关于模糊神经网络在图像质量评价问题上的应用
he Application Base On The Fuzzy and BP Neural Network In The Image Quality Evaluation
针对当前图像质量评价方法上所存在的不足，本文提出一种基于模糊神经网络进行图像质量评价的新方法。首先，用一个可行的主观保真度绝对评价尺度来建立评价模型，并用小波变换提取图像的亮度、模糊度和相关度指标来综合表示图像的质量，并进行评分。然后，根据以上评价结果建立相应的评价矩阵Ak,Bk。用模糊联想器FAM建立二者的关联矩阵记录下每个图像的综合评分，为图像评价建立了更加客观的评价体系。通过FAM实现对图像质量进行评价，将模糊理论和神经网络相结合，有效避免了传统评价方法的主观性和简单性，使模糊神经网络在图像质量评价问题上有着良好的结果
Because of the defects of the current method of image quality evaluation ,this article put forward a new method about image quality evaluation through technology of fuzzy andneural networks.First, establish evaluation model with a viable Subjective fidelity-Absolute Evaluation Scale, and uses the wavelet transformation to extract brightness, the ambiguity and the degree of correlation target from images for carrying on grades.Then, establish corresponding appraisal matrix Ak, Bk according to above measure results.With FAM eestablishe the two\\\
何巍
计算技术、计算机技术自动化技术、自动化技术设备
模糊神经网络，图像质量评价，FAM网络
Fuzzy and Neural NetworksImage Quality Evaluation FAM network model
何巍.关于模糊神经网络在图像质量评价问题上的应用[EB/OL].(2008-04-21)[2025-08-22].http://www.paper.edu.cn/releasepaper/content/200804-764.点此复制
评论