Font Size: a A A

Dsp-based Image Coding And Realization

Posted on:2006-11-17Degree:MasterType:Thesis
Country:ChinaCandidate:Y L SuFull Text:PDF
GTID:2208360152999720Subject:Detection Technology and Automation
Abstract/Summary:PDF Full Text Request
This paper studies the network transporting and implementation of image compression with ICETEK-C6711-IDK-YUV image process platform based on TMS320C6711 chip, by realizing remote video traffic monitoring system, using JPEG criterion as a base. Namely transporting image data compressed to RCM3200 module by UART, terminal visits image compressed through IE browser.Firstly this paper introduces significance of image coding, theory of image coding, international criterion of image coding, then introduces the features of c6000s DSP and environment of software developing. Realizing JPEG compression encoder in C language on PC after studying JPEG criterion. But the purpose is to realize image coding by hardware, so the program modified must be transplanted to DSP. Mostly file operation are modified to memory operation of DSP.The source code designed on PC is inefficient for DSP platform, so code optimizing is necessary. Based on the standard procedure and supported by powerful development tools and image lib of C6000, the paper partly optimizes the code as follows: a) adopting the compiling options to optimize program; b) using image lib supported by TI; c) assigning the memory sector; d) using DMA to transmit the date between inter and outer memories; e) modifying the functions partly; f) using the key words volatile and far. The optimize program perform 10 times faster then the original. Time needed to achieve a compression ruduce to 0.3s from 3s.Image data Compressed by DSP transfer to the net controller implemented by RCM3200, which provides net service based on HTTP protocol. Then the user can browses these images by terminal PC with the internet browse software like IE.Images with different compression ratio are gained by modify the Q (quality) parameter during the experiment applied on this image process platform. Results of the experiment imply that decline of the image quality is unconspicuous when the Q parameter ranges from 100 to 30. When the Q parameter is set to 25, the image quality drops distinctly. Then a conclusion is drew that Q parameter should be set about 50 to obtain both...
Keywords/Search Tags:Video monitoring, DSP, JPEG, Optimize, Network
PDF Full Text Request
Related items