摘要
采用带有光照自适应的基于高斯混合模型的背景减法提取移动物体的图像.通过只更新感兴趣的高斯分布减少运算量;对基于改进的高斯混合模型算法提取的图像,当光照快速变化时进行光照补偿.实验结果表明,改进的高斯混合模型算法,对光照快速变化的场景有较好的适应性,提取的图像效果较好.
It presented a light adaptive background subtraction algorithms that based on Gaussian mixture model for detecting moving objects.By only updating interesting Gaussian distributions reduced the computation and then performed a illumination compensation for the background obtained from the model.The algorithm achieved a better adaptability in situation such as light mutation,also a better result and efficiency under some conditions compared with traditional Gaussian mixture model.
出处
《湖北大学学报（自然科学版）》
CAS
2012年第3期355-359,共5页
Journal of Hubei University：Natural Science
基金
国家自然科学基金(60973034)资助
关键词
高斯混合模型
光照突变
背景减法
感兴趣的高斯分布
背景的高斯分布
Gaussian mixture model
light mutant
background subtraction
Gaussian distribution of interested
Gaussian distribution of background
