摘要
在皮革、纺织,食品、冶金和农林牧产品加工等行业中,大背景中微小缺陷的检测大量存在。本文模仿人类的视觉注意机制,提出了一种基于注视机制的机器视觉检测系统及算法。系统由多个低分辨率、低成本的摄像头获取不规则的图像,用主成分分析法对原始样本数据提取特征,然后由BP神经网络对特征进行分类识别以确定可疑区域位置,再控制从动摄像头获取目标区域的细节图像,解决了传统机器视觉系统固有的图像冗余数据问题。
There are large requirements of inspecting small defects in a large background in leather manufacturing, fibers industry, food industry, metallurgical industry, agriculture, forestry and animal husbandry. By simulating human＇visual attention function, this experment proposes a machine vision system with an algorithm, which consists of cameras with low resolution and low cost for acquisition of irregular images. The system uses principal component analysis and BP neural networks to locate the suspectable target area, and then drives the related camera to capture the detail image around the target. In addition, it can solve the problem of redundant image data in the traditional machine vision system.
出处
《皮革科学与工程》
CAS
2008年第6期67-70,共4页
Leather Science and Engineering
基金
国家自然科学基金(60875022)
关键词
机器视觉
注意机制
主元分析
BP神经网络
machine vision
attention mechanism
principal components analysis
BP neural network
