Print

发布时间: 2017-07-16
摘要点击次数:
全文下载次数:
DOI: 10.11834/jig.160659
2017 | Volume 22 | Number 7




    遥感图像处理    




  <<上一篇 




  下一篇>> 





残差点退化的统计费用网络流机载相位解缠算法
expand article info 刘怡君1, 韩春明1,2, 岳昔娟1
1. 中国科学院遥感与数字地球研究所, 北京 100093;
2. 三亚中科遥感所, 三亚 572029

摘要

目的 相位解缠是InSAR干涉数据处理的关键步骤,而解缠不连续(即相位跳变)问题却普遍存在,尤其在机载InSAR系统中,由于数据的高分辨率,使得低矮地物如树木带在数据中表现为相位不一致,因而相位跳变问题更加显著。星载InSAR相位解缠广泛使用统计费用网络流(SNAPHU)算法[1],借鉴其经验将SNAPHU算法引入高分辨机载InSAR相位解缠。而残差点退化方法能有效补偿局部相位不一致区域。因此本文提出一种结合残差点退化方法与SNAPHU算法的高分辨率机载InSAR相位解缠算法。 方法 将原始InSAR数据滤波且去除平地相位,再对其进行残差点退化处理。残差点退化包含残差点定位,及残差点补偿两部分。根据残差点及其邻域像元的性质,对残差点进行补偿使其退化为非残差点,不断迭代这一过程,以减少图像中的残差点,优化局部数据。根据机载InSAR系统定标参数,修正SNAPHU算法中的参数及几何模型,使用修正后算法进行相位解缠。 结果 利用2011年四川江油地区的单轨双天线X波段机载InSAR数据进行了试验,试验结果表明,在相位不一致,相干性低的连续树木带区域,该算法显著缩小了解缠相位不连续区域,修正了大面积的相位跳变。 结论 验证了残差点退化方法结合统计费用网络流算法可有效解决解缠相位大面积跳变问题,且对噪声具有鲁棒性。

关键词

机载干涉合成孔径雷达; 相位解缠; 残差点; 统计费用网络流(SNAPHU)

Residue degradation statistical-cost network-flow phase unwrapping algorithm for airborne SAR interferometry
expand article info Liu Yijun1, Han Chunming1,2, Yue Xijuan1
1. Institute of Remote Sensing and Digital Earth, Chinese Academy of Sciences, Beijing 100093, China;
2. Sanya Institute of Remote Sensing, Sanya 572029, China
Supported by: Key Research and Development Plan of Hainan Province (SY16ZYD2132)

