摘要
户外视觉系统在雨(雪)天气条件下获取的视频图像会产生严重退化.提出一种基于小波融合的视频图像去雨(雪)方法,从频率域角度分析,采用小波多层分解和小波融合技术:先确定雨(雪)噪声所在的具体层,再制定一种基于雨(雪)噪声污染度的融合规则,最后对多幅连续退化图像的特定层进行小波融合,以达到去除雨(雪)的目的.仿真结果表明,本文方法的复原结果理想,不受噪声强度制约.
Due to the serious degradation of the video images captured by the outdoor vision system in rain（ snow） weather,a method of removing rain（ snow） from video images based on wavelet fusion is proposed.This method analyzes from the perspective of frequency domain,uses multilevel wavelet decomposition and wavelet fusion.First of all,it determines where the specific layer number of rain（ snow） noise is,and secondly,it formulates a fusion rule based on the degree of rain（ snow） noise pollution,thirdly,it fuses the specific layers of multiple continuous degraded images to remove rain（ snow）.The simulation result shows that the restoration result of this method is ideal and free from noise intensity.
出处
《北华大学学报（自然科学版）》
CAS
2018年第1期135-140,共6页
Journal of Beihua University（Natural Science）
基金
吉林市科技局杰出青年研究项目(20156412)
关键词
雨(雪)去除
图像复原
小波多层分解
小波融合
融合规则
雨(雪)噪声污染度
rain （ snow） removal
image restoration
wavelet multi-level decomposition
wavelet fusion
fusion rule
rain （ snow） noise pollution
