论文部分内容阅读
本文一改以往的以仿真的感性效果作为信号边缘检测质量的效果评价方法,提出小波变换边缘检测定位精度和抗噪声能力量化分析方法。基于小波变换的边缘检测算法,物理意义上是一个先平滑,再进行边缘检测的过程,其边缘检测特性与边缘类型和尺度大小有关。随尺度增大,定位偏差增大,反映了小波变换局部化特征强弱对边缘检测特性的影响。本文给出了不含噪声和含有噪声情况下,典型边缘定位精度的量化表述。
This paper amends the previous method of evaluating the effect of signal edge detection by using the perceptual effect of simulation, and proposes a quantitative analysis method of wavelet transform edge detection location accuracy and anti-noise ability. Edge detection based on wavelet transform algorithm, the physical sense is a first smooth, then the edge detection process, the edge detection characteristics and the edge type and size of the size. As the scale increases, the positioning deviation increases, reflecting the influence of the wavelet transform local feature strength on the edge detection characteristics. In this paper, a quantitative description of typical edge location accuracy without noise and noise is given.