摘要
针对机载光电传感器获得的图像信息的不确定性与模糊性,提出利用D-S证据理论对多图像传感器信息进行融合处理的方法。首先对各图像传感器获得的同一场景图像分别进行预处理,然后对预处理后的图像提取其矩特征,通过与目标图像数据库中图像特征的对比分析,利用灰色系统理论方法获取各传感器的基本概率赋值,最后根据D-S证据理论的组合规则与决策规则获得多图像传感器的融合识别结果。仿真结果表明该方法能提高目标识别的准确度,降低传感器获取的信息的不确定性。
Aim at the uncertainty and faintness of images information which are gained with aerial Optical- Electro Sensors, the image fusion method based on D-S evidence theory is brought forward. Firstly, the pretreatment of each image got by corresponding image sensor is carried. Afterwards, the matrix character of the image was gained and compared with every possible image in image database. The basic probability evaluated of every sensors is gained by grey system theory. Lastly, the target recognition result of multi-sensors obtained through the combination rules and decision rules of D-S evidence theory. The simulation results showed that the methods could greatly improve the efficiency of target recognition and decrease the uncertainty of sensors information.
出处
《传感技术学报》
CAS
CSCD
北大核心
2009年第7期973-977,共5页
Chinese Journal of Sensors and Actuators
关键词
图像融合
光电传感器
D-S证据理论
目标识别
Image Fusion
Optical Electronic Sensors
D-S Evidence Theory
Target Recognition
