工业环境中运动目标检测的模糊自适应高斯背景模型方法

刘新海; 方康玲; 吴谨

doi:10.11834/jig.200608189

学术论文与技术报告 | 浏览量 : 0 下载量: 190 CSCD: 0

PDF
导出
分享
收藏
专辑

工业环境中运动目标检测的模糊自适应高斯背景模型方法
A Fuzzy Adaptive Gaussian Model Method for Motion Object Detection in Industrial Field
2006年11卷第8期页码：1114
纸质出版：2006
DOI： 10.11834/jig.200608189
稿件说明：

移动端阅览

刘新海, 方康玲, 吴谨. 工业环境中运动目标检测的模糊自适应高斯背景模型方法[J]. 中国图象图形学报, 2006,11(8):1114. DOI： 10.11834/jig.200608189.

A Fuzzy Adaptive Gaussian Model Method for Motion Object Detection in Industrial Field[J]. Journal of Image and Graphics, 2006, 11(8): 1114. DOI： 10.11834/jig.200608189.

摘要

根据工业环境中成像环境差、图像模糊的特点，提出了一种对运动目标检测的模糊自适应高斯背景模型。首先建立3种主要背景状态下的模糊高斯背景模型；为了减少在线检测过程中的运算量，选择背景图像中一些独立、相对静止区域代替背景图像；最后用模糊识别的方法选择出合适的高斯背景模型。该算法不仅能够自适应环境的变化对背景图像的影响，而且满足实际生产过程中在线检测的要求。

Abstract

A fuzzy adaptive Gaussian model is presented in this paper. The method which is based on the fuzzy characteristic of image is employed to detect the moving object in the industrial field. Firstly three fuzzy Gaussian models under three main background states are set up. Then some independent

relatively still regions in the image are chosen to replace the whole background image in order to reduce the heavy calculation in the on-line detecting. At last

the suitable Fuzzy Gaussian background model matching current background is determined by fuzzy recognition. The method can not only be adaptive to environmental transformations but also satisfy the requirement of real-time detecting in the industrial field.