摘要
Contourlet变换具有各向异性和方向选择性特征,使其在特征提取和图像识别中拥有天然的优势。介绍了Contourlet变换的原理和构造方法,重点讨论了基于Contourlet变换的图像特征提取,并把此方法应用于无人机可着陆的草地地貌图像的识别。经仿真测试,基于Contourlet变换的草地地貌识别精度可达到93%。
Contourlet transform has the characteristics of anisotropy and directional selectivity,which make it owns the borned advantage in image feature extraction and recognition.This paper introduces the transformation principle and construction method of contourlet transform,emphasizes on the image feature extraction by contourlet transform.,and applies this method in the grassland image identification for UAV landing.The simulation experiment result demonstrates that the highest identification accuracy of grassland landforms based on contourlet transformation is upto 93%.
出处
《沈阳航空工业学院学报》
2009年第5期31-34,共4页
Journal of Shenyang Institute of Aeronautical Engineering
