关于运动图像模糊区域边界优化降噪仿真

摘要

Reducing the noise from blurred region boundary of motion image can effectively improve the overall visual effect of motion image.In traditional methods,the clustering of feature vectors of pixel is ignored,which resuits in unideal denoising effect.This article presents a method for reducing the noise in boundary of blurred region of motion image based on wavelet transform.The smooth preprocessing of image is carried out,and the image is divided into a plurality of super pixels.Then,the class indicator is added on each super pixel,and category clusters of different pixels are formed.Moreover,the wavelet transform is used to extract low frequency and high frequency characteristics of each cluster,and the fuzzy means method is used to cluster different types of pixel feature vectors.Thus,the noise reduction of blurred region boundary in image is completed.Simulation shows that the proposed method has high accuracy of optimization and noise reduction of blurred region boundary in image.%对运动图像模糊区域边界去噪,能够有效提高视频运动图像整体视觉效果.对运动图像模糊区域的优化降噪,首先需要提取每个集群的低频和高频特征,对不同类型像素特征向量进行聚类,完成运动图像模糊区域的优化降噪.传统方法计算运动图像的粗阚值,给出图像域统计特性近似高斯分布的特点,但忽略了对像素特征向量的聚类,导致图像降噪效果不理想.提出基于小波变换的运动图像模糊区域边界优化降噪方法.对图像进行平滑预处理,将图像划分为多个超像素,对各个超像素增加类别标志,形成不同像素的类别集群,采用小波变换提取每个集群的低频和高频特征,利用模糊均值方法对不同类型像素特征向量进行聚类,完成运动图像模糊区域边界优化降噪.仿真证明,所提方法图像模糊区域边界优化降噪精度较高,有效地完善了模糊区域边界优化降噪图像的视觉效果.