摘要
针对局部失焦模糊图像的模糊程度测量问题,提出一种基于BP神经网络的图像局部失焦模糊测量方法,该方法通过提取图像块频域和空域的多个特征组成特征向量,并为每个特征向量设置模糊程度标签,然后通过BP神经网络训练分类器,实现待测图像块的模糊程度估计。实验结果表明,该方法可以有效测量出局部失焦模糊图像的模糊区域以及模糊程度。
As for the problem of local defocus blur image blur degree measurement,presents a kind of based on BP neural network of local defocus blur image measurement method.Firstly,the extracted multiple features from image blocks frequency domain and the airspace consist of feature vector,and labels blur degree for each feature vector.Then BP neural network-based classifier is trained to estimate the blur mea?surement of each image block.Experimental results show that the method can effectively measure the blur regions and degree of local defo?cus blur images.
作者
戴柳云
DAI Liu-yun(College of Computer and Information Science,Chongqing Normal University,Chongqing 401331)
出处
《现代计算机》
2018年第20期26-30,共5页
Modern Computer
关键词
BP神经网络
模糊测量
失焦模糊
模糊特征
Back-Propagation
Blur Measurement
Defocus Blur
Extraction of Characteristics
