基于FSCD-CNN的深度图像快速帧内预测模式选择算法

doi:10.3969/j.issn.0255-8297.2021.03.009

摘要/Abstract

摘要： 针对3D-HEVC的多视点视频加深度图的编码格式和四叉树编码结构所带来的编码复杂度问题，提出了一种深度图像快速帧内预测模式选择算法。首先，从深度视频序列中以最优的深度图最大编码单元（largest coding unit，LCU）划分深度为标签获取训练集；其次，构建了适用于LCU的Cu深度快速选择卷积神经网络（fast selecting Cu’s depth-convolutionalneural network，FSCD-CNN）；最后，对深度图LCU进行划分深度预测，跳过部分编码模式决策，实现最佳LCU划分。实验结果表明，与相关文献对比，所提算法在保持了编码性能的同时平均减少了15%的编码时间，实验验证了其有效性和可靠性。

关键词: 3D-HEVC, 深度图, 最大编码单元, 卷积神经网络, 编码复杂度

Abstract: In view of the coding complexity caused by the encoding format of multiview video plus depth map and the quadtree coding structure in 3D-HEVC, a fast intra prediction mode selection algorithm for depth images based on FSCD-CNN (fast selecting cu’s depth-convolutional neural network) is proposed. First, a training set is obtained by dividing the depth of the optimal depth map LCU (largest coding unit) of a depth video sequence as labels. Second, a FSCD-CNN network is constructed, which is suitable for deep decision-making of LCU. At last, the optimal division of LCU is achieved by carrying out the depth-division prediction of depth map LCU and skipping some coding mode decisions. Experimental results show that the proposed algorithm could reduce the coding time by 15% on average while maintaining the same coding performance as other relevant literatures, and verify the effectiveness and reliability of this method.

Key words: 3D-HEVC, depth map, largest coding unit (LCU), convolutional neural network (CNN), coding complexity

中图分类号:

TP319.4

崔鹏涛, 张倩, 刘敬怀, 周超, 王斌, 司文. 基于FSCD-CNN的深度图像快速帧内预测模式选择算法[J]. 应用科学学报, 2021, 39(3): 433-432.

CUI Pengtao, ZHANG Qian, LIU Jinghuai, ZHOU Chao, WANG Bin, SI Wen. FSCD-CNN Based Fast Mode Decision Algorithm for Intra-prediction in Depth Map Coding[J]. Journal of Applied Sciences, 2021, 39(3): 433-432.

参考文献

[1] Song Y, Ho Y S. Unified depth intra coding for 3D video extension of HEVC[J]. Signal Image & Video Processing, 2014, 8(6): 1031-1037.
[2] Saldanha M, Zatt B, Porto M, et al. Solutions for DMM-1 complexity reduction in 3DHEVC based on gradient calculation[C]//IEEE the 7th Latin American Symposium on Circuits & Systems (LASCAS), Florianopolis, Brazil, 2016: 211-214.
[3] Zhang H B, Chan Y L, Fu C H, et al. Quadtree decision for depth intra coding in 3DHEVC by good feature[C]//IEEE International Conference on Acoustics, Speech and Signal Processing, Shanghai, China, 2016: 1481-1485.
[4] Jaballah S, Larabi M C, Tahar J B. Heuristic inspired search method for fast wedgelet pattern decision in 3D-HEVC[C]//The 6th European Workshop on Visual Information Processing (EUVIP), Marseille, France, 2016: 1-6.
[5] Guo L, Tian X, Chen Y. Simplified depth intra coding for 3D-HEVC based on gray-level cooccurrence matrix[C]//IEEE International Conference on Signal and Image Processing (ICSIP), Beijing, China, 2016: 328-332.
[6] Li T, Yu L, Wang S, et al. Simplified depth intra coding based on texture feature and spatial correlation in 3D-HEVC[C]//Data Compression Conference, Snowbird, UT, 2018: 421.
[7] Fu C H, Zhao Y W, Zhang H B, et al. Depth modeling mode decision for depth intra coding via good feature[C]//2017 IEEE International Conference on Image Processing (ICIP), Beijing, China, 2017: 4018-4022.
[8] Zhang H B, Fu C H, Chan Y L, et al. Probability-based depth intra-mode skipping strategy and novel VSO metric for DMM decision in 3D-HEVC[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2018, 28(2): 513-527.
[9] Zhang Q, Jing R H, Wang B, et al. Fast mode decision based on gradient information in 3D-HEVC[J]. IEEE Access, 2019, 7: 135448-135456.
[10] Ren H, Bai H, Lin C, et al. Just noticeable difference based fast coding unit partition in 3DHEVC intra coding[C]//Data Compression Conference (DCC), Snowbird, UT, 2016: 629-629.
[11] Avila G, Conceicao R, Bubolz T, et al. Complexity reduction of 3D-HEVC based on depth analysis for background and ROI classification[C]//The 25th European Signal Processing Conference (EUSIPCO), Kos, Greece, 2017:1031-1035.
[12] Hamout H, Elyousfi A. Fast texture intra size coding based on big data clustering for 3D-HEVC[C]//IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Calgary, AB, 2018: 1728-1732.
[13] Saldanha M, Sanchez G, Marcon C, et al. Fast 3D-HEVC depth maps intra-frame prediction using data mining[C]//IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Calgary, AB, 2018: 1738-1742.
[14] Liu X, Li Y, Liu D, et al. An adaptive CU size decision algorithm for HEVC intra prediction based on complexity classification using machine learning[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2017, 29(1): 144-155.
[15] Xu M, Li T, Wang Z, et al. Reducing complexity of HEVC: a deep learning approach[J]. IEEE Transactions on Image Processing, 2018, 27(10): 5044-5059.
[16] Li Y, Liu Z, Ji X, et al. CNN based CU partition mode decision algorithm for HEVC inter coding[C]//2018 25th IEEE International Conference on Image Processing (ICIP), Athens, Greece, 2018: 993-997.
[17] Feng Z, Liu P, Jia K, et al. Fast intra CTU depth decision for HEVC[J]. IEEE Access, 2018, 6: 45262-45269.
[18] Katayama T, Kurod K, Shi W, et al. Low-complexity intra coding algorithm based on convolutional neural network for HEVC[C]//2018 International Conference on Information and Computer Technologies (ICICT), DeKalb, 2018: 115-118.
[19] Wei Y, Wang Z, Xu M, et al. An LSTM method for predicting CU splitting in H.264 to HEVC transcoding[C]//2017 IEEE Visual Communications and Image Processing (VCIP), Petersburg, 2017: 1-4.
[20] Jing R H, Zhang Q, Wang B, et al. CART-based fast CU size decision and mode decision algorithm for 3D-HEVC[J]. Signal, Image and Video Processing, 2019, 13(2): 209-216
[21] Peng K K, Chiang J C, Lie W N. Low complexity depth intra coding combining fast intra mode and fast CU size decision in 3D-HEVC[C]//IEEE International Conference on Image Processing (ICIP), Phoenix, AZ, 2016: 1126-1130.