期刊文献+

基于深度学习的车牌超分辨率重建

Super-resolution Reconstruction of License Plates Based on Deep Learning
下载PDF
导出
摘要 车牌图像重建是实现智能交通的重要步骤.在经过不断的重复实验后,本文提出了一种新的基于生成对抗网络(GAN)的超分辨率车牌图像重建模型.所提出的办法主要包括4个部分:(1)预处理输入图像,包括调整图片大小和筛选对比度差的图片;(2)引入了残差密集网络,能够充分提取车牌图像特征;(3)引入渐进式采样进行图片重建,因其具有较大的感受野,能提供更多的信息细节;(4)引入基于PatchGAN的鉴别器模型,该模型能更加精准地判断,从而引导生成器进行更高质量、更多细节的图像重建.通过在CCPD数据集上与目前较优的算法进行比较,证明本文的模型重建的车牌图像具有较高的PSNR和SSIM,分别达到了26.80和0.77,而且重建单帧图像的花费时间更少,仅为0.06 s,进而证明了我们算法的可行性. License plate image reconstruction plays an important role in the intelligent transportation system.After repeated experiments,a super-resolution image reconstruction method for license plates is proposed with the help of generative adversarial networks(GANs).The method mainly consists of four parts:(1)pretreatment of the input image,including image resizing and filtering of images with poor contrast;(2)image feature extraction using a residual dense network;(3)introduction of progressive sampling,which can provide a larger receptive field and more information details;(4)introduction of a discriminator based on PatchGAN to make a more accurate judgment,which guides the generator to reconstruct images with higher quality and more details.The comparison with a current superior algorithm on the Chinese City Parking Dataset(CCPD)proves that the proposed model has higher PSNR and SSIM(26.80 and 0.77,respectively)and less time of reconstructing a single-frame image(only 0.06 s),which verifies the feasibility of the proposed approach in license plate image reconstruction.
作者 刘良鑫 林勉芬 周成菊 潘家辉 LIU Liang-Xin;LIN Mian-Fen;ZHOU Cheng-Ju;PAN Jia-Hui(School of Software,South China Normal University,Foshan 528225,China)
出处 《计算机系统应用》 2022年第2期234-240,共7页 Computer Systems & Applications
基金 广州市科技计划项目重点领域研发计划(202007030005) 广东省自然科学基金面上项目(2019A1515011375) 国家自然科学基金面上项目(62076103)。
关键词 超分辨率图像重建 生成对抗网络(GAN) 残差密集网络 渐进式上采样 super-resolution reconstruction generative adversarial network(GAN) residual dense network progressive upsampling
  • 相关文献

参考文献5

二级参考文献25

  • 1韩玉兵,陈小蔷,吴乐南.一种视频序列的超分辨率重建算法[J].电子学报,2005,33(1):126-130. 被引量:8
  • 2H S Hou, H C Andrews. Cubic spline for image interpolation and digital filtering [J]. IEEE Transaction on Signal Pressing, 1978,26(6) :508 - 517.
  • 3S Mallet, Guoshen Yu. Super-Resolution with sparse mixing es- timators [ J]. IEEE Transactions on Image Processing, 2010, 19 ( 11 ) : 2889 - 2900.
  • 4W T Freeman, T R Jones, E C Pasztor. Example-based super- resolution [ J ]. IEEE Computer Graphics and Applications, 2002,22(2) :56 - 65.
  • 5M Elad, D Datsenko. Example-based regularization deployed to super-resolution reconstruction of a single image [ J ]. The Computer Journal, 2007,50(4) : 1 - 16.
  • 6Yang Jian-chao, J Wright, T S Huang, Yi Ma. Image super-res- olution via sparse representation [J]. 1EEE Transaction on Im-age Procesfing,2010,19(ll):2861 - 2873.
  • 7Yang Jian-chao, J Wright, T S Huang, Yi. Ma, Image super- resolution as sparse representation of raw image patches [ A]. Proceedings of the 1F, IEEE Conference on Computer Vision and Pattern Recognition[ C]. Anchorage, AK, 2008.1 - 8.
  • 8R Zeyde, M Elad, M Protter. On single image scale-up using sparse-representations [ A] .Proceedings of the 7th International Conference on Curves and Surfaces [ C ]. Avignon: Avignon, France, 2010.
  • 9M Aharon, M Elad, A Bruckstein, The K-SVD: an algorithm for designing of overcomplete dictionaries for sparse represen- tation [ J 3. IEEE, Transaction on Signal Processing, 2006, 54 (11) :4311 - 4322.
  • 10R Rubinstein, M Zibulevsky,M Elad. Efficient implementation of the K-SVD algorithm using batch orthogonal matching pur- suit [ J/OL ]. http://www, cs. le, chnionac, il/N ronrubin/Publications/KSVD--OMP- v2. pdf, 2008-03-15.

共引文献91

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部