Abstract

Objective Phase unwrapping is one of the key steps in InSAR data processing, and its precision directly affects the accuracies of DEM generation or surface deformation monitoring. The phase extracted from a complex SAR interferogram is wrapped because it represents a measure; modulo 2π and multiples of 2π must be added or subtracted. Ideally, when the sampling rate of the image satisfies the Nyquist sampling theorem, phase unwrapping can be achieved by a simple 2D integration. However, phase inconsistencies caused by overlay and shadow appear in the actual data. When the integration path passes through the abovementioned area, the phase error will spread, leading to unwrapping errors. Numerous phase unwrapping algorithms are proposed to solve the above problems. The statistical-cost network-flow algorithm [1](SNAPHU) is a kind of network flow method that is widely used in InSAR data processing. Based on the SNAPHU algorithm, this study aims to find a phase unwrapping algorithm for airborne SAR interferometry. While applying the SNAPHU to high-resolution airborne InSAR data, if the interferogram has phase inconsistencies caused by lines of trees, then the unwrapped result will show large areas of phase jump along the inconsistencies correspondingly. SNAPHU, which has limitations in managing local phase inconsistencies, is a global optimization algorithm. Local phase inconsistencies can be optimized by a local algorithm. Phase inconsistencies can be described by residues. Local data can be optimized by residue compensation. This paper proposes an airborne unwrapping algorithm combining the residue degradation with the SNAPHU, which has advantages of the local and the global optimum. Method Residues are degraded to non-residues according to the values of residues and neighboring pixels, and the local data are optimized. The modified SNAPHU algorithm is applied to the degraded image. The proposed algorithm can be divided into three steps.(1) Residue degradation. First, the filtered interferogram is flattened. Residue degradation is applied to the flattened interferogram. This process contains residue detection and compensation. Phase inconsistencies lead to the presence of residues. Phase inconsistencies mean the absolute phase difference between two neighboring pixels exceeding π. Therefore, the pair of pixels in the residue, whose phase is inconsistent, is first located to degenerate the residue, and then a compensation constant C is set to compensate the residue. The residue can be degraded by multiple iterations because the compensation changes the value of phase and the phase difference between the neighboring pixels. A threshold N for the remaining residues is set to balance the accuracy and efficiency. When the number of residues is less than N, the compensation stops to control the degree of degradation.(2) Phase unwrapping of airborne SNAPHU algorithm. In the airborne system, the incidence angle ${\theta _i} $ calculated by the original geometric model of SNAPHU is close to zero, which is obviously incorrect, because the height of the platform is far less than the radius of the Earth. Simultaneously, the earth curvature can be ignored to simplify the model with a narrow mapping bandwidth in the airborne system, thus the parameters and the geometric model in the SNAPHU algorithm are modified according to the calibration parameters of the airborne InSAR system. The modified SNAPHU algorithm is employed for phase unwrapping.(3) Median filter. A median filter is applied to the unwrapped result, which effectively reduces phase noise, and the edge information is well preserved. Therefore, a 5×5 median filter is applied to the unwrapped image to obtain the final result. Result The efficiency and accuracy of the proposed method is tested and validated by using the single-pass dual antenna airborne InSAR data covering Jiangyou, Sichuan areas in 2011. Unwrapping results have a significant error region using SNAPHU directly. A large area of the phase jumping exists. However, error unwrapped regions are significantly shrunk after residue degradation, and phase jumping regions are effectively corrected. After a 5 × 5 median filter, the partial phase noise is removed, and the unwrapping result is smooth. The performances of the improved methods are evaluated by the number of discontinuous points in azimuth and range directions. When the number of discontinuous point is minimal, the anti-phase distortion performance is improved, and the unwrapping quality is high. The residue degradation effectively reduces the number of discontinuous points in the interferogram. The number of discontinuous points in either direction of improved airborne SNAPHU algorithm is lower than that of the simple airborne SNAPHU algorithm. The anti-phase distortion performance and the quality of the unwrapping results are obviously improved. Conclusion The residue degradation process effectively solves the problem of large unwrapped phase jump and becomes further robust to noise. This algorithm combines the advantages of local and global optimization, which has a good overall performance, and effectively solves the problem of large unwrapped phase jump. Moreover, median filter can effectively reduce the noise and is robust to noise.

Key words

airborne SAR interferometry; phase unwrapping; residue; statistical-cost network-flow(SNAPHU)

0 引言

合成孔径雷达干涉测量(InSAR)是一项利用复图像相位差作为信息源提取地表3维信息的技术[2]。目前InSAR技术主要运用于生成大面积高分辨率的数字高程模型(DEM)和检测地表形变。相位解缠是InSAR技术的关键步骤,其精度直接影响DEM和地表形变精度[3]

根据两幅SAR复图像获得的干涉相位差值$\psi $是周期折叠后的相位主值,值域为(-π,π]。其与真实相位差值$\emptyset $,相差$2k\pi $

$ \emptyset = \psi + 2k\pi \;\;k = 0, \pm 1, \pm 2, \cdots $ (1)

只有获得目标在两次成像中的真实相位差,才能得到目标的正确高度信息,因此必须恢复$\psi $被模糊掉的相位周期。这种由相位主值得到真实相位值的过程称为相位解缠[4-5]。为后文叙述一致,$\emptyset $为解缠相位,$\psi $为缠绕相位。

