计算机科学 ›› 2024, Vol. 51 ›› Issue (5): 92-99.doi: 10.11896/jsjkx.231100067
宋建锋, 张文英, 韩露, 胡国正, 苗启广
SONG Jianfeng, ZHANG Wenying, HAN Lu, HU Guozheng, MIAO Qiguang
摘要: 针对黑白电影的上色过程中,自动上色模型只生成一种结果导致上色结果单一、基于参考示例上色方法需要用户指定参考图像、参考图像的高要求会耗费大量人力的问题,提出了一种多阶段的黑白影像智能色彩修复算法(A Multi-Stage Intelligent Color Restoration Algorithm for Black-and-White Movies,MSICRA)。首先,使用VGG19网络将电影分割为多个场景片段;其次,将每个场景片段逐帧切割,将每帧图像的边缘强度和灰度差作为图像清晰度评判指标,筛选出每个场景中清晰度位于[0.95,1]区间的图像;然后,选择筛选出的图像中的第一张,使用不同的渲染因子值进行上色,利用饱和度进行上色效果的评估,选择合适的渲染因子值对筛选出的图像上色;最后,利用上色前和上色后图像之间的均方误差选择上色质量较好的图像作为该场景片段上色的参考图像。实验结果表明,所提算法在黑白电影《雷锋》和《永不消逝的电波》的PSNR上分别提高了1.32%和2.15%,SSIM分别提高了1.84%和1.04%。该算法不仅可以实现全自动上色,而且颜色真实,符合人们的认知。
中图分类号:
[1]LEI C Y,CHEN Q F.Fully Automatic Video Colorization with Self-Regularization and Diversity[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.Long Beach,USA,2019:3753-3761. [2]LIU Y,ZHAO H,CHAN K C K,et al.Temporally consistent video colorization with deep feature propagation and self-regularization learning[J].arXiv:2110.04562,2021. [3]ZHAO Y,PO L M,YU W Y,et al.VCGAN:video colorization with hybrid generative adversarial network[J].IEEE Transactions on Multimedia,2022,25:3017-3032. [4]ZHAO Y,PO L M,CHEUNG K W,et al.SCGAN:Saliency Map-guided Colorization with Generative Adversarial Network [J].IEEE Transactions on Circuits and Systems for Video Technology,2020,31(8):3062-3077. [5]AKIMOTO N,HAYAKAWA A,SHIN A,et al.Reference-based video colorization with spatiotemporal correspondence[J].arXiv:2011.12528,2020. [6]ZHANG B,HE M M,LIAO J,et al.Deep Exemplar-based VideoColorization[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.Long Beach,USA,2019:8052-8061. [7]SHI M,ZHANG J Q,CHEN S Y,et al.Reference-based deepline art video colorization[J].IEEE Transactions on Visualization & Computer Graphics,2023,29(6):2965-2979. [8]HE M M,CHEN D D,LIAO J,et al.Deep exemplar-based colo-rization[J].ACM Transac tions on Graphics (TOG),2018,37(4):1-16. [9]WAN Z Y,ZHANG B,CHEN D D,et al.Bringing Old FilmsBack to Life[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.New Orleans,USA,2022:17694-17703. [10]JI X,JIANG B,LUO D,et al.ColorForm er:Image colorization via color memory assist ed hybrid-attention transformer[C]//European Conference on Computer Vision.Tel-Aviv,Israel,2022:20-36. [11]ANTIC J.DeOldify [EB/OL].(2018-10-31) [2021-12-10].https://github.com/jantic/DeOld ify. [12]SIMONYAN K,ZISSERMAN A.Very deep convolutional networks for large-scale image recognition[J].arXiv:1409.1556,2014. [13]WANG X M,ZHANG S Y,ZHANG J,et al.Contour Recon-struction Method for Noisy Image Based on Depth Residual Learning[J].Journal of Xidian University,2020,47(3):66-71. [14]VASWANI A,SHAZEER N,PARMAR N,et al.Attention Is All You Need[C]//Proceedings of the Thirty-first Conference on Neural Information Processing Systems.Long Beach,USA,2017:1-11. [15]RONNEBERGER O,FISCHER P,BROX T.U-net:Convolutional networks for biomedical image segmentation[C]//Medical Image Computing and Computer-Assisted Intervention(MICCAI 2015).Munich,Germany,2015:234-241. [16]HE K M,ZHANG X Y,REN S Q,et al.Deep Residual Learning for Image Recogni tion[C]//Proceedings of the IEEE Confe-rence on Computer Vision and Pattern Recognition.Las Vegas,USA,2016:770-778. [17]JIANG G Y,HUANG D J,WANG X,et al.Overview on Image Quality Assessment Methods[J].Journal of Electronics & Information Technology,2010,32(1):219-226. |
|