首页> 外文OA文献 >High efficiency image restoration using mirror images and adaptive Kalman filtering
【2h】

High efficiency image restoration using mirror images and adaptive Kalman filtering

机译:使用镜像和自适应卡尔曼滤波的高效图像恢复

摘要

The purpose of this paper is image restoration by means of fast parallel Kalman filtering . This algorithm was introduced byudBiemond et al. in 1983 [13] for image restoration . The problem is the appearance of ringing in the restored image. It resultsudfront the implicit assumption of rows periodicity, when FFT is applied on the image rows. Since left and right boundaries valuesudare différent, discontinuity points appear . In order to achieve boundary condition suitable for the ID horizontal FFT processing,udwe use in the algorithm a mirror image . We first give the description of this mirror image, and then we exploit in the algorithmudits horizontal properties of boundaries continuity and symmetry.udThen, an adaptive Kalman filtering is introduced to avoid non consistency of the Kalman gains factors on some frequencies, andudto reduce the remaining ringing.udFinally results are shown for some images and PSF .
机译:本文的目的是利用快速并行卡尔曼滤波进行图像复原。该算法由 udBiemond等人介绍。于1983年[13]用于图像恢复。问题是在还原的映像中出现了振铃。当在图像行上应用FFT时,它会导致周期性地隐含了行周期性的假设。由于左边界值和右边界值有差异,因此出现了不连续点。为了获得适合于ID水平FFT处理的边界条件, udwe在算法中使用镜像。首先,我们对该镜像进行描述,然后在算法中 udits边界连续性和对称性的水平属性。 ud然后,为了避免在某些频率上Kalman增益因子的不一致性,引入了自适应Kalman滤波,并且 ud减少剩余的响铃。 ud最后显示了某些图像和PSF的结果。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号