MPEG Video Streams Research Articles

The Algorithm of Scenario Summary Synthesis Based on Semantics

The MPEG video data contains three types of frames, which are I frame, P frame and B frame. However, the I frame records the main information of video data, and the P frame and B frame are just as motion compensation of the I frame. The paper presents the method that analyzes the MPEG video stream in the compressed domain, and then according to the key technique of semantic-based video retrieval the paper presents directly extracting I-frames in the compressed domain, and using the DC coefficient of I frame to establish the model of information system. Then reduce I Frame by the theory of attribute reduction based on RS, and produce a key frame that reflects the main content of the shots. Finally, the key frame’s DC coefficient has constructed the mathematical modeling, and has classified the shot according to the differences of frames. So by the polymerization of I frames of similar shots, we could obtain the scenario summary. Experiment indicated that this method can be realized in the compressed MPEG video automatically, and can lay the foundation for the video processing in the future.

A reliability enhanced video storage architecture in hybrid SLC/MLC NAND flash memory

Abstract Video-based applications have been widely used in embedded mobile devices. In such devices, NAND flash memory chip is normally used as the storage media to store videos. To enlarge the capacity of NAND flash memory, MLC (multi-level cell) technology is widely adopted with low storage reliability compared with the classical SLC (single-level cell) technology. Thus, the inherent features of NAND flash memory will significantly affect the storage reliability of video files. In this paper, we propose a reliability enhanced MPEG video storage architecture (REVSA) in hybrid SLC/MLC storage systems. In this architecture, the video stream is reconstructed as the key cluster and the non-key cluster according to the importance of video restoration. The key cluster is stored in SLC blocks to ensure the reliability, while the non-key cluster is stored in MLC blocks to utilize its capacity. To achieve this, we introduce a new layer in the storage hierarchy called video bit-stream analyzer, which includes a segmentation pipeline and a merging pipeline for MPEG video stream to build clusters and merge them to original video stream. We conducted experiments in a hybrid SLC/MLC storage system to store the car driving videos. The experimental results show that the proposed scheme can significantly improve the storage reliability of the MPEG video files and prolong the lifetime of the storage system.

Simulated performance of TCP, SCTP, DCCP and UDP protocols over 4G network

Forth Generation (4G) network support for wide geographical locations proves its use as a more advanced wireless technology. The Long-Term Evolution (LTE) is a 4G mobile communications standard set by International Communication Union (ITU), specifically ITU Radio communication Sector (ITU-R). At the moment, video traffic and telecommunications grow under the expansion of LTE, which is considered as the actual motivating access technology of 4G network. Throughout the deployment of LTE, various transport protocols are advised and broadly experimented, for instance, TCP, SCTP, DCCP and UDP that may execute differently on 4G networks subject to the network scenarios and parameter settings. Even though the deployment of LTE is swiftly enhanced, there is a lack of performance evaluation of its protocols. Hence, a widespread scrutiny is required for the evaluation of the operation of numerous protocols for high-end applications such as multimedia. Adopting these applications with flexible quality of service constraints with improved usage of resources is a challenging task. In this paper, the output results of different transport protocols for multimedia streaming applications, e.g., video, through extensive simulations are analyzed. The performance of an MPEG-4 video streaming is evaluated using NS-3. The performance metrics used are delay, jitter, throughput, and packet loss. These metrics are evaluated at the base station via TCP, SCTP, DCCP and UDP protocols over the 4G-LTE technology. The obtained results show that the DCCP performs the best in throughput improvement with the minimization of delay and jitter as compared to UDP, TCP, and SCTP.

