一种深度图像帧内编码单元快速划分算法

doi:10.11896/jsjkx.180701337

计算机科学 ›› 2019, Vol. 46 ›› Issue (10): 286-294.doi: 10.11896/jsjkx.180701337

一种深度图像帧内编码单元快速划分算法

朱威^1,2, 易瑶¹, 王图强¹, 郑雅羽^1,2

(浙江工业大学信息工程学院杭州310023)¹
(浙江省嵌入式系统联合重点实验室杭州310023)²

收稿日期:2018-07-20 修回日期:2018-11-15 出版日期:2019-10-15 发布日期:2019-10-21
通讯作者: 朱威(1982-),男,博士,副教授,硕士生导师,主要研究方向为视频编解码和智能视觉处理,E-mail:weizhu@zjut.edu.cn。
作者简介:易瑶(1992-),女,硕士生,主要研究方向为三维视频编码;王图强(1995-),男,硕士生,主要研究方向为三维视频处理;郑雅羽(1978-),男,博士,副研究员,硕士生导师,主要研究方向为视频信息处理。
基金资助:
本文受浙江省自然科学基金(LY17F010013),国家自然科学基金(61401398)资助。

Fast Coding Unit Partition Algorithm for Depth Maps

ZHU Wei^1,2, YI Yao¹, WANG Tu-qiang¹, ZHENG Ya-yu^1,2

(College of Information Engineering,Zhejiang University of Technology,Hangzhou 310023,China)¹
(United Key Laboratory of Embedded System of Zhejiang Province,Hangzhou 310023,China)²

Received:2018-07-20 Revised:2018-11-15 Online:2019-10-15 Published:2019-10-21

摘要/Abstract

摘要： 新一代的三维视频编码标准——3D-HEVC (3D High Efficient Video Coding)为了显著减少视点个数,增加了包含视频场景几何信息的深度图像,但深度图像编码的计算复杂度非常高,其编码时间是彩色图像的4倍左右。为了降低深度图像编码的计算复杂度,文中提出了一种基于纹理特征分析的深度图像帧内编码单元(CU)快速划分算法。首先,对深度图像的编码树单元(CTU)进行初级纹理特征分析,根据深度图像的纹理变化特征,在大津法的基础上对全局灰度进行分级,再通过判断CTU内采样点的纹理复杂度以及纹理方向标识来确定当前CTU的划分趋势。然后,对纹理复杂度高的CTU进行CU级别的精细纹理特征分析,利用CU内部像素分布的统计特征,自底向上计算不同尺寸的CU的纹理划分标识。最后,根据CTU的纹理复杂度、纹理方向标识以及CU的纹理划分标识预测当前CTU的划分深度范围,并判断是否提前终止CU划分。实验结果表明,与3D-HEVC参考模型中的原始算法相比,所提算法在平均增加0.8%左右码率的同时,能够降低45%左右的编码时间,同时保持了良好的编码率失真性能;与现有的3种快速算法相比,所提算法在整体序列上分别降低了约12%,3%,4%的编码时间,而在大分辨率序列上则分别降低了14%,11%,10%的编码时间,并具有相近的编码率失真性能。

关键词: 3D-HEVC, 深度图像, 纹理特征分析, 帧内编码

Abstract: The 3D high efficient video coding (3D-HEVC) is the new generation of video coding standard,which adopts depth maps to reduce the number of viewpoints significantly.Depth maps contain geometric information which can improve the video compression efficiency,but the depth image encoding has a heavy computation and takes about 4 times as long as the color image encoding.In this paper,in order to reduce the computational complexity of the depth image coding,a coding unit (CU) partition algorithm based on texture analysis was proposed for 3D-HEVC intra coding.Firstly,the rough texture analysis is performed for the depth map,the global grayscale classification based on the OTSU method is calculated through the texture characteristics of the depth map,and the texture complexity and the direction identification of CTU are determined by the sample points in CTU.Then,the fine texture analysis is performed on a CTU with high texture complexity,and the statistical features of the pixel distribution in the CUs are used to compute the textural division flags from bottom to top for different size of CUs.Finally,the texture complexity of CTU,texture direction flags and CU texture division flags are utilized to predict the depth range of current CTU and decide whether to terminate the division of CU.Compared with the original algorithm in 3D-HEVC test model,the proposed algorithm can reduce 45% encoding time on average with only 0.8% increase in Bjontegaard delta bit rate under the all-intra configuration.Compared with three state-of-the-art algorithms,the proposed algorithm reduces the encoding time by about 12%,% and 4% respectively for overall sequences,and 14%,11% and 10% respectively for the large-resolution sequences,with a similar encoding rate distortion performance.

