基于边缘模型的图像后处理算法的FPGA实现  

The FPGA Realization for Digital Image Post-processing Algorithms of Model Edge-based

作　　者：许全泉[1] 黑菲宾[1] 朱红[1] 

机构地区：[1]西安电子科技大学模式识别与智能研究所,西安710071

出　　处：《遥测遥控》2007年第2期22-26,52,共6页Journal of Telemetry,Tracking and Command

摘　　要：在传统图像后处理算法中,性能和复杂度往往难以兼顾。从FPGA硬件实现出发,在原有的基于边缘模型的数字图像后处理算法基础上,提出一种新的基于模型匹配的重建模型算法,并且根据边缘的宽度和边缘退化的主要形式建立两个边缘重建模型。仿真实验结果表明,改进的算法不仅满足工程实时性要求,而且一定程度上抑制了高倍压缩后的数字图像中出现的振铃效应和模糊效应,起到了改善图像主观和客观质量的作用。In the traditional digital image post-processing algorithms,performance and complexity always can not be satisfied simulta- neously.In this paper on the basis of FPGA a new model for construction based model-matching is presented in the original digital image post-processing algorithms of model-based edge,and two construction models-based edge are set up according to both the width of edge and the main form of edge degeneration.The results of simulation and experiment indicate that the improued algorithms can not only process re- al-time,but also improve the quality of the highly compressed digital image by limiting the ringing and blurring effect.

关 键 词：图像后处理 FPGA 边缘 振铃 模糊 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]