理想情况下,当图像采样率满足Nyquist采样定理时,通过简单2维积分即可实现相位解缠。但实际数据中存在由叠掩、阴影等导致的相位不一致($\left| {\Delta \emptyset } \right| > \pi $),积分路径在通过上述区域时,将产生误差的传递,出现解缠错误[6]。为解决上述问题,国内外学者提出了几十种相位解缠算法,主要分为三大类:1) 路径跟踪法[7-8],以枝切法为代表,基于残差点确定积分路径;2) 最小范数法[9-11],以最小二乘法为代表,基于最小范数准则,通过缠绕相位差估计解缠相位;3) 网络流法[12-14],以统计费用网络流(SNAPHU)算法为代表,基于网络流理论。第1类路径跟踪法,为局部算法,将可能的误差传递限制在噪声区内,通过选择合适的积分路径,隔绝噪声区,阻止相位误差的全程传递,但在高密度残差点区域,容易形成解缠孤岛;第2类最小范数法,为全局算法,稳定性高,解缠结果有很好的连续性和平滑性,但易造成全程的误差传递。第3类网络流法,为全局算法,相较而言,其兼具了解缠连续性与限制误差传递的优点。

SNAPHU算法为网络流法,被广泛运用于InSAR数据处理中,本文以SNAPHU算法为机载相位解缠算法的研究基础。SNAPHU本为星载相位解缠算法,其参数与几何模型与机载InSAR系统并不完全匹配,因此根据机载系统的特性,修正其参数与几何模型。此外,机载InSAR系统获取的数据分辨率更高,相较星载数据更易出现相位不一致,尤其在山区,树木繁茂区域相干性低,连续的树木带将产生严重的相位不一致。相位不一致为局部问题,当干涉数据存在严重相位不一致时,SNAPHU作为一种全局算法,并不能很好地解决该问题,解缠结果往往存在沿相位不一致为分界的大面积相位跳变。

为解决上述问题,结合路径跟踪法和网络流法的优点,提出一种残差点退化的修正几何模型的统计费用网络流机载相位解缠算法。该算法根据残差点及其邻域像元的性质对残差点进行处理,使其退化为非残差点,使局部数据得以优化;再采用机载统计费用网络流法对退化后的图像进行解缠。该算法可分为3步:1) 残差点退化;2) 机载统计费用网络流相位解缠;3) 中值滤波。该算法结合了局部求优与全局求优算法的特点,从而在整体上具有良好的解缠性能,有效地解决了解缠相位大面积跳变的问题;此外,中值滤波能有效降低噪声的影响,对噪声具有鲁棒性。

1 残差点退化

1.1 残差点检测

残差点的定义如图 1所示。残差点表示为一个由4相邻像元构成的2×2矩阵,$i $为行,$j $为列。

图 1 残差点结构图
Fig. 1 The structure chart of residue

残差点检测表述为

$ \begin{array}{l} {\Delta _1} = W\left[{\psi \left( {i, j + 1} \right)-\psi \left( {i, j} \right)} \right]\\ {\Delta _2} = W\left[{\psi \left( {i + 1, j + 1} \right)-\psi \left( {i, j + 1} \right)} \right]\\ {\Delta _3} = W\left[{\psi \left( {i + 1, j} \right)-\psi \left( {i + 1, j + 1} \right)} \right]\\ {\Delta _4} = W\left[{\psi \left( {i, j} \right)-\psi \left( {i + 1, j} \right)} \right] \end{array} $ (2)

式中,$\emptyset $为像元点的缠绕相位,${\Delta _i} $为2点间的缠绕相位差[15]$W $为缠绕函数,将相位缠绕在主值区间(-π,π),即

$ W\left[{\psi \left( {m + 1} \right)-\psi \left( m \right)} \right] = \left\{ \begin{array}{l} \psi \left( {m + 1} \right) -\psi \left( m \right)\;\;\;\;\;\left| {\psi \left( {m + 1} \right) -\psi \left( m \right)} \right| < \pi \\ \psi \left( {m + 1} \right) -\psi \left( m \right) - 2\pi \;\;\psi \left( {m + 1} \right) - \psi \left( m \right) > \pi \\ \psi \left( {m + 1} \right) - \psi \left( m \right) + 2\pi \;\;\;\psi \left( {m + 1} \right) - \psi \left( m \right) < - \pi \end{array} \right. $ (3)

其中$q $表示的4点的相位差值和,即

$ q = \sum\limits_i^4 {{\Delta _i}} $ (4)