Key words: 3D-HEVC, Depth map, Intra coding, Texture analysis

中图分类号:

TP391

朱威, 易瑶, 王图强, 郑雅羽. 一种深度图像帧内编码单元快速划分算法[J]. 计算机科学, 2019, 46(10): 286-294. https://doi.org/10.11896/jsjkx.180701337

ZHU Wei, YI Yao, WANG Tu-qiang, ZHENG Ya-yu. Fast Coding Unit Partition Algorithm for Depth Maps[J]. Computer Science, 2019, 46(10): 286-294. https://doi.org/10.11896/jsjkx.180701337

参考文献

[1]UREY H,CHELLAPPAN U H,ERDEN E,et al.State of the Art in Stereoscopic and Autostereoscopic Displays[J].Procee-dings of the IEEE,2011,99(4):540-555.
[2]MÜLLER K,SCHWAR H,MARPE D,et al.3D High-Efficiency Video Coding for Multi-View Video and Depth Data[J].IEEE Transactions on Image Processing,2013,22(9):3366-3378.
[3]MÜLLER K,SCHWAR H,MARPE D,et al.Overview of the Multiview and 3D Extensions of High Efficiency Video Coding[J].IEEE Transactions on Image Processing,2016,26(1):35-49.
[4]SULLIWAN G J,OHM J R,HAN W J,et al.Overview of the High Efficiency Video Coding(HEVC) Standard[J].IEEE Transactions on Circuits and Systems for Video Technology,2012,22(12):1649-1668.
[5]KAUFF P,ATZPADIN N,FEHN C,et al.Depth Map Creation and Image-based Rendering for Advanced 3DTV Services Providing Interoperability and Scalability [J].Signal Processing:Image Communication,2007,22(2):217-234.
[6]CEN Y F,WANG W L,YAO X W,et al.Decision Tree Based Coding Unit Splitting Algorithm for HEVC[J].Computer Scie-nce,2016,43(4):308-312.(in Chinese)
岑跃峰,王万良,姚信威,等.基于决策树的HEVC编码单元划分算法[J].计算机科学,2016,43(4):308-312.
[7]LIU S,PENG Z J,CHEN J L,et al.Multi-class Support Vector Maschine-based Fast Algorithm for 3D-HEVC Depth Video Intra Coding[J].Journal on Communications,2016,37(11):181-188.(in Chinese)
刘晟,彭宗举,陈嘉丽,等.基于多类支持向量机的3D-HEVC深度视频帧内编码快速算法[J].通信学报,2016,37(11):181-188.
[8]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:IEEE Press,2016:1126-1130.
[9]CHUNG K L,HUANG Y H,LIN C H,et al.Novel Bitrate Sa-ving and Fast Coding for Depth Videos in 3D-HEVC[J].IEEE Transactions on Circuits and Systems for Video Technology,2016,26(10):1859-1869.
[10]ZHANG H B,CHAN Y L,FU C H,et al.Quadtree Decision for Depth Intra Coding in 3D-HEVC by Good Feature[C]//IEEE International Conference on Acoustics,Speech and Signal Processing (ICASSP).Shanghai:IEEE Press,2016:1481-1485.
[11]SILVA T L,AGOSTINI L V,CRUZ L A S.Fast Intra Prediction Algorithm Based on Texture Analysis for 3D-HEVC Encoders[J].Journal of Real-Time Image Processing,2016,12(2):357-368.
[12]TOHIDYPOUR H R,POURAZAD M T,NASIOPOULOS P.Online-Learning-Based Complexity Reduction Scheme for 3D-HEVC[J].IEEE Transactions on Circuits and Systems for Vi-deo Technology,2016,26(10):1870-1883.
[13]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.
[14]GUO R X HE G,LI Y S,et al.Fast Algorithm for Prediction Unit and Mode Decisions of Intra Depth Coding in 3D-HEVC[C]//IEEE International Conference on Image Processing (ICIP).Phoenix:IEEE Press,2016:1121-1125.
[15]FU C H,ZHAO Y W,ZHANG H B,et al.Depth Modelling Mode Decision for Depth Intra Coding via Good Feature[C]//IEEE International Conference on Image Processing (ICIP).Beijing:IEEE Press,2017:4018-4022.
[16]JABALLAH S,LARABI M C,TAHAR J B.Heuristic Inspired Search Method for Fast Wedgelet Pattern Decision in 3D-HEVC[C]//European Workshop on Visual Information Processing (EUVIP).Marseille:IEEE Press,2016:1-6.
[17]LEI J J,SUN Z Y,GU Z Y,et al.Simplified Search Algorithm for Explicit Wedgelet Signalization Mode in 3D-HEVC[C]//IEEE International Conference on Multimedia and Expo (ICME).Dhaka:IEEE Press,2017:805-810.
[18]OHTSU N.A Threshold Selection Method from Gray-Level Histograms[J].IEEE Transactions on systems,Man,and Cybernetics,1979,9(1):62-66.
[19]MÜLLER K,VETRO A.Common Test Conditions of 3DV Core Experiments[C]//ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG 11 11th Meeting.Geneva,2014:1-7.
[20]BJONTEGAARD G.Calculation of Average PSNR Differences between RD Curves [C]//Proceedings of the VCEG-M33.Austin,2001:1-4.

