Jian Feng

Scene change detection algorithm for MPEG video sequence

A new scene change detection algorithm is developed for MPEG encoded video sequences. The proposed method is very simple that it only requires the bit-rate information at macroblock level as well as the number of various motion-predicted blocks for detection. Those information can be extracted easily from the encoded bit-stream without decompression. Simulation results on different sequences show that the proposed algorithm is effective for detecting abrupt scene changes in MPEG video.

Adaptive block matching motion estimation algorithm using bit-plane matching

Motion estimation/compensation techniques are extensively used in video compression because of their abilities to exploit the high temporal correlation between successive frames of an image sequence. We first propose a new matching criterion called bit-plane matching (BPM) for distance measurement in the block matching algorithm. This criterion not only considers the magnitude but also the directive information of the pixel elements in the block. A simple adaptive block matching algorithm is also developed to adjust the search origin and search range at a block-by-block basis. Simulation results show that the proposed algorithm achieves similar MSE performance for motion-predicted images but requires about 8 times less the number of computations than the full-search method.

Error concealment for MPEG video transmissions

Packet loss is a critical problem for video transmission on packet-switched networks. An error compensation method called the modified boundary matching algorithm is developed for recovering lost video due to packet loss at the video decoder. Simulation results demonstrate that the proposed method preserves good image quality even when the cell loss rate is very high.

Cell loss concealment method for MPEG video in ATM networks

Cell loss is inevitable in ATM networks due to network congestion and appears to be a critical factor affecting the quality of the video signal. For transmitting MPEG video over ATM networks, cell loss causes error propagation in both the spatial and temporal domain, which in turn leads to severe degradation in the image quality. We first introduce a structural method for packing MPEG video bit-stream into ATM cells. With this method, the impact of cell loss can be limited to the lost cells only. A cell loss compensation method called modified boundary matching algorithm is then proposed for transmitting the MPEG video bit-stream in ATM networks. Simulation results demonstrate that the proposed method yields satisfactory results even when the cell loss rate is very high.

Loss recovery techniques for transmission of MPEG video over ATM networks

Cell loss is a critical problem for transmission of video over ATM networks. We first introduce a structural method for packing MPEG video bit-stream into ATM cells. With this method, the impact of cell loss can be limited to the lost cells only. A compensation method called the modified boundary matching algorithm is then proposed for recovery of the lost video information. Simulation results demonstrate that the proposed method yields satisfactory results even when the cell loss rate is very high.

Motion adaptive classified vector quantization for ATM video coding

In this paper, a two-layer motion adaptive classified vector quantizer (MACVQ) is proposed for ATM video coding. The proposed method classifies image blocks into three different motion classes: slow-motion, medium-motion and fast-motion. Specified vector quantizers are then designed to encode different motion blocks. A simple edge-based classified vector quantizer is developed to encode the motion-compensated prediction (MCP) error. Compared with the MPEG-type DCT interframe coder, the proposed MACVQ coder has better MSE performance and less accumulation error between consecutive frames. >

Simulation analysis on statistical multiplexing of MPEG video sources

In this paper, we first investigate the impacts of different interleaving and synchronization schemes on the performance of an ATM multiplexer using real MPEG data. A source scheduling method is then developed to improve the performance of the multiplexer. Experiment results show that the level of burstiness for the aggregated MPEG traffic is reduced and the network performance in terms of cell loss probability and statistical multiplexing gain is greatly enhanced.

Predictive mean search algorithm for vector quantization of images

A simple coding algorithm called the predictive mean search (PMS) algorithm is proposed for vector quantization of images. The proposed PMS algorithm, which exploits the interblock correlation, requires less computation than several existing VQ encoding algorithms and also achieves a higher compression ratio than the standard VQ scheme.<<ETX>>

Two-layer MPEG video coding algorithm for ATM networks

In this paper a two-layer MPEG video coding algorithm is presented for video communications over ATM networks. A new classification method is proposed to partition image data into two layers with different priorities. A study on the statistical characteristics of the output bit-stream such as bit-rate distribution and autocorrelation at both bit-level and cell-level is given in the paper.

Improved minimum distortion encoding algorithm for image vector quantization

An improved minimum distortion encoding method called predictive mean search (PMS) algorithm is proposed for image vector quantization in this paper. With the proposed method, the minimum distortion codeword can be found by searching only a portion of the codebook and its relative address to the origin of search is sent instead of absolute address. Two schemes are proposed to transmit the relative address of the minimum distortion codeword. Simulation results show that a significant reduction in both computations and bit-rates is achieved by using the proposed methods.