计算机工程 ›› 2019, Vol. 45 ›› Issue (3): 212-216,224.doi: 10.19678/j.issn.1000-3428.0048890

• 图形图像处理 • 上一篇    下一篇

基于深度学习的图像美观度评价

费延佳,李福翠,邵枫   

  1. 宁波大学 信息与工程学院,浙江 宁波 315211
  • 收稿日期:2017-10-10 出版日期:2019-03-15 发布日期:2019-03-15
  • 作者简介:费延佳(1992—),男,硕士研究生,主研方向为图像质量评价;李福翠,讲师;邵枫,教授。
  • 基金项目:

    国家自然科学基金(61622109)。

Image Aesthetic Assessment Based on Deep Learning

FEI Yanjia,LI Fucui,SHAO Feng   

  1. Department of Information Science and Engineering,Ningbo University,Ningbo,Zhejiang 315211,China
  • Received:2017-10-10 Online:2019-03-15 Published:2019-03-15

摘要:

基于深度卷积神经网络的特征提取方法比传统手工特征提取方法更加贴近人类大脑的视觉感受。为此,建立一种两通道组合图像美观度评价模型。使用美学信息通道和场景信息通道的组合来自动提取图像中美学信息和场景类别信息,通过融合两类信息最终形成美感分类器。在AVA库上进行训练和测试,结果表明,与图像局部特征提取方法相比,该模型结构较简洁,且具有较高的分类准确率。

关键词: 美观度评价, 卷积神经网络, 场景识别, 两通道, 分类网络

Abstract:

The method of extracting features based on deep Convolutional Neural Network(CNN) is closer to the visual perception of the human brain than the traditional manual extraction feature method.Therefore,a two-channel combination model is proposed.The combination of the aesthetic information channel and the scene information channel is used to automatically extract the aesthetic information and scene category information in the image,and finally combine the two types of information to form an aesthetic classifier.The training and testing on the AVA library show that compared with the image local feature extraction method,the model structure is simple and has high classification accuracy.

Key words: aesthetic assessment, Convolutional Neural Network(CNN), scene recognition, two-channel, classification network

中图分类号: