Low-memory video compression architecture using strip-based processing for implementation in wireless multimedia sensor networks

Abstract

This paper presents a very low-memory video compression architecture for implementation in a wireless multimedia sensor network. The approach employs a strip-based processing technique where a group of image sequences is partitioned into strips, and each strip is encoded separately. A new one-dimensional, memory-addressing method is proposed to store the wavelet coefficients at predetermined locations in the strip buffer for ease of coding. To further reduce the memory requirements, the video-coding scheme uses a modified set-partitioning in hierarchical trees algorithm to give a high compression performance. The proposed work is implemented using a soft-core microprocessor-based approach. Simulation tests conducted have verified that even though the proposed video compression architecture using strip-based processing requires a much less complex hardware implementation and its efficient memory organisation uses a lesser amount of embedded memory for processing and buffering, it can still achieve a very good compression performance

Adaptive scanning methodology for object‐oriented visual information processing

“…As the trend of multimedia processing moves towards a higher resolution, better quality, lower bit rate and lower complexity, and coding efficiency have become the top priorities. Recent research work has emphasised the visual coding components such as compression theory [4,5], object-detection and tracking [6,7] as well as region-of-interest coding [8], whereas less attention has been placed on the pre-processing of the visual data.In this Letter, an investigation that has been carried out on various types of image data capture sequences is presented. The performance of image coding using the traditional raster scan approach and the diagonal snake scan approach is compared to the proposed work where the image pixels are collected starting from the centre of the image area and progressively extended outwards, which is similar to the water drop ripple effect.…”

“…As the trend of multimedia processing moves towards a higher resolution, better quality, lower bit rate and lower complexity, and coding efficiency have become the top priorities. Recent research work has emphasised the visual coding components such as compression theory [4,5], object-detection and tracking [6,7] as well as region-of-interest coding [8], whereas less attention has been placed on the pre-processing of the visual data.…”

“…It is also shown from the simulation results that the proposed scanning approach works particularly well for object-oriented images with a smoother background such as Lenna, Airplane, Zelda and video conferencing sequences such as Claire, Miss-America and Mother-Daughter. Object-oriented visual compression: Several research works on improving the coding efficiency of a visual information processing system have recently been published [3][4][5][6][7][8]. In [3] and [4], a strip-based image compression scheme based on the zerotree coding is presented.…”

“…Object-oriented visual compression: Several research works on improving the coding efficiency of a visual information processing system have recently been published [3][4][5][6][7][8]. In [3] and [4], a strip-based image compression scheme based on the zerotree coding is presented. One of the advantages of the zerotree coding is that the zerotrees with more significant data are given priority in the encoding process.…”

“…One of the advantages of the zerotree coding is that the zerotrees with more significant data are given priority in the encoding process. Here, the zerotree-based image compression performance presented in [3,4] is evaluated using the different types of scanning methodologies. In this simulation, image pixels are input to the compression system using the raster scan, the diagonal snake scan and the proposed scanning methods.…”