Limits for the real-time simulation of video services over commodity hardware

The results from event-driven simulations with video services can present limitations because of the cost of developing accurate models. Creating a video server model precise in how streams are transmitted is generally as expensive as the construction of the real server, thus several simplifications are generally introduced to reduce the effort. In this context, the lack of accuracy could be devastating if metrics such as jitter are imprecise. One potential alternative to solve these issues is the combination of real services with Real-Time (R-T) capable network simulators. This paper evaluates R-T simulation capabilities of two of the most popular tools, NS-3 and OPNET to support the performance evaluation of video services using a commodity hardware testbed. Taking a video on demand service as a test case, we have studied the use of NS-3 and OPNET in situations where their more profitable characteristics, such as instance creation simplicity or the use of analytic traffic, could produce the best results. The performed tests show how NS-3 is able to replicate an access network topology with up to 10 MPEG-4 video streams or OPNET has been able to replicate a ring core network with up to 60 of the same streams without affecting important metrics as important as jitter. These results point out the validity of R-T simulation in the context of services that have streams as a main component if their strengths are taken into account.

High-quality streamable free-viewpoint video

We present the first end-to-end solution to create high-quality free-viewpoint video encoded as a compact data stream. Our system records performances using a dense set of RGB and IR video cameras, generates dynamic textured surfaces, and compresses these to a streamable 3D video format. Four technical advances contribute to high fidelity and robustness: multimodal multi-view stereo fusing RGB, IR, and silhouette information; adaptive meshing guided by automatic detection of perceptually salient areas; mesh tracking to create temporally coherent subsequences; and encoding of tracked textured meshes as an MPEG video stream. Quantitative experiments demonstrate geometric accuracy, texture fidelity, and encoding efficiency. We release several datasets with calibrated inputs and processed results to foster future research.

MPEG-4 Video Transmission Over IEEE 802.11e Wireless Mesh Networks Using Dynamic-Cross-Layer Approach

This paper proposes a dynamic cross-layer approach to enhance quality of MPEG-4 video stream transmission over an IEEE 802.11e wireless mesh networks. IEEE 802.11e standard either classifies video data to a specific access category (AC) which causes high packet loss at heavy load conditions or uses static mapping to map video data to lower priority AC which causes unnecessary transmission delays and packet losses in case of light load network. Some developed techniques dynamically map video frames to the other priority queues causing highly degradation in video quality. The paper work proposes a new approach that combines a new mapping cross-layer between the medium access control and the application layer for dynamically mapping video frames based on frame priority and queue load congestion with low priority queue parameters adaptation such as minimum and max contention window to reduce queuing delay. Proposed approach is compared with already developed schemes. Simulation results give high performance under light and heavy loads network traffic. All implementations are simulated using NS2 Simulator tool with required configuration to examine all possible types of traffics.

An Algorithm of Extracting I-Frame in Compressed Video

The MPEG video data includes three types of frames, that is: I-frame, P-frame and B-frame. However, the I-frame records the main information of video data, the P-frame and the B-frame are just regarded as motion compensations of the I-frame. This paper presents the approach which analyzes the MPEG video stream in the compressed domain, and find out the key frame of MPEG video stream by extracting the I-frame. Experiments indicated that this method can be automatically realized in the compressed MPEG video and it will lay the foundation for the video processing in the future.

REVAMPING QUALITY OF SERVICE OF VIDEO STREAMING OVER WIRELESS LAN

Just like video streaming over a wired network, streaming can also be done over a wireless LAN. However, real-time streaming over wireless networks is a challenging proposition due to the highly variable nature of wireless links and the resource-poor nature of mobile devices. In such a context, transmission control schemes have to dynamically adapt both to the application requirements and the channel conditions. In this paper, we put forward a short review of some recent innovations, which have been devised to thoroughly revamp the Quality of Service (QoS) of video streaming over WLANs. Depending on their applications, the solutions have focused mainly on the following three network layers: Media Access Control layer (MAC), Application layer (APP) and Physical layer (PHY). In this paper, we propose an adaptive cross-layer quality-of-service (QoS) scheme for wireless channel and streaming applications. For the sake of adaptive QoS, the cross-layer architecture assumes that layer information could be exchanged between application layer and lower layers. In addition, priority-based adaptive QoS scheduling for MPEG video streams is proposed here. It considers the frame type of the MPEG-4 video file so as to efficiently provide non-similar priorities to important packets of video. IEEE 802.11e protocol assigns top priority to video applications. It does this in order to reduce delay and packets losses, which could happen due to other competing traffic. Simulation results performed with the network simulator ns-2 will show that the cross-layer architecture allows a good performance under both, light and heavy loads, while minimizing the mean packet delay and frame jitter. We aim to minimize the dropping of frames and frame jitter while gracefully degrading video quality to enable the same.

A Burst Effort Broadcasting Approach of MPEG-4 Video Transmission for Intervehicle Communication

Efficient and real-time video broadcasting helps to improve driving safety and to make traveling fun for drivers. However, it is hard to structure vehicles into permanent network topologies and schedules to broadcast video streaming real-time in intervehicle communication (IVC) in a high-mobility vehicular environment. In order to achieve this, we have designed a burst effort broadcasting (BEB) approach for IVC, which considers the challenges of high mobility and multihop broadcast, as well as the features of MPEG-4 video streams to adapt to highway scenarios. The BEB approach is distributed in real time without protocol overhead and comprises a queuing procedure and a scheduling scheme. The queuing procedure consists of the preprocess of video transmission, including the video shaping of groups of pictures and sequential reordering video frames. Based on the queuing procedure, a mobility-adaptive scheduler is applied to handle the broadcast and rebroadcast of the video stream. The concept of macroscopic broadcast is utilized to increase the broadcast performance and the video perceived quality of service (PQoS), and to reduce the number of unnecessary redundant broadcasts. As an evaluation, the real MPEG-4 video was conducted in simulation and the broadcast performance was compared with another protocol by the metrics of peak-signal-to-noise ratio (PSNR) and loss of video frames in different broadcasting scenarios. Then, the results were analyzed. The simulation proved that this approach is an efficient multihop broadcast solution that does indeed provide a realistic solution to promote a higher degree of video PQoS on highways.

Research on MPEG Video Stream Transmission Performance Based on NS-2

Following the deployment of network multimedia applications, such as video conference, simulation research on video stream characteristic is more and more important. MPEG standard encodes the video sequences to frame I, P, and B, and has higher compression rate and bandwidth saving character. MPEG video traffic characteristic and simulation methods are researched in this paper. Network simulation version 2(NS-2) kernels is extended using C/C++ programming language. An MPEG video traffic generator modules and simulation interfaces are designed in NS-2 kernel. Simulation results show that the proposed extension method is feasible and effective.

Online OLED dynamic voltage scaling for video streaming applications on mobile devices

This work proposes an online DVS approach for OLED-based mobile video applications to reduce display power consumption. A time-efficient representative-region based DVS scheme is developed and applied in MPEG video streaming. Based on the proposed scheme, flexible DVS solutions can be adaptively derived according to timing constraints. Experimental results show a 26.9% power reduction while keeping 99.3% frames displayed in high quality.

Fast Channel Zapping with Destination-Oriented Multicast for IP Video Delivery

Channel zapping time is a critical quality of experience (QoE) metric for IP-based video delivery systems such as IPTV. An interesting zapping acceleration scheme based on time-shifted subchannels (TSS) was recently proposed, which can ensure a zapping delay bound as well as maintain the picture quality during zapping. However, the behaviors of the TSS-based scheme have not been fully studied yet. Furthermore, the existing TSS-based implementation adopts the traditional IP multicast, which is not scalable for a large-scale distributed system. Corresponding to such issues, this paper makes contributions in two aspects. First, we resort to theoretical analysis to understand the fundamental properties of the TSS-based service model. We show that there exists an optimal subchannel data rate which minimizes the redundant traffic transmitted over subchannels. Moreover, we reveal a start-up effect, where the existing operation pattern in the TSS-based model could violate the zapping delay bound. With a solution proposed to resolve the start-up effect, we rigorously prove that a zapping delay bound equal to the subchannel time shift is guaranteed by the updated TSS-based model. Second, we propose a destination-oriented-multicast (DOM) assisted zapping acceleration (DAZA) scheme for a scalable TSS-based implementation, where a subscriber can seamlessly migrate from a subchannel to the main channel after zapping without any control message exchange over the network. Moreover, the subchannel selection in DAZA is independent of the zapping request signaling delay, resulting in improved robustness and reduced messaging overhead in a distributed environment. We implement DAZA in ns-2 and multicast an MPEG-4 video stream over a practical network topology. Extensive simulation results are presented to demonstrate the validity of our analysis and DAZA scheme.

