摘要
根据室外环境的特点,提出一种基于朴素贝叶斯光照模式分类和导航线检测相结合的方法,以解决光照变化对导航线提取的影响。首先运用朴素贝叶斯对不同光照条件下的图像进行分类,然后根据图像中导航线区域颜色特征信息,结合边缘特征信息来检测导航线。经实验测试,在阴影遮挡、光照变化等情况下能够准确地提取出导航线,正确率达90%以上,该方法具有较强的鲁棒性。
According to features of outdoor environment, a kind of method based on na？ vc Bayesian illumination mode classification and navigation line detection is proposed to solve the problem of influence on navigation line extraction by change of illumination. Firstly, ha？ ve Bayesian is used for classifying images under different illumination conditions, and then ac- cording to color feature information in navigation line area and combining with edge feature information, detection on navi- gation line is conducted. Experimental testing proves that it is able to correctly extract navigation line under conditions of shading and change of illumination and accuracy is over 90% which means that this method has strong robustness.
出处
《广东电力》
2015年第12期13-18,共6页
Guangdong Electric Power
基金
国家自然科学基金资助项目(51405286)
上海市自然科学基金资助项目(13ZR1417800)
上海市科委重点科技攻关计划项目(14110500700)
上海市电站自动化技术重点实验室(13DZ2273800)
