Font Size: a A A

Optimization And Implementation Of JPEG Encoding Based On Embedded Multi-core Architecture

Posted on:2017-12-14Degree:MasterType:Thesis
Country:ChinaCandidate:X Y LiaoFull Text:PDF
GTID:2348330503493276Subject:Control Science and Engineering
Abstract/Summary:PDF Full Text Request
With the development of embedded technology, derived a range of products such as smartphones, digital cameras,etc which often need to encode the image information, in order to achieve the storage, transport and other purpose. As the mainstream image codec technology, JPEG standard has already widely used in embedded system image processing.To meet the higher throughput and real-time requirements, the traditional method of improving single-core CPU frequency have significant defects in size and power consumption. However, the way through the multi core architecture chip not only greatly improves the performance but also has the characteristics of small size and low power consumption. For this purpose, it is of great significance for the embedded multimedia processing to realize the image encoding algorithm and optimization by multi-core embedded structure.In this paper, we have implemented and optimized the JPEG encoding algorithm based on the embedded structure and the JPEG baseline encoding algorithm. Firstly, the paper introduces the background, research status and development trend of image encoding and embedded multi core architecture. Then, the basic principle of JPEG baseline encoding algorithm is analyzed. In the third part, two kinds of embedded multi-core architecture, the physical structure, the parallel programming model and the evaluation index of parallel computing are compared. It established a general idea of homogeneous multi-core as the hardware platform, Pthread model as the means of programming and the speedup as a parallel evaluation index. The fourth part of the paper gives the overall framework of the parallel encoding, then analyzes the feasibility of the algorithm. The data partition module,JPEG encoding module and sub graph fusion module are described in detail at the sametime.In addition, the JPEG encoding module has a number of code-level and algorithm-level optimization. After testing, using the multi core parallel encoding algorithm and JPEG encoding module encodes a resolution of 24 bit RGB color 1280*720 picture on the iTop-4412 platform only takes 56ms.
Keywords/Search Tags:JPEG encoding, Embedded system, Parallel, Multi core, Optimization
PDF Full Text Request
Related items