Improved Motion Compensation Temporal Filtering (MCTF) for Scalable Video Coding (H.264/SVC) Standards

被引：0

作者：

Vora, Ankur ^{[1
]}

Chandramouli, Krishna ^{[2
]}

机构：

[1] Sch Informat Technol & Engn, Vellore, Tamil Nadu, India

[2] Vellore Inst Technol, Vellore, Tamil Nadu, India

来源：

2015 2ND INTERNATIONAL CONFERENCE ON ELECTRONICS AND COMMUNICATION SYSTEMS (ICECS) | 2015年

关键词：

Dyadic partition; Lifting Based Approach; Fine Granular Scalability (FGS); MODEL;

D O I：

暂无

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

0808 ; 0809 ;

摘要：

The newer of the ITU/MPEG standardization effort of H.264/AVC scalable extension, Joint Scalable Video Model (JSVM) has been made for efficient Scalable Video Coding (SVC). The Motion Estimation (ME) and Motion Compensation (MC) techniques are applied to exploit the redundancy between two consecutive frames. This paper deals with application of Multi-resolutional Discerete Wavelet Transform (DWT) to improve channel bandwidth aspects and the visual quality of the video in error prone network. A gradient based approach is used to recover the lost pixels via finding the direction of maximum amount of correctly received pixels and concoles the lost pixels. The algorithm applied in a temporal domain on the spatial frequency of the Motion Vectors (MV). To evaluate the proposed algorithm an open source reference software for SVC called Joint Scalable Video Model (JSVM) developed by ITU is used. In the results shows the proposed algorithm gets the significant improvement in visual quality for error prone networks with reduction in bandwidth & improves PSNR.

引用

页码：602 / 606

页数：5