基于PCA和分块FFT的快速模板匹配算法

Fast template matching algorithms based on PCA and blocking FFT

图像