”可以是第一特征在第二特征正下方或斜下方，或仅仅表示第一特征水平高度小于第二特征。In the present invention, unless otherwise clearly specified and limited, the first feature may be in direct contact with the first feature or the first and second feature may be in direct contact with the second feature through an intermediary. touch. Moreover, "above", "above" and "above" the first feature on the second feature may mean that the first feature is directly above or obliquely above the second feature, or simply means that the first feature is higher in level than the second feature. "Below", "beneath" and "beneath" the first feature may mean that the first feature is directly below or obliquely below the second feature, or simply means that the first feature is less horizontally than the second feature.

在本说明书的描述中，参考术语“一个实施例”、“一些实施例”、“示例”、“具体示例”、或“一些示例”等的描述意指结合该实施例或示例描述的具体特征、结构、材料或者特点包含于本发明的至少一个实施例或示例中。在本说明书中，对上述术语的示意性表述不必须针对的是相同的实施例或示例。而且，描述的具体特征、结构、材料或者特点可以在任一个或多个实施例或示例中以合适的方式结合。此外，在不相互矛盾的情况下，本领域的技术人员可以将本说明书中描述的不同实施例或示例以及不同实施例或示例的特征进行结合和组合。In the description of this specification, descriptions referring to the terms "one embodiment", "some embodiments", "example", "specific examples", or "some examples" mean that specific features described in connection with the embodiment or example , structure, material or characteristic is included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms are not necessarily directed to the same embodiment or example. Furthermore, the described specific features, structures, materials or characteristics may be combined in any suitable manner in any one or more embodiments or examples. In addition, those skilled in the art can combine and combine different embodiments or examples and features of different embodiments or examples described in this specification without conflicting with each other.

尽管上面已经示出和描述了本发明的实施例，可以理解的是，上述实施例是示例性的，不能理解为对本发明的限制，本领域的普通技术人员在本发明的范围内可以对上述实施例进行变化、修改、替换和变型。Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present invention, and those skilled in the art can make the above-mentioned The embodiments are subject to changes, modifications, substitutions and variations.

Claims

1. A polarized SAR image bridge detection method based on level set segmentation is characterized by comprising the following steps:

carrying out level set segmentation on the polarized SAR image according to the regional statistical characteristics to obtain land and a water area;

extracting characteristic points of the contour of the water area, and determining an interested area in the land according to the distance of the characteristic points to serve as a suspected bridge area;

removing false alarms in the suspected bridge area to realize bridge detection;

and carrying out constant false alarm detection on the suspected bridge area after the false alarms are eliminated, and distinguishing the strong scatterer bridge.

2. The method for detecting the polarized SAR image bridge based on the level set segmentation as claimed in claim 1, characterized in that the feature points of the contour of the water area are extracted by a digital curve splitting and merging algorithm.

3. The level set segmentation based polarized SAR image bridge detection method of claim 1, characterized in that the region of interest that is not connected is used as the false alarm.

4. The method for detecting a polarized SAR image bridge based on level set segmentation as claimed in claim 1, wherein the performing of the constant false alarm detection on the suspected bridge area after the false alarm is removed specifically comprises:

and taking the suspected bridge area after the false alarm is removed as a point target, representing the point target by an area average coherence matrix, and detecting the point target by a polarized whitening filter.

5. The utility model provides a polarize SAR image bridge detection device based on level set segmentation which characterized in that includes:

the segmentation module is used for carrying out level set segmentation on the polarized SAR image according to the regional statistical characteristics to obtain land and a water area;

the determining module is used for extracting characteristic points of the contour of the water area and determining an interested area in the land as a suspected bridge area according to the distance of the characteristic points;

the rejecting module is used for rejecting false alarms in the suspected bridge area to realize bridge detection;

and the detection module is used for carrying out constant false alarm detection on the suspected bridge area after the false alarms are eliminated so as to distinguish the strong scatterer bridge.

6. The apparatus of claim 5, wherein the determining module extracts feature points of the contour of the water area through a digital curve splitting and merging algorithm.

7. The level set segmentation-based polarized SAR image bridge detection device of claim 5, wherein the culling module takes the region of interest that is not connected as the false alarm.

8. The apparatus according to claim 5, wherein the detection module is specifically configured to: