Amina Kessentini

Statistical analysis of SHVC encoded video

Scalable extension (SHVC) of the High Efficiency Video Coding HEVC was developed to ameliorate the coding efficiency for different types of scalability. In this paper, a survey for SHVC extensions is investigated. We describe its scalability types and explain the different additional coding features that further improve the Enhancement Layer (EL) coding efficiency. Furthermore, we evaluate and prove the effectiveness of the SHVC through experimental results for different coding configurations reduction of about. Based on statistical analysis, depth 0 (64×64) can reach more than 75% of use in B1 frame. This statistics helped us explore deeply on the prediction mode percentages.

Scalable extension of the high efficiency video coding SHEVC performance study

Scalable High Efficiency Video Coding (SHVC) is the extension of the High Efficiency Video Coding (HEVC). This standard is developed to ameliorate the coding efficiency for the spatial and quality scalability. In this paper, we investigate a survey for SHVC extension. We describe also its types and explain the different additional coding tools that further improve the Enhancement Layer (EL) coding efficiency. Furthermore, we assess through experimental results the performance of the SHVC for different coding configurations. The effectiveness of the SHVC was demonstrated, using two layers, by comparing its coding adequacy compared to simulcast configuration and HEVC for enhancement layer using HM16 for several test sequences and coding conditions.

Depth Partitioning and Coding Mode Selection Statistical Analysis for SHVC

The Scalable High Efficiency Video Coding (SHVC) has been proposed to improve the coding efficiency. However, this additional extension generally results an important coding complexity. Several studies were performed to overcome the complexity through algorithmic optimizations that led to an encoding time reduction. In fact, mode decision analysis is imperatively important in order to have an idea about the partitioning modes based on two parameters, such as prediction unit size and frame type. This paper presents statistical observations at two levels: coding units (CUs) and prediction units (PUs) selected by the encoder. Analysis was performed for several test sequences with different motion and texture characteristics. The experimental results show that the percentage of choosing coding or prediction unit size and type depends on sequence parameters, frame type, and temporal level.

Fast inter-prediction algorithms for spatial Scalable High efficiency Video Coding SHVC

AbstractScalable High Efficiency Video Coding (SHVC) has aggregated an exhaustive algorithm for mode decision based on a recursive quad-tree structure coding unit. However, this comes at the cost of high computational complexity. Therefore, in order to improve the SHVC in terms of encoding time, fast inter-mode decision algorithms for spatial scalability were suggested in this paper. This study focuses on using statistical analysis and the correlation between base layer and enhanced layer modes to speed up the mode decision algorithm for inter-mode. Consequently, the proposed method concerns only the inter-prediction module. It is devoted to three principal levels: CUs level, PUs, and skip mode. Experimental results for different proposed methods showed an important improvement in terms of encoding time which is the most important constraint for any real-time implementation. Moreover, the combination of these three methods could achieve up to 48% of time reduction with insignificant loss in video quality.

DSP-based down-sampling process using lanczos filter bank

Re-sampling process is the technique used to create a new version of the image with different width and/or height in terms of pixels count. It may be either increasing the size of image so called upsampling or reducing its size so called down-sampling. In this paper we interest on down-sampling process. Hence, a DSP TMS320C6472 implementation of the down-sampling module is performed. The down-sampling method is based on poly-phase filter bank. A software design for this poly-phase filter interpolation on two levels is proposed in order to satisfy a significant time saving. Results show the efficiency of the down sampler filter especially for removing the aliasing artifacts. A good speed performance is performed for the two implementation design. For execution time results, we might attaint a gain of 33 % for multi-core design execution compared with the one core design.

Performance analysis of inter-layer prediction for H264/SVC

With the introduction of diverse variety of display transmission and resolutions channel capacities, the joint Video Team has developed a new scalable video coding standard (H264/SVC) as an extension of H.264/AVC. In fact, it provides a single compressed bit-stream with several scalability levels. The spatial scalability is mainly based on coding tools such as inter-layer predictions that offer a better coding efficiency. In this paper, we present the impact of different inter layer prediction modes for spatial scalability. As a first step, analysis results show the significant role of inter-layer prediction process for coding efficiency. As a second step, role of each inter-layer prediction type is investigated. Simulations results demonstrate the significance of inter-layer intra prediction and inter-layer residual prediction when targeting a better coding efficiency.