$q = 0 $时,该点为非残差点。当$q \ne 0 $时,该点为残差点,当$q \le-2\pi $时该点称之为正残差点;当$q \ge-2\pi $时该点称之为负残差点。

1.2 残差点补偿

相位不一致导致了残差点的存在。相位不一致,即为相邻像元间的绝对相位差超过π值,即$\left| {\psi \left( {m + 1} \right)-\psi \left( m \right)} \right| > \pi $。因此本文退化残差点的思路:首先定位残差点中相位不一致的像元点对,而后设置补偿步长$C $,将其补偿到定位点上。由于对残差节点的补偿修改了像元相位值,影响了周围像元间的相位差,通过多次迭代补偿,可使孤立残差点退化,相邻极性相反的残差点彼此抵消,从而达到使所有残差点退化的目的。

图 2所示,$\Delta {\psi _i}$表示2点之间的相位差值。计算残差节点内两两像元之间的相位差值,即

图 2 残差点相位差
Fig. 2 The phase difference of residue

$ \begin{array}{l} \Delta {\psi _1} = \psi \left( {i, j + 1} \right)-\psi \left( {i, j} \right)\\ \Delta {\psi _2} = \psi \left( {i + 1, j + 1} \right)-\psi \left( {i, j + 1} \right)\\ \Delta {\psi _3} = \psi \left( {i + 1, j} \right)-\psi \left( {i + 1, j + 1} \right)\\ \Delta {\psi _4} = \psi \left( {i, j} \right) - \psi \left( {i + 1, j} \right) \end{array} $ (5)

$\Delta {\psi _1} $为例,阐述像元点补偿机制,${{\bar \psi }_{i, j}} $为像元点$\left( {i, j} \right) $8邻域相位均值,即

$ \begin{array}{l} {{\bar \psi }_{i, j}} = \left( {\psi \left( {i-1, j-1} \right) + \psi \left( {i-1, j} \right)} \right.\\ \psi \left( {i - 1, j + 1} \right) + \psi \left( {i, j - 1} \right) + \psi \left( {i, j + 1} \right) + \\ \psi \left( {i + 1, j - 1} \right) + \psi \left( {i + 1, j} \right) + \\ \left. {\psi \left( {i + 1, j + 1} \right)} \right)/8 \end{array} $ (6)

如果$\Delta {\psi _1} > \pi $,则

$ \left\{ \begin{array}{l} \psi \left( {i, j} \right) = \psi \left( {i, j} \right) + C\;\;\;\;\;\;\;\;\;\;{{\bar \psi }_{i, j}} > \pi \\ \psi \left( {i, j + 1} \right) = \psi \left( {i, j + 1} \right)-C\;\;\;{{\bar \psi }_{i, j}} \le \pi \end{array} \right. $ (7)

如果$\Delta {\psi _1} <-\pi $,则

$ \left\{ \begin{array}{l} \psi \left( {i, j + 1} \right) = \psi \left( {i, j + 1} \right) + C\;\;\;\;\;\;\;\;\;\;{{\bar \psi }_{i, j}} > \pi \\ \psi \left( {i, j} \right) = \psi \left( {i, j} \right)-C\;\;\;{{\bar \psi }_{i, j}} \le \pi \end{array} \right. $ (8)

为平衡精度与效率,设置残留残差点阈值$N $,控制残差点退化的程度,当残留残差点数小于$N $,则停止补偿。

2 机载统计费用网络流方法

2.1 SNAPHU算法

SNAPHU属于网络流法,其思路归结为求解最小费用流的网络规划问题:1) 构建网络模型;2) 建立费用函数;3) 获取最小费用流;4) 通过最小费用流求得解缠相位差;5) 沿路径积分,求取解缠相位。

2.1.1 网络模型

1998年,Costantini[1]首次提出用于解决相位解缠的网络模型,如图 3所示。其中四条双向弧围成的小正方形表示像元,正方形内的数字为相位值$\psi $。四相邻的2×2像元构成节点,节点由圆圈表示。连通节点的为弧,弧对应像元间的相位差,弧上流的方向和大小分别对应相位差的方向和大小。

