摘要
在对图像特征进行有效分割的基础上,提出了一种基于红外成像障碍物检测的改进Flood Filling算法,并首次将该算法应用到障碍物检测领域中。通过对模拟以及真实红外图像中目标的快速成功提取,不仅验证了该方法的有效性,同时实验结果也表明了其较之原始算法对时效性的提高。
Based on the effective segment of image feature, this paper raised an efficient fast method to detect the possible obstacles by firstly applying Flood Filling algorithm into object detection domain. The successful extraction of the obstacles in both testing stimulant and real infrared images proved not only the validity and reliability of this novel method but also the enhancement of computing speed.
出处
《传感技术学报》
CAS
CSCD
北大核心
2009年第1期85-89,共5页
Chinese Journal of Sensors and Actuators
基金
教育部博士点资助项目(20050699037)
