基于FBM分形向量特征的图象匹配

IMAGE MATCHING BASED ON FBM FEATURE VECTOR

由于自然景物图象的灰度与边缘特征不稳定，基于灰度与边缘的匹配算法对于这一类图象难以生效。本文通过分析ＦｒａｃｔａｌＢｒｏｗｎｉａｎＭｏｔｉｏｎ（ＦＢＭ）的Ｗｅｉｅｒｓｔｒａｓｓ－Ｍａｎｄｅｌｂｒｏｔ随机分形逼近函数的频谱，给出了分形向量特征的定义和快速判别图象的ＦＢＭ区域的方法。在ＦＢＭ区域内采用该特征进行匹配能克服复杂自然景物图象中灰度与边缘特征的不稳定性。实验表明采用ＦＢＭ分形向量特征的匹配方法能够获得比较传统匹配方法。

Due to the unstable grey level and edge features of digital images on nature scence,matching methods based on grey level and edges of image is not feasible,In theis paper we proposed a fractal feature vector based on FBM model、analysis of a FBM's approximation function Weierstrass Mandelbrot Randon Fractal Function,presented a image matching approch using this vectors in FBM regions and resampling the images Our method can work robustly when the great of changes of image grey levels and features are existed in the natural scences Experiments have showed that our approch is more efficient and has higher matching probability than often used approches based on grey level and edges,such as MAD (Mean Absolute Difference) algorithm