Research On InSAR Phase Unwrapping Based On Ant Colony Genetic Fusion Algorithm

Since the 21 st century,China has been threatened by geological disasters such as earthquakes and landslides,and there is an urgent need for effective real-time monitoring and early warning.Interferometric Synthetic Aperture Radar(English name Interferometry Synthetic Aperture Radar,In SAR for short)is a space-to-ground observation technology that can monitor the ground 24 hours a day in real time,and can realize effective early warning of geological disasters.The core step of In SAR data processing is phase unwrapping.The pros and cons of the unwrapping algorithm will directly affect the digital elevation model(DEM)or the measured value of surface deformation.Since the 1980 s,many researchers at home and abroad have conducted in-depth research and analysis of unwrapping algorithms.After induction unwrapping methods can be divided into three categories,namely path-based tracking algorithm,minimum norm-based algorithm and network-based Flow minimum cost flow algorithm.This article will focus on In SAR phase unwrapping and mainly study path-based tracking algorithms.The main work is as follows:The first: the basic principle of this technique is explained,and the differences and specific steps of the three unwrapping methods are explained in detail.Among them are branch cutting,ant colony algorithm and many other methods.The second: aiming at the traditional Goldstein branch cutting algorithm,it is easy to cause the short tangent of the branch when searching for the residual points,which leads to the "island" effect of the unwrapping result.For this problem,an ant colony genetic algorithm was designed in addition to the traditional calculation method.Joint algorithm.After applying this method,it can effectively avoid the shortcomings of too many optimized parameters,slow calculation speed and low efficiency in the ant colony algorithm,reduce the time consumed by the ant colony algorithm in solving the local optimal solution,and improve the locality of the ant colony algorithm.Search ability,establish shorter branch tangent line than traditional branch method,prevent “islanding”effect during unwrapping process and improve phase unwrapping accuracy.Finally,the calculated value of Matlab software is compared with the measured value,which proves that this method is feasible and effective.The third: aiming at the problem that the operation efficiency of most In SAR phase unwrapping algorithms is relatively slow when processing large In SAR data,a parallel antcolony genetic joint algorithm based on Matlab is proposed.The parallel pool is started through Matlab,the number of computer CPU physical cores is used as the number of parallel cores,and the parfor loop is started to optimize the ant colony algorithm.This not only reduces the corresponding calculation time,but also compares it after actual measurement to demonstrate this calculation.The feasibility of the approach.