摘要
本文分析了实现目标检测并达到自动取景和自动变焦目的的各种方法及算法,构架了一种新型的基于声音检测和图像分析的自动监控系统。通过声音定位,实现系统的初步取景,后经过图像检测,进一步精确范围,同时确定变焦倍数,特别是在图像检测方面,对运动检测和肤色匹配进行了算法研究。实验结果表明,本系统充分利用景物运动特性,突破了摄像机视场角的限制,提高了自动取景的准确度及稳定性,实现了有效的自动变焦。本系统在各种监控场所将会有很好的应用前景.
Different automatic detection methods and algorithms that can be used for automatic viewfinder and zoom are discussed. A new automatic detection system based on sound detection and image analysis is introduced preliminary viewfinder is realized by sound detection and more precise range is obtained by image detection. ~ The Image detection also determines the zoom ratio. Especially, we made lots of researches on motion detection and complexion. The results show that this system breaks the limitation of the camera' s angle of view, improves the accuracy and stability of the automatic viewfinder and achieves automatic zoom. This proposed system will be applied well in various monitoring situations.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2008年第11期2373-2377,共5页
Chinese Journal of Scientific Instrument
关键词
自动取景
运动区域检测
肤色检测
自动变焦
automatic viewfinder
motion detection
complexion detection
automatic zoom