| Navigation science based on natural characteristics plays a very important role in scientific research and has very important application value.At present,navigation science has a wide range of applications in military and civil fields.In recent years,based on the symmetrical characteristics of the polarization pattern of sky light,a new polarization navigation method using the symmetrical line-solar meridian as the benchmark to obtain direction information has become a hotspot in the field of polarization light navigation.This method of course information acquisition has the advantages of high stability and strong anti-jamming ability due to the use of all-sky polarization information,and has been used in military and civil navigation fields.It has far-reaching research value.At present,most imaging polarization navigation algorithms use the symmetry of polarization azimuth image to approximately extract the solar meridian,and retain the polarization information in the solar meridian and its vicinity.The advantages of this method lie in its convenience,less data processing and faster calculation,but most of the polarization information in polarization image is lost,and its robustness is poor.The calculation error is large.In this paper,an algorithm for estimating the relative rotation angle of the solar meridian is proposed,which is suitable for various imaging polarization navigation sensors.Unlike previous algorithms,this algorithm is based on the whole polarized image,and takes the smallest unit pixel of the image as the object of study,which greatly improves the utilization of polarization information in polarized image.Firstly,the polar coordinate system is established on the polarization azimuth image,so that every pixel in the polarization azimuth image can be represented by polar radius and polar angle in the polarization coordinate system.The purpose of establishing polar coordinate system is to establish a one-to-one correspondence between each pixel in polarized image and polar radius and polar angle,and lay a foundation for extracting single-pixel rings.Then the extraction process of single-pixel rings is carried out.The whole polarization azimuth image is traversed on the basis of polar coordinate system,and the single-pixel rings under a polar radius are extracted.Then the whole polarized image is traversed to match the pixels.The whole process includes the matching of the pixels in the single-pixel ring with a certain inner radius and the matching of all the outer pixel rings.In this way,most of the pixels in the whole image are calculated,which greatly improves the utilization of polarization information.In order to ensure the efficiency of the algorithm in the traversal process,this paper optimizes the tabulation method in the search process of the algorithm,which greatly improves the efficiency of the algorithm.Finally,simulation experiments and outdoor static and dynamic experiments are carried out.The simulation experiment uses the polarization azimuth image constructed to discuss the influence of different gray resolution on the calculation results of the algorithm.The experiment shows that the improvement of gray resolution is helpful to improve the accuracy of the algorithm.The outdoor static experiment uses the full polarization imaging detector of the research group to collect and process the polarization images of the sky at different times in the outdoor open environment.The outdoor feasibility of the algorithm is demonstrated by experiments,which show that the error of the algorithm is within 0.63° in outdoor static experiments.In outdoor dynamic experiments,the polarization information acquisition device is mounted on the turntable to rotate and collect sky polarization images.The polarization azimuth before coordinate conversion is used as the real value of the experiment.The error of the algorithm in the experimental results is between-0.7°~+0.5°,which proves that the algorithm is effective.It proves the feasibility and robustness of the algorithm. |
PDF Full Text Request