首页 > 论文 > 激光与光电子学进展 > 56卷 > 13期(pp:131001--1)

基于卷积神经网络和边缘检测的自然纹理合成算法

Natural Texture Synthesis Algorithm Based on Convolutional Neural Network and Edge Detection

  • 摘要
  • 论文信息
  • 参考文献
  • 被引情况
  • PDF全文
分享:

摘要

基于卷积神经网络(CNN)的VGG-19(Visual Geometry Group)模型,研究了卷积神经网络对输入纹理进行卷积时,输入纹理特征图的边缘信息对生成自然纹理效果的影响。在使用卷积神经网络的VGG对输入图像进行卷积时,为了防止过拟合现象,采用平均池化的方式对特征图进行处理,在一定程度上保护了特征图的边缘信息,相对采用最大池化处理特征图取得了更好的生成效果。同时,提取各层特征图的边缘信息并将其叠加到特征图中,能很好地保留纹理图像的边缘结构信息。实验结果表明,改进后的方法能取得较为理想的纹理生成效果。

Abstract

Based on the Visual Geometry Group (VGG-19) model of convolutional neural networks (CNN), influences of the edge information in an input texture feature map on the natural texture are studied when the CNN convolves the input texture. When the input image is convoluted by the VGG using the CNN, the feature map is processed in an average pooling manner to prevent overfitting, which protects the edge information of the feature map to some extent and the generation effect is better than that obtained via max-pooling processing. The edge information of each layer of feature map is extracted and superimposed on the feature map, which preserves the edge structure information of the texture image well. Experimental results demonstrate that the proposed method achieves a good texture generation effect.

Newport宣传-MKS新实验室计划
补充资料

DOI:10.3788/LOP56.131001

所属栏目:图像处理

基金项目:贵州省教育厅青年科技人才成长项目、贵州省科技合作计划、国家自然科学基金;

收稿日期:2018-11-13

修改稿日期:2019-01-22

网络出版日期:2019-07-01

作者单位    点击查看

张定祥:贵州电子信息职业技术学院电子信息工程系, 贵州 凯里 556000
谭永前:凯里学院大数据工程学院, 贵州 凯里 556011

联系人作者:谭永前(tanyongqian1@163.com)

备注:贵州省教育厅青年科技人才成长项目、贵州省科技合作计划、国家自然科学基金;

【1】Wei L Y and Levoy M. Texture synthesis over arbitrary manifold surfaces. [C]∥The 28th Annual Conference on Computer Graphics and Interactive Techniques, August 12-17, 2001, Los Angles. New York: ACM. 355-360(2001).

【2】Tong X, Zhang J D, Liu L G et al. Synthesis of bidirectional texture functions on arbitrary surfaces. [C]∥The 29th Annual Conference on Computer Graphics and Interactive Techniques, July 23-26, 2002, San Antonio, Texas. New York: ACM. 665-672(2002).

【3】Ashikhmin M. Synthesizing natural textures. [C]∥Proceedings of the 2001 Symposium on Interactive 3D Graphics, March 19-21, 2001, North Carolina. New York: ACM. 217-226(2001).

【4】Efros A A and Freeman W T. Image quilting for texture synthesis and transfer. [C]∥The 28th Annual Conference on Computer Graphics and Interactive Techniques, August 12-17, 2001, Los Angles. New York: ACM. 341-346(2001).

【5】Xue F, Cheng C and Jiang J L. Wang Tile-based improved texture synthesis. Journal of Computer Applications. 30(8), 2098-2100, 2156(2010).
薛峰, 成诚, 江巨浪. 基于Wang Tile的改进纹理合成算法. 计算机应用. 30(8), 2098-2100, 2156(2010).

【6】Kwatra V, Schoed L A, Essa I et al. Graphcut textures image and video synthesis using graph cuts. [C]∥The 30th Annual Conference on Computer Graphics and Interactive Techniques, July 27-31, 2003, San Diego, California. New York: ACM. 277-286(2003).

【7】Gatys L A and Ecker A S. -12-06)[2018-11-01]. https:∥arxiv. org/abs/1505, (2015).

【8】Cadieu C F and Hong H. Yamins D L K, et al. Deep neural networks rival the representation of primate IT cortex for core visual object recognition. PLoS Computational Biology. 10(12), (2014).

【9】Simonyan K. -04-10)[2018-11-01]. https:∥arxiv. org/abs/1409, (2015).

【10】Portilla J and Simoncelli E P. A parametric texture model based on joint statistics of complex wavelet coefficients. International Journal of Computer Vision. 40(1), 49-70(2000).

【11】LeCun Y A, Bottou L, Orr G B et al. . Efficient backprop. ∥Montavon G, Orr G B, Müller K R. Neural networks: tricks of the trade. Heidelberg: Springer. 7700, 9-48(2012).

【12】Jaderberg M, Vedaldi A and Zisserman A. Speeding up convolutional neural networks with low rank expansions[C]∥Proceedings of the British Machine Vision Conference 2014, September 1-5, 2014, Nottingham, UK. Durham, England,. (2014).

【13】Jia Y Q, Shelhamer E, Donahue J et al. Caffe: convolutional architecture for fast feature embedding. [C]∥The 22nd ACM International Conference on Multimedia, November 3-7, 2014, Orlando, Florida. New York: ACM. 675-678(2014).

【14】Okazawa G, Tajima S and Komatsu H. Image statistics underlying natural texture selectivity of neurons in macaque V4. Proceedings of the National Academy of Sciences. 112(4), E351-E360(2015).

【15】Xie J, Lu Y, Gao R, generator networks[EB/OL] et al. -10-29)[2018-11-01]. https:∥arxiv. org/abs/1609, (2018).

引用该论文

Dingxiang Zhang, Yongqian Tan. Natural Texture Synthesis Algorithm Based on Convolutional Neural Network and Edge Detection[J]. Laser & Optoelectronics Progress, 2019, 56(13): 131001

张定祥, 谭永前. 基于卷积神经网络和边缘检测的自然纹理合成算法[J]. 激光与光电子学进展, 2019, 56(13): 131001

您的浏览器不支持PDF插件,请使用最新的(Chrome/Fire Fox等)浏览器.或者您还可以点击此处下载该论文PDF