基于态势理论的Agent模型及受损图像的恢复  

The Agent Model based on the Ecostate-Ecorole Theory and Restoring the Damaged Picture

作　　者：刘洪[1] 梁栋[1] 何辅云[1] 

机构地区：[1]合肥工业大学计算机与信息学院,合肥230009

出　　处：《计算机工程与应用》2002年第8期82-84,共3页Computer Engineering and Applications

摘　　要：文章提出了一种基于态势理论的Agent计算模型,并将这种模型成功地用于对局部受损图像的恢复。文中所提出的局部受损图像恢复方法是根据受损区域的特点,产生一系列随机Agent,再对它进行繁殖和进化。在不断地进化中分割出受损区域,用受损像素周围的某种均值对其进行恢复。实验证明此方法具有计算量小,快速,高效,自适应的特点。This paper presents an Agent model based on the Ecostate-Ecorole theory,and uses it to restore the local damaged pic ture successfully.The method of restoring local damaged picture produce some random Agents based on the character of the damaged regions then breed and evolve them.In the course of the gradual evolution,it can segment the damaged region,then restore the damaged pixels by replacing them with some sort of average in the neighboring region.The experiment proves this method has the small computation cost,fast,good effect,adaptive characters.

关 键 词：AGENT 态势理论 受损图像 图像恢复 图像分割 

分 类 号：TN911.73[电子电信—通信与信息系统] TP391.41[电子电信—信息与通信工程]