期刊文献+

基于预滤波迭代梯度法的运动估计

Motion estimation based on pre-filter iterative gradient method
下载PDF
导出
摘要 提出一种对小平移和大平移图像序列进行运动估计的新算法 预滤波迭代梯度法,预滤波迭代梯度法就是先对图像进行低通滤波,然后用迭代梯度法进行运动估计。结果表明,该方法比梯度法、预滤波梯度法和迭代梯度法具有更高的估计精度,而且可以很好地抑制噪声的影响。 The pre-filter iterative gradient method for estimating motion in image sequences with small and large displacements is presented, which pre-filter the images with a low pass filter and motion estimation is performed using the iterative gradient method. The experimental results show that the new algorithm is more accurate than gradient method, pre-filter gradient method and iterative gradient method and can greatly retrain the effect of noise.
出处 《光学技术》 CAS CSCD 2004年第5期538-540,共3页 Optical Technique
基金 国家自然科学基金委员会与中国节能投资公司联合研究资金(项目号:60107003) 陕西省自然科学研究项目资助
关键词 预滤波迭代梯度法 运动估计 梯度法 pre-filter iterative gradient method motion estimation gradient method
  • 相关文献

参考文献6

  • 1[1]Altunbasak Y, Merserau R M, Patti A J. A fast parametric motion estimation algorithm with illumination and lens distortion correction[J]. IEEE Transaction on Image Processing, 2003, 12(4): 395-408.
  • 2[2]Alam M S, Bognar G J, Hardie R C, et al. High resolution infrared image reconstruction using multiple, randomly shifted, low resolution, aliased frames[J]. Proc. of the SPIE, 1997,3063:102-112.
  • 3[3]Kaltenbacher Eric, Hardie R C. High resolution infrared image reconstruction using multiple, low resolution, aliased frames[J]. Proc. of the SPIE, 1996, 2751: 142-152.
  • 4[4]Irani M, Peleg S. Improving resolution by image registration[J]. CVGIP:Graphical Models and Image Processing, 1991,53(3):231-239.
  • 5[5]Schaum A, McHugh M. Analytic methods of image registration: Displacement estimation and resampling[R]. Naval Research Report9298, 1992.28.
  • 6[6]Bergen J R, Burt P J, Hingorani R, et al. A three-frame algorithm for estimating two-component image motion[J]. IEEE Transaction on Pattern Analysis and Machine Intelligence,1992,14(9):886-896.

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部