摘要
本文提出了一种在图象处理和模式识别中非常有效的新方法——可变参数的逐级离散相关法。这种方法在快速的逐级运算基础上,对每级扩展增长率具有任意的可变性。它是对逐级离散相关法的拓广。在具体应用中,该方法体现出灵活多变、计算快速、有效等特点。此外,在相关窗的选取上采用近似高斯分布状的加权窗,对实际图象信号的处理起到了滤去噪声的作用。
In this paper, a new and highly efficient algorithm (which can be) used in image processing and pattern recognition is presented. The algorithm, named Variable Parameter Hierarchical Discrete Correlation (VPHDC), is an extension of HDC, and has some distinguishing features in its applications, such as flexibility, efficiency, fastness and so on. Besides, it also has an effect of filtering on noise because it uses Gaussian-like weighted window.
出处
《自动化学报》
EI
CSCD
北大核心
1990年第1期57-61,共5页
Acta Automatica Sinica
关键词
图象处理
模式识别
目标匹配
Image processing
pattern recognition
object matching.
