摘要
重点研究了旋转图像间的匹配方法,通过改变影像的结构将旋转状态变换为平移状态,计算改动后影像的傅氏变换幅度谱,用幅度谱代替原始灰度进行相关。同时,为了提高速度,采用分层搜索法分两个阶段完成整个匹配过程。同时进行大量的实验,从抗旋转性、匹配速度和正确率等方面进行分析。实验数据表明,该算法可抗任意角度旋转、匹配速度较快、正确率较高,而且对噪声小的尺度变化具有良好的鲁棒性。
This paper focused on the study of rotated image matching method. It transformed the rotated status to translation status by changing the image structure, and then calculated the Fourier transforming magnitude spectrum of the changed image and correlated the magnitude spectral values instead of original gray levels. Meanwhile, for improving the process speed, hierarchy - searching strategy with two steps was applied throughout the matching. Then it carried out many experiments for the proposed method and analyzed its performance of anti - rotation, matching speed, and matching correctness. The experimental results showed that the algorithm can against arbitrary rotation with a rather high speed and matching rate, and it is robust to scale transform with low noise.
出处
《测绘与空间地理信息》
2007年第4期104-107,共4页
Geomatics & Spatial Information Technology
关键词
图像匹配
旋转
圆形模板
傅里叶变换
灰度
image matching
rotation
circle template
Fourier transform
gray level
