摘要
该文在考虑二维图象失真满足二维联立自回归（ＳＡＲ）模型基础上，提出自适应空域递归算法Ⅱ（ＳＡⅡ）。该算法与ＳＡⅠ算法不同，ＳＡⅠ为逐行迭代，每次修正所有的参数，而ＳＡⅡ逐点迭代，但每次只修正１组参数。ＳＡⅡ的收敛速度并没有因此而损失，但运算量大大减少，且稳定收敛，数值稳定性好，对背景噪声不敏感。研究用ＳＡⅡ实现图象滤波的方法，用计算机进行仿真，对结果进行总结分析。
In considering the fact that the two-dimensional （2D） image deformations satisfy the 2D serial autoregressive （SAR） model, adaptive spatial recursive algorithm Ⅱ （SAⅡ） is presented in this paper. The algorithm is different from SA Ⅰ algorithm. SAⅠ is iterating per line and adjusting all of the parameters, but SAⅡ is iterating per point and adjusting only a group of the parameters. The convergence speed of SAⅡ is not then affected, but its calculating operations have greatly decreased. SAⅡ stably converges and has the numeric stabilitines,and it is also not sensitive to baekground noises. The usages of SAⅡ to image filtering are analysed in paper, and computer simulations are performed,the results are analysed and concluded.
出处
《南京理工大学学报》
CAS
CSCD
1996年第2期151-154,共4页
Journal of Nanjing University of Science and Technology
基金
国家自然科学基金资助课题