图 3 网络模型
Fig. 3 Network model

2.1.2 费用函数

SNAPHU费用函数表示,在已知前提下,解缠相位差$\Delta \mathit{\boldsymbol{ \boldsymbol{\varPhi} }} $与真实相位差$\Delta \mathit{\boldsymbol{ \boldsymbol{\hat \varPhi} }} $的接近程度。该费用函数为条件概率密度函数$f\left( {\Delta \mathit{\boldsymbol{ \boldsymbol{\varPhi} |}}I, \rho } \right) $的负对数[1],即

$ \mathit{\boldsymbol{G}}\left( \cdot \right) =-\log \left( {f\left( {\Delta \mathit{\boldsymbol{ \boldsymbol{\varPhi} }}|I, \rho } \right)} \right) $ (9)

$\mathit{\boldsymbol{G}}\left( \cdot \right) $表示费用函数,${\Delta \mathit{\boldsymbol{ \boldsymbol{\varPhi} }}} $为解缠相位差,$I $为干涉图像的平均亮度,$\rho $为干涉图像的复相干系数。$f\left( {\Delta \mathit{\boldsymbol{ \boldsymbol{\varPhi} |}}I, \rho } \right) $表示,在已知图像亮度和相干性的前提下,解缠相位差${\Delta \mathit{\boldsymbol{ \boldsymbol{\varPhi} }}} $的条件概率。文中大写字母表集合,小写字母表元素,${\Delta \mathit{\boldsymbol{ \boldsymbol{\varPhi} }}} $为解缠相位差$\Delta \emptyset $的集合。

2.1.3 最小费用流

构建好费用函数后,须求得函数最小值,以及费用最小时,网络流的布局。SNAPHU在动态消圈法(DCC)的基础上,提出了一种非线性的网络流算法——Pivot-and-Grow算法[1],并由Pivot-and-Grow算法解得各节点的最小费用流流量$n $

由于流量$n $表示缠绕相位差$\Delta \psi $与解缠相位差$\Delta \emptyset $之差($n = \Delta \psi-\Delta \emptyset $),因此$\Delta \psi $加上$n $即为解缠相位差$\Delta \emptyset $。对缠绕相位$\psi $,沿路径积分解缠相位差$ \Delta \emptyset $,即可得到解缠相位。

2.2 机载SNAPHU几何模型

2.2.1 SNAPHU原几何模型

SNAPHU算法原适用于星载InSAR系统,在求解费用函数过程中,需要代入视角$\theta $以及入射角${\theta _i} $的值。为计算$\theta $${\theta _i} $值,原SNAPHU几何模型如图 4所示。

图 4 SNAPHU几何模型
Fig. 4 The geometric model of SNAPHU

图 4${\rm{O}} $为地心;${\rm{P}} $为目标点;${A_1} $${A_2} $为主副天线;$\theta $为天线${A_1} $对目标点${\rm{P}} $成像时的视角;$R $为主天线到目标点的距离;$H $为天线${A_1} $到地心${\rm{O}} $的距离;${\theta _i} $为入射角。根据几何模型,${\theta _i} $$\theta $可表示为

$ {\theta _i} = \pi-{\cos ^{-1}}\left( {\left( {{R^2} + R_a^2-{H^2}} \right)/2R\;{R_a}} \right) $ (10)

$ \theta = {\sin ^{-1}}\left( {{R_a}\sin {\theta _i}/H} \right) $ (11)

2.2.2 机载几何模型

在机载系统中,由于平台高度远小于地球半径,按星载几何模型计算出的入射角${\theta _i} $接近于0,显然不正确。同时由于机载测绘带宽较窄,可简化模型,不计地球曲率,机载几何模型如图 5所示。

图 5 机载几何模型
Fig. 5 The geometric model of airborne system

图 5${\rm{P}} $为目标点;${A_1} $${A_2} $为主副天线;$\theta $为天线${A_1} $对目标点${\rm{P}} $成像时的视角;$R $为主天线到目标点的距离;$H $为天线${A_1} $的高程;${\theta _i} $为入射角。根据机载几何模型,${\theta _i} $可等同于视角$\theta $,在已知高程和斜距的前提下,入射角${\theta _i} $与视角$\theta $可表示为

