摘要
总结了当前图像目标识别算法的现状;从生物医学角度出发,分析了人类对图像信息的获取原理,提出一种基于仿生思维的图像亮度自适应调整算法;从人的视觉系统机制出发,实现一种从各种复杂的图像背景中分辨出同类目标的视觉仿生分类算法。实验结果表明,对于针对各种复杂的、具有多目标的图像,与其它目标分类算法比较,视觉仿生算法的分类效果更好。
Bionic technology exhibits great advantages in the application of image processing. Image target classification is the prerequisite of image processing algorithm. Firstly a survey of the status quo of image target recognition algorithm is offered; then, with a thorough analysis of human beings＇ image information acquisition mechanism based on bio-medical perspectives, an adaptive brightness adjustment algorithm of bionic thinking is proposed. Setting forth from the mechanism of human vision system, a visual bionic algorithm is realized, which can distinguish similar targets from the background of a complicated image. Experimental results prove that the algorithms show more accurate recognition effects when identifying targets against complicated images with multiple targets.
出处
《计算机工程与设计》
CSCD
北大核心
2014年第5期1722-1726,共5页
Computer Engineering and Design
基金
华侨大学引进人才科研启动费基金项目(12Y0316)
泉州市科研基金项目(24201305)
中央高校基本科研业务费基金项目(JB-ZR1202)
物联网云计算平台建设基金项目(2013H2002)
关键词
图像处理
视觉仿生
目标分类
亮度调整
仿生思维
image processing
vision bionic
target classification
brightness adjustment
bionic thinking
