摘要
针对天空背景下低信噪比的飞行器,提出了一种基于SUSAN算法、灰色系统理论和数学形态学相结合的飞行器结构特征提取的新方法。在Visual C++6.0平台下,首先利用SUSAN算法从背景中提取飞行器的结构边缘信息,并与原图像相加实现目标增强;然后用灰色系统理论检测出飞行器的结构特征边缘;最后利用条件膨胀和重构算法,实现云层的抑制,并重构出飞行器目标。实验结果表明:该方法对于实现飞行器的跟踪、结构特征提取以及事后判读有重要的意义,同时验证了该方法的可行性。
An algorithm based on SUSAN algorithm, gray system theory and mathematical morphology for extraction of the structure feature of aircraft with sky background and low SNR image was given. At the platform of Visual C ＋ ＋ 6.0 firstly the structure edge of aircraft was extracted from the background using SUSAN operator, and the target was strengthened with the origin image. Then gray system theory was applied to detect the edge of aircraft. After binafization conditioned inflation was used to restrain cloud and noise and reconstruction the aircraft. The experimental result indicats that the algorithm has very important meaning to realize target tracking, structure feature detecting and interpretation. The feasibility of the method is proved.
出处
《光学技术》
EI
CAS
CSCD
北大核心
2007年第3期406-408,共3页
Optical Technique
基金
中科院二期创新项目(C04708Z)
关键词
成像光学
SUSAN算法
灰色系统理论
特征提取
关联度
条件膨胀
image optics
SUSAN algorithm
gray system theory
feature extraction
relevant degree
conditioned inflation