$ {\theta _i} = \theta = {\cos ^{-1}}\left( {H/R} \right) $ (12)

3 实验与分析

为验证算法对机载InSAR系统的有效性,采用2011年四川江油地区的双天线X波段机载InSAR数据:滤波后的干涉复图像(分辨率为0.5 m,1 500×1 500像素)、同区域相干图。图 6(a)为树木茂密,细节丰富的干涉复图像(已滤波),图 6(b)为相干图,图 6(c)图 6(a)对应的残差点图(白色区域为残差点分布区域,残差点表示相位不一致,即相位不正确区域),图 6(d)图 6(a)干涉复图像对应的缠绕相位图,图 6(e)为去除平地相位[16]后的缠绕相位图。

图 6 江油地区机载InSAR数据
Fig. 6 The airborne InSAR data of Jiangyou((a) filtered interferogram; (b) coherence image; (c) residues image; (d) wrapped phase image; (e) flattened wrapped phase image)

图 6(a)(c)可看出,树木繁茂区域相干性低,残差点密集,残差点图与树木茂盛区域相吻合,可见树木是影响高分辨率机载InSAR数据解缠质量的重要因素。图 6(d)条纹密集,高程信息被平地相位所掩盖,无法直观反映地表高程,较之图 6(e)去除平地相位后,地形信息凸显,能直观反映高程变化。

采用机载SNAPHU算法和结合残差点退化的机载SNAPHU算法方法进行解缠,解缠结果如图 7(a)(c)。Christophe Magnard等验证了解缠后,对其结果使用5×5窗口中值滤波,能有效降低相位噪声,且边缘信息[17]保存完好。本文受其启发,对解缠结果进行了5×5中值滤波。

图 7 解缠结果图
Fig. 7 Unwrapped results((a) airborne SNAPHU algorithm; (b) airborne SNAPHU algorithm with residues degradation; (c) unwrapping results after the 5×5 median filtering)

图 7(a)所示,图像左右侧灰度差异明显,左侧较亮区域突变为较暗区域,左右两侧的相位值相差2π,可见直接采用SNAPHU算法,解缠结果有明显的误差传递,存在大面积的相位跳变。而图 7(b)所示,图像左右两侧灰度一致,不存在灰度突变,即相位变化连续,可见残差退化方法有效修正了相位跳变区域。由图 7(c)可见,5×5中值滤波后,部分相位噪声被去除,解缠结果更为平滑。

下面从方位向和距离向上的不连续点数目来评价改进方法的性能[18]

不连续点数越少,说明抗相位畸变的性能越好,解缠质量越高。表 1给出机载SNAPHU算法和残差点退化后的机载SNAPHU算法的不连续点数目。图 8图 9分别给出原始缠绕相位、机载SNAPHU算法和改进后的机载SNAPHU算法的距离向和方位向的不连续点图。图 8(a)图 9(a)分别展示的是原始InSAR数据(未去除平地相位)中的距离向和方位向不连续点的分布,其分布与图 6(e)缠绕相位图中相位跳变的位置相吻合,可见为获取准确的地表高程变化,平地相位的去除是必要的。图 8(b)图 9(b)分别展示机载SNAPHU算法(修正参数与模型的SNAPHU)解缠去平干涉数据,其解缠结果中距离向与方位向不连续点的分布。将图 8(b)图 9(b)对照图 6(a)可看出,不连续点的分布与连续树木带的位置相吻合;同时对照图 6(b)可发现,树木区域的相干性低,是影响解缠质量的重要因素,而就图 8(b)图 9(b)的结果而言,SNAPHU算法并不能很好地解决由长距离连续树木带所导致的相位不连续问题。观察图 8(c)图 9(c),可直观看出距离向与方位向的不连续点数明显减少,对照图 8(b)图 9(b),树木带处的密集不连续点几乎都已消失,可见残差退化方法很好地解决了由长距离连续树木带所导致的相位不连续问题。

