摘要
为降低海运事故发生率,提高海上交通安全,运用VC++和计算机视觉处理技术OpenCV,结合肤色的YCbCr模型开发船用值班视频监控的核心算法.该算法以人脸识别为核心,用OpenCV结合肤色的YCbCr模型定位和标识值班人员的位置,根据其精神状态判断是否警示.该算法精度高、运算速度快,且系统的实现成本低.
To reduce the ratio of marine accidents and improve the safety of maritime traffic,a ship watch-keeping surveillance algorithm based on VC++ OpenCV was accomplished.The face detecting is the core of algorithm which uses OpenCV skin color model in YCbCr space to locate the position of the operator's face and then estimates the active state so as to decide whether the alarm is triggered or not.The algorithm has high accuracy,fast computation speed and its implement cost is low.
出处
《大连海事大学学报》
CAS
CSCD
北大核心
2010年第3期57-60,共4页
Journal of Dalian Maritime University
基金
国家自然科学基金资助项目(50979009)
教育部博士点基金资助项目(200801510002)
国家重点基础研究发展计划(973)子项目(2009CB320805)