Unequal error protection under bitrate constraint for video streaming over internet

This article describes a simple packet-level FEC system suitable for unequal error protection of layered video streams, that we called TAPIOCA (in French, Transport Audiovisuel avec Protection Inegale des Objets et Controle d'Admission). It is designed in a way that the FEC overhead induced by redundant packets is perfectly controlled by the sender. In order to achieve that, TAPIOCA calculates on-the-fly the optimal erasure code to be used, video data unit by video data unit, under a given bitrate constraint. In addition, and contrary to the well-known PET (Priority Encoding Transmission) system, the video data units of each layer are encoded separately. This is especially useful when all layers are not output from the video coder at the same time. Simulation results for MPEG-4 video streaming show that the proposed FEC system can be very efficient even if packet losses are due to network congestion. Moreover, comparison with PET system shows that TAPIOCA exhibits better performance, considering criteria including the decodable frame rate, protection system efficiency and computational cost.

New pixel based approach for reverse play of MPEG video for streaming system

In macroblock based architecture of MPEG video streaming system with backward playback support, the macroblocks of the frames are divided into two categories: Backward Macroblock (BMB) and Forward/Backward Macroblock (FBMB). The BMB and FBMB are processed differently. This approach reduces the network bandwidth and buffer size requirements. In this paper, we propose a pixel based approach that accesses less data from the server and hence further saves the network bandwidth and buffer requirement. The I- or P-frame (say, frame n−1) is reverse-predicted from the currently decoded frame (i.e., frame n) that is stored in the frame buffer at the client system. Using the motion vector information of frame n, the positions of its various pixels are found in the frame n−1 and their exact values are found using the prediction errors of the frame n by just subtracting them from their values in the frame n. Thus, most of the pixels of frame n−1 are predicted from the frame n. Experimental results show that on an average 93.4% of pixels in the previous I- or P-frame can be reverse-predicted from the current P-frame. The rest 6.6% unpredicted pixels are accumulated in the form of blocks and those blocks are requested from the server. We also propose new Block Identification Algorithms to identify different blocks of unpredicted pixels in an image. They are Maximum Height Minimum Width (MHMW), First Come First Serve (FCFS), Maximum Area (MA), and Biggest Block (BB) algorithms. The server processes the MPEG video stream and returns the desired blocks. We also discuss a FindBlock algorithm for extracting a small block of any frame from the MPEG video stream. Thus, only the motion vectors, prediction errors of frame n, and the unpredicted pixels of the frame n−1 need to be transmitted to the client system. This makes considerable saving in system resources.

Improving transmission quality of MPEG video stream by SCTP multi-streaming and differential RED mechanisms

SCTP is an emerging transmission protocol with the advanced advantage of supports multi-homing and multi-streaming, which can provide high availability and increase reliability. To improve the performance of MPEG video over Internet, the present study modifies the multi-streaming mechanisms of SCTP, which enable SCTP to use network resources efficiently and provide a differential transmission priority for the encoding frame types of MPEG video stream. Furthermore, to further enhance the decodable frame ratio, the present study also modifies the queue management mechanism of RED which enable the modified RED to provide differential stream protection for MPEG video stream. The simulation results presented in this study confirmed that the proposed scheme can improve the performance of MPEG video stream.