表 1 不连续点数目表
Table 1 The number of discontinuous point

下载CSV
解缠算法 距离向不连续点数(距离向
不连续点数目占所有像元的百分比)
方位向不连续点数(方位向
不连续点数占所有像元的百分率)
机载SNAPHU算法 3 927(0.175%) 3 288(0.146%)
残差点退化后的机载SNAPHU算法 724(0.032%) 570(0.025%)
图 8 距离向不连续点图
Fig. 8 The discontinuity maps of range gradients((a) original wrapped phase; (b) airborne SNAPHU algorithm; (c) improved airborne SNAPHU algorithm)
图 9 方位向不连续点图
Fig. 9 The discontinuity maps of azimuth gradients((a) original wrapped phase; (b) airborne SNAPHU algorithm; (c) improved airborne SNAPHU algorithm)

从数学统计而言,原始缠绕相位的距离向不连续点数为42 546,方位向不连续点数为32 723。使用机载SNAPHU算法,解缠结果中的距离向不连续点数为3 927,方位向不连续点数为3 288,相较原始缠绕相位,其数量已大幅减少。而残差点退化后的机载SNAPHU算法,其不连续点数目较之机载SNAPHU算法进一步减少。表明本文算法抗相位畸变性能和解缠质量都明显提高。

4 结论

通过结合残差点退化以及修正SNAPHU几何模型来得到一种适用于机载InSAR系统的相位解缠算法。修正SNAPHU几何模型,使其数学逻辑更为准确,符合机载系统的几何关系。残差点退化有效减少了干涉图像中的不连续点数,解决了解缠结果存在的大面积相位跳变问题。利用真实数据实验验证了该算法的可行性和有效性。利用不连续点数证明了本文所提出的结合残差点退化及修正SNAPHU几何模型的相位解缠流程,对解缠结果有所改进。

但本文所做的几何模型修正,考虑的是理想情况,即目标点处于水平面上,没有考虑目标点物的高程,数学模型不够完善,这也是下一步需要继续研究的内容。同时残差点退化方法,虽然能有效修正解缠相位的大面积跳变,但解缠结果还不够平滑,修正的区域边界仍较为粗糙,不能准确地解决小面积的解缠相位跳变。这些仍是后续需要改进的地方。

