摘要
由于断层、断裂等地貌特征,使利用InSAR技术获得的火山地形包裹相位图中易出现条纹密集、混叠等现象,为图像处理带来很大困难,很难准确恢复出火山真实地形的相位信息.为此,采用基于四向梯度自适应结合均值滤波的横向剪切四向加权最小二乘相位解缠图像处理方法,提出基于四向梯度的自适应滤波,利用一阶差分法判断图像边缘位置信息,有效去除噪声的同时提高边缘保持能力,解决边界模糊问题,结合均值滤波,更好地保留细节信息,提高图像清晰度.再采用横向剪切四向加权最小二乘算法进行相位解包裹,基于干涉图计算相位梯度值,利用离散余弦变换求解泊松方程获取真实相位.经仿真模拟实验验证,该算法运行时间短,解缠精度高,可准确地获取火山真实地形的相位信息.
Due to the geomorphological features such as faults and fractures,the volcanic terrain parcel phase map obtained by InSAR technology is prone to the phenomena of dense stripes and aliasing,which brings great difficulties to image processing and is difficult to accurately recover the phase information of volcanic real terrain.For this reason,proposes a transverse shearing four-way weighted least squares phase unwrapping image processing method based on four-way gradient adaptive combination with mean filtering.An adaptive filter based on four-way gradient is proposed.First-order difference method is used to judge the edge position information of the image,which can effectively remove noise while improving the edge retention ability,solve the problem of boundary blur, and combine with mean filter to better retain detail information and improve image clarity.Then the transverse shearing four-way weighted least square algorithm is used to unwrap the phase,the phase gradient value is calculated based on the interferogram,the Poisson equation is solved by discrete cosine transform to obtain the true phase.The simulation results show that the algorithm has short running time and high unwrapping precision,and can accurately obtain the phase information of volcanic real terrain.
引文
[1] Sparks R S J,高玲,魏海泉.预报火山喷发[J].世界地震译丛,2005(4):1-13
[2] LAWINFJ,FORSSéNPE,OVRéN.Efficientmulti-frequency phase unwrapping using kernel density estimation[C]//European Conference on Computer Vision.Springer International Publishing,2016:170-185
[3]许景波,袁怡宝,崔晓萌,等.表面测量中高斯滤波中线的有理逼近实现[J].吉林大学学报,2014,44(5):1347-1352
[4]靳国旺,张红敏,徐青,等.多波段In SAR差分滤波相位解缠方法[J].测绘学报,2012,41(3):434-440,448
[5]郭媛,杜松英.局部噪声斑块In SAR干涉图相位解缠算法[J].光子学报,2018,47(12):98-105
[6]李喆,李建增,扈琪.鲁棒自适应加权的引导滤波算法[J].电光与控制,2019,26(1):26-30
[7]王永红,陈维杰,钟诗民,等.相位解包裹技术及应用研究进展[J].测控技术,2018(12):1-7
[8]齐艳妮,刘亭,王光宇.InSAR图像的梯度自适应滤波方法[J].西部探矿工程,2016(12):129-132
[9]郭媛,吴全,毛琦,等.基于横向剪切的四向最小二乘相位解包裹算法[J].光电子·激光,2015(10):1953-1959
[10] Yan Liping,Zhang Haiyan,Zhang Rui,et al.A robust phase unwrapping algorithm based on reliability mask and weighted minimum least-squares method[J].Optics and Lasers Engineering,2019(112):39-45