基于粉尘浓度的图像退化模型  

Image Degraded Model Based on Concentration of Dust

作　　者：王园宇[1] 刘钹 刘杰[2] 张海超[3] 

机构地区：[1]太原理工大学计算机科学与技术学院,山西太原030024 [2]太原理工大学软件工程学院,山西太原030024 [3]太原理工大学电气工程学院,山西太原030024

出　　处：《机械工程与自动化》2015年第5期7-8,11,共3页Mechanical Engineering & Automation

基　　金：山西省自然科学基金资助项目(2012011014-2)

摘　　要：针对基于视觉的目标自动识别与跟踪在粉尘环境中遇到的图像退化问题,提出了一种粉尘浓度对图像退化的影响模型,用于模拟一定粉尘浓度下图像的退化。首先,通过自主设计的粉尘环境模拟装置对粉尘图像进行了采集;其次,建立了粉尘图像退化的初步模型;然后,以浓度为自变量对干扰图像逐像素点进行一元二次非线性回归,求得参数,并进行显著性检验;最后,修正粉尘图像的退化模型。实验证明,本模型在距离一定的情况下可以对给定粉尘浓度的退化图像进行模拟。In order to simulate the degraded problem of dust image,the image degraded model based on concentration of dust was proposed.Firstly,dust images were acquired in a self-made instrument in which dust environment could be simulated quantitatively;Secondly,a rough degraded model of dust was established;Then,the relationship between the concentration of dust and the disturb image was set up by unary quadratic non-linear regression method;Finally,the image degraded model based on the concentration of dust was got.The experiments show that the model is able to simulate the degraded image of dust of certain dust concentration under the assumption of fixed distance.

关 键 词：数字图像处理 图像退化模型 图像模拟 粉尘浓度 

分 类 号：TP317.4[自动化与计算机技术—计算机软件与理论]