参考文献

  • [1] Chen C W. Statistical-cost network-flow approaches to two-dimensional phase unwrapping for radar interferometry[D]. California: Stanford University, 2001.
  • [2] Zhu T, Wang Z Y, Xie M, et al.Research of four directions weighted least-square algorithm for phase unwrapping[J]. Journal of Sichuan University: Natural Science Edition, 2009, 46(2): 372–376. [朱挺, 王正勇, 谢明, 等. 四向加权最小二乘法相位解缠研究[J]. 四川大学学报:自然科学版, 2009, 46(2): 372–376. ] [DOI:10.3969/j.issn.0490-6756.2009.02.022]
  • [3] Chen Q, Yang Y H, Liu G X, et al.InSAR phase unwrapping using least squares method with integer ambiguity resolution and edge detection[J]. Acta Geodaetica et Cartographica Sinica, 2012, 41(3): 441–448. [陈强, 杨莹辉, 刘国祥, 等. 基于边界探测的InSAR最小二乘整周相位解缠方法[J]. 测绘学报, 2012, 41(3): 441–448. ]
  • [4] Ghiglia D C, Pritt M D. Two-Dimensional Phase Unwrapping: Theory, Algorithms, and Software[M]. New York: John Wiley & Sons, 1998: 122-135.
  • [5] Xu J Y, An D X, Huang X T, et al.Phase unwrapping for large-scale P-band UWB SAR interferometry[J]. IEEE Geoscience and Remote Sensing Letters, 2015, 10(12): 2120–2124. [DOI:10.1109/LGRS.2015.2451991]
  • [6] Zhong H P, Tang J S, Zhang S, et al.A quality-guided and local minimum discontinuity based phase unwrapping algorithm for InSAR/InSAS interferograms[J]. IEEE Geoscience and Remote Sensing Letters, 2014, 11(1): 215–219. [DOI:10.1109/LGRS.2013.2252880]
  • [7] Goldstein R M, Zebker H A, Werner C L.Satellite radar interferometry: two-dimensional phase unwrapping[J]. Radio Science, 1988, 23(4): 713–720. [DOI:10.1029/RS023i004p00713]
  • [8] Zhong H P, Tang J S, Zhang S.A combined phase unwrapping algorithm based on quality-guided and minimum discontinuity for InSAR[J]. Journal of Electronics & Information Technology, 2011, 33(2): 369–374. [钟何平, 唐劲松, 张森. 一种基于质量引导和最小不连续合成的InSAR相位解缠算法[J]. 电子与信息学报, 2011, 33(2): 369–374. ] [DOI:10.3724/SP.J.1146.2010.00440]
  • [9] Wang H Y, Liu Z Q, Liao W, et al.Comparison of four phases unwrapping algorithm based on method of minimum norm[J]. Chinese Journal of Lasers, 2014, 41(2): #0209016. [王华英, 刘佐强, 廖薇, 等. 基于最小范数的四种相位解包裹算法比较[J]. 中国激光, 2014, 41(2): #0209016. ] [DOI:10.3788/CJL201441.0209016]
  • [10] Pritt M D.Phase unwrapping by means of multigrid techniques for interferometric SAR[J]. IEEE Transactions on Geoscience and Remote Sensing, 1996, 34(3): 728–738. [DOI:10.1109/36.499752]
  • [11] Guo R H, Li J X, Zhu R H.Seed point unwrapping algorithm based on DCT algorithm[J]. Acta Optica Sinica, 2012, 32(2): #0212006. [郭仁慧, 李建欣, 朱日宏. 基于DCT算法的种子点相位解包算法[J]. 光学学报, 2012, 32(2): #0212006. ] [DOI:10.3788/AOS201232.0212006]
  • [12] Costantini M.A novel phase unwrapping method based on network programming[J]. IEEE Transactions on Geoscience and Remote Sensing, 1998, 36(3): 813–821. [DOI:10.1109/36.673674]
  • [13] Chen C W, Zebker H A.Network approaches to two-dimensional phase unwrapping: intractability and two new algorithms[J]. Journal of the Optical Society of America A, 2000, 17(3): 401–414. [DOI:10.1364/JOSAA.17.000401]
  • [14] Wang L J, Jiang H J, Da H Q, et al.Parallel computing based on statistical-cost network-flow algorithm for phase unwrapping[J]. Geomatics and Information Science of Wuhan University, 2010, 35(9): 1061–1064. [王丽娟, 蒋厚军, 达汉桥, 等. 统计费用网络流相位解缠并行处理[J]. 武汉大学学报:信息科学版, 2010, 35(9): 1061–1064. ]
  • [15] Wang Z Y, Zhu T, He X H, et al.A singularity point degradation four directions least-square algorithm for InSAR phase unwrapping[J]. Journal of Sichuan University: Engineering Science Edition, 2010, 42(1): 185–190. [王正勇, 朱挺, 何小海, 等. 一种残差点退化的四向最小二乘InSAR相位解缠算法[J]. 四川大学学报:工程科学版, 2010, 42(1): 185–190. ]
  • [16] Zhang Y, Wan D.InSAR imaging technology[J]. Electronic Science & Technology, 2012, 25(3): 107–109. [张跃, 万昳. 干涉合成孔径雷达成像技术[J]. 电子科技, 2012, 25(3): 107–109. ] [DOI:10.3969/j.issn.1007-7820.2012.03.032]
  • [17] Magnard C, Frioud M, Small D, et al.Processing of MEMPHIS Ka-band multibaseline interferometric SAR data: from raw data to digital surface models[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2014, 7(7): 2927–2941. [DOI:10.1109/JSTARS.2014.2315896]
  • [18] Liu G L, Hao H D, Yan M, et al.Phase unwrapping algorithm by using Kalman filter based on topographic factors[J]. Acta Geodaetica et Cartographica Sinica, 2011, 40(3): 283–288. [刘国林, 郝华东, 闫满, 等. 顾及地形因素的卡尔曼滤波相位解缠算法[J]. 测绘学报, 2011, 40(3): 283–288. ]