On the Use of Motion Vectors for 2D and 3D Error Concealment in H.264/AVC Video

Author(s):  
Hugo R. Marins ◽  
Vania V. Estrela

The fundamental principles of the coding/decoding H.264/AVC standard are introduced emphasizing the role of motion estimation and motion compensation (MC) in error concealment using intra- and inter-frame motion estimates, along with other features such as the integer transform, quantization options, entropy coding possibilities, deblocking filter, among other provisions. Efficient MC is one of the certain reasons for H.264/AVC superior performance compared to its antecedents. The H.264/AVC has selective intra-prediction and optimized inter-prediction methods to reduce temporal and spatial redundancy more efficiently. Motion compensation/prediction using variable block sizes and directional intra-prediction to choose the adequate modes help decide the best coding. Unfortunately, motion treatment is a computationally-demanding component of a video codec. The H.264/AVC standard has solved problems its predecessors faced when it comes to image quality and coding efficiency, but many of its advantages require an increase in computing complexity.

Author(s):  
Hugo R. Marins ◽  
Vania V. Estrela

The fundamental principles of the coding/decoding H.264/AVC standard are introduced emphasizing the role of motion estimation and motion compensation (MC) in error concealment using intra- and inter-frame motion estimates, along with other features such as the integer transform, quantization options, entropy coding possibilities, deblocking filter, among other provisions. Efficient MC is one of the certain reasons for H.264/AVC superior performance compared to its antecedents. The H.264/AVC has selective intra-prediction and optimized inter-prediction methods to reduce temporal and spatial redundancy more efficiently. Motion compensation/prediction using variable block sizes and directional intra-prediction to choose the adequate modes help decide the best coding. Unfortunately, motion treatment is a computationally-demanding component of a video codec. The H.264/AVC standard has solved problems its predecessors faced when it comes to image quality and coding efficiency, but many of its advantages require an increase in computing complexity.


Author(s):  
Minesh Patel ◽  
Anand Darji

Extensive use of digital multimedia has led to the development of advance video processing techniques for development of multimedia applications. Application such as video surveillance requires 247 recording and streaming. So, the bandwidth and storage costs become significant. With introduction of video streaming over internet, where different kinds of end users request same content with different available bandwidth, it requires scalable video coding (SVC). These challenges can be overcome by developing new techniques to reduce redundancy in subsequent frames and to improve the coding efficiency. In this paper, overlapping weighted linear sum (OWLS) pre-processing method and its hardware architecture are proposed. It is implemented using field progrmmable gate array (FPGA) and the application specific integrated circuit (ASIC) is also developed using TSMC180nm technology standard cell library. Results show improvement in terms of power and area as compared to the existing work. In motion compensated temporal filtering (MCTF), wavelet transform is implemented by temporal filters. Architecture for 5/3 Lifting MCTF is also implemented and compared with baseline H.264 video codec. Simulation results show that the average peak signal to noise ratio (PSNR) improvement is 2.36[Formula: see text]dB. The MCTF design using 5/3 Lifting filter is synthesized for Virtex-5 FPGA and compared with the existing close-loop architecture with better performance.


Author(s):  
Hao Zheng ◽  
Yizhe Zhang ◽  
Lin Yang ◽  
Peixian Liang ◽  
Zhuo Zhao ◽  
...  

3D image segmentation plays an important role in biomedical image analysis. Many 2D and 3D deep learning models have achieved state-of-the-art segmentation performance on 3D biomedical image datasets. Yet, 2D and 3D models have their own strengths and weaknesses, and by unifying them together, one may be able to achieve more accurate results. In this paper, we propose a new ensemble learning framework for 3D biomedical image segmentation that combines the merits of 2D and 3D models. First, we develop a fully convolutional network based meta-learner to learn how to improve the results from 2D and 3D models (base-learners). Then, to minimize over-fitting for our sophisticated meta-learner, we devise a new training method that uses the results of the baselearners as multiple versions of “ground truths”. Furthermore, since our new meta-learner training scheme does not depend on manual annotation, it can utilize abundant unlabeled 3D image data to further improve the model. Extensive experiments on two public datasets (the HVSMR 2016 Challenge dataset and the mouse piriform cortex dataset) show that our approach is effective under fully-supervised, semisupervised, and transductive settings, and attains superior performance over state-of-the-art image segmentation methods.


2019 ◽  
Vol 16 (22) ◽  
pp. 20190500-20190500 ◽  
Author(s):  
Prayline Rajabai Christopher ◽  
Sivanantham Sathasivam

Sign in / Sign up

Export Citation Format

Share Document