恶劣天气下的仿生偏振定向方法

针对偏振光导航在恶劣天气下精度显著下降的问题，提出了一种可以在恶劣天气下基于大气偏振模式的定向算法。与现有方法相比，首次将三维块匹配与canny边缘检测结合的思想应用于修复被不同天气破坏的偏振角度图像中。具体而言，将偏振角度图像的修复分为噪声粗处理和边缘提取处理去噪两部分。在噪声粗处理阶段采用三维块匹配算法,在边缘提取处理去噪部分，利用canny边缘检测算法对偏振角度图像进行二次去噪。实验结果表明，该方法不仅能够提高晴朗天空下的定向精度，而且能够显著提升在阴天、沙尘、雾霾等恶劣天气条件下的导航定向精度，即使在偏振角图像对称∞模式被破坏的情况下，航向角精度仍可由9.4470°提高到1.6859°。

Bionic Polarization Orientation Method Under Severe Weather

Geomagneticindoorpositioninghasbeenwidelyconcernedforitslowcost,widerangeof applicationandlowcomplexity. However,itspositioningaccuracyisnotstable,andthefingerprintcollectionworkloadisheavy. Inthispaper,afeaturebasedonmagneticfieldinformationof proximitypointcanbeobtainedfromfingerprintdata,andthefeaturerangeis0~1.Experiments showthatthecharacteristicsofadjacentpointsarepositivelycorrelatedwiththeaccuracyofpositioning. Thelargertheproximityfeature,thesmallerthepositioningerrorofthetestpoint,and thehighertheaccuracyofthepositioning.Thelargertheaverageofproximityfeature,thesmaller theindoorpositioningerror.Combiningthecharacteristicsofadjacentpointswiththemethodof quickconstructionoffingerprintdatabasecanrealizeprejudgmentofpositioningaccuracyoftheindoorarea, andprovideabasisfortheselectionoftheindoorpositioningmethod.Inviewoftheindoorareawithlargeproximityfeature, itisproposedtouseinterpolationalgorithmtoreducethe collectionoffingerprintdatabase.Inourexperiment,only29.6%oftheoriginalfingerprintpoints arecollected,whichhaslittleeffectonthepositioningerror.Theefficiencyofgeomagneticindoor positioningcanbegreatlyimprovedbyusingproximityfeatures.