相关文章 10

[1]	刘月红, 牛少华, 神显豪. 基于卷积神经网络的虚拟现实视频帧内预测编码 Virtual Reality Video Intraframe Prediction Coding Based on Convolutional Neural Network 计算机科学, 2022, 49(7): 127-131. https://doi.org/10.11896/jsjkx.211100179
[2]	孟文权, 武利生. 基于深度特征的足底曲面三角剖分重构 Triangulation Reconstruction of Plantar Surface Based on Depth Feature 计算机科学, 2019, 46(6A): 205-207.
[3]	宋相法,张延锋,郑逢斌. 基于L_2,1范数稀疏特征选择和超法向量的深度图像序列行为识别 Activity Recognition from Depth Image Sequences Based on L_2,1-norm Sparse Feature Selection and Super Normal Vector 计算机科学, 2017, 44(2): 306-308. https://doi.org/10.11896/j.issn.1002-137X.2017.02.052
[4]	李文莉,高宏伟,冀大雄,李岩. 基于遗传算法的海底沉积物纹理特征优化方法 Optimization Method of Seabed Sediment Texture Feature Based on Genetic Algorithm 计算机科学, 2016, 43(Z6): 130-133. https://doi.org/10.11896/j.issn.1002-137X.2016.6A.031
[5]	张鸿宇,刘威,许炜,王辉. 基于深度图像的多学习者姿态识别 Depth Image Based Gesture Recognition for Multiple Learners 计算机科学, 2015, 42(9): 299-302. https://doi.org/10.11896/j.issn.1002-137X.2015.09.059
[6]	郭庆慧,梁秀霞,张锐. 基于稀疏特征点的单视点深度图像校准 Calibration for Single Viewpoint Depth Image Based on Sparse Feature Points 计算机科学, 2014, 41(Z6): 191-195.
[7]	李康,尚鹏,耿国华. 基于深度图像的人脸模型特征点自动标定 Automatic Location of Feature Points on Three-dimensional Facial Model Based on Depth Image 计算机科学, 2014, 41(4): 287-291.
[8]	杨全,彭进业. 采用SIFT-BoW和深度图像信息的中国手语识别研究 Chinese Sign Language Recognition Research Using SIFT-BoW and Depth Image Information 计算机科学, 2014, 41(2): 302-307.
[9]	李幸,赵耀,林春雨,姚超. 基于小波子带系数预测的深度图像编码 Depth Map Coding Based on Wavelet Inter-subband Coefficients Prediction 计算机科学, 2014, 41(10): 134-138. https://doi.org/10.11896/j.issn.1002-137X.2014.10.031
[10]	董鹏宇,林涛. 无再损帧内编码 Re-loss-free Intra Coding Using Integer Linear Programming for H. 264/MPEG-4 AVC 计算机科学, 2009, 36(9): 258-261.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

一种深度图像帧内编码单元快速划分算法

Fast Coding Unit Partition Algorithm for Depth Maps

PDF (PC)

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 10

Metrics

本文评价

推荐阅读 0