Data obfuscation with network coding

Network coding techniques such as fountain codes are a promising way to disseminate large bulks of data in a multicast manner over an unreliable medium. In this work we investigate how to conceal such an encoded data stream on its way to numerous receivers with a minimum investment. Compared to conventional ‘encrypt – encode/decode – decrypt’ approaches, our solution is preferable for two reasons: First, it causes less CPU investment for encryption and decryption proportional to the ratio of the payload length to the signaling data length. Second, besides obfuscating the payload data from an eavesdropper, we hide the coding information from an eavesdropper. We evaluate the approach with respect to its application to various data types like MPEG-2 video streams and Java bytecode.

Adaptive FEC algorithm based on prediction of video quality and bandwidth utilization ratio

In this paper, a channel adaptive FEC algorithm is proposed which balances the trade-off between the QoS of video transmission and the bandwidth utilization ratio in wireless IP networks. Our algorithm can dynamically adjust to a suboptimal number of FEC redundant packets to cater to the time-varying wireless channel. For the sake of obtaining the suboptimal amount of FEC redundant packets, we derive two analytical models, one is the playable frame rate in MPEG video stream, another is the effective utilization ratio of FEC. Based on these analytical models, a suboptimal value of redundant packets, which makes both the quality of video stream and the effective utilization ratio of FEC approximate their maximum, is calculated by predicting the quality of video stream and effective utilization ratio of FEC under different network conditions (characterized by packet loss probability in this paper). The simulation results indicate that the QoS of wireless video transmission and the bandwidth utilization ratio are improved by employing the proposed algorithm.

Cross-Layer Measurement on an IEEE 802.11g Wireless Network Supporting MPEG-2 Video Streaming Applications in the Presence of Interference

The performance of wireless local area networks supporting video streaming applications, based on MPEG-2 video codec, in the presence of interference is here dealt with. IEEE 802.11g standard wireless networks, that do not support QoS in according with IEEE 802.11e standard, are, in particular, accounted for and Bluetooth signals, additive white Gaussian noise, and competitive data traffic are considered as sources of interference. The goal is twofold: from one side, experimentally assessing and correlating the values that some performance metrics assume at the same time at different layers of an IEEE 802.11g WLAN delivering video streaming in the presence of in-channel interference; from the other side, deducing helpful and practical hints for designers and technicians, in order to efficiently assess and enhance the performance of an IEEE 802.11g WLAN supporting video streaming in some suitable setup conditions and in the presence of interference. To this purpose, an experimental analysis is planned following a cross-layer measurement approach, and a proper testbed within a semianechoic chamber is used. Valuable results are obtained in terms of signal-to-interference ratio, packet loss ratio, jitter, video quality, and interference data rate; helpful hints for designers and technicians are finally gained.

Extraction of Desired Partial Block of a Frame from MPEG Video Stream

In MPEG video encoding, the next frame is encoded based on the previous frame. For forward replay of the video, extraction of data of the next frame from the current frame is a trivial process. However, for backward replay, it may not always be possible to extract data of the next (previous) frame completely from the currently displayed frame. In this paper, we propose a method for extracting unpredictable portion of a frame from MPEG video. This helps in extracting the desired portion rather than decoding the entire frame.

A semantic framework for video genre classification and event analysis

Semantic video analysis is a key issue in digital video applications, including video retrieval, annotation, and management. Most existing work on semantic video analysis is mainly focused on event detection for specific video genres, while the genre classification is treated as another independent issue. In this paper, we present a semantic framework for weakly supervised video genre classification and event analysis jointly by using probabilistic models for MPEG video streams. Several computable semantic features that can accurately reflect the event attributes are derived. Based on an intensive analysis on the connection between video genres and the contextual relationship among events, as well as the statistical characteristics of dominant event, a hidden Markov model (HMM) and naive Bayesian classifier (NBC) based analysis algorithm is proposed for video genre classification. Another Gaussian mixture model (GMM) is built to detect the contained events using the same semantic features, whilst an event adjustment strategy is proposed according to an analysis on the GMM structure and pre-definition of video events. Subsequently, a special event is recognized based on the detected events by another HMM. The simulative experiments on video genre classification and event analysis using a large number of video data sets demonstrate the promising performance of the proposed framework for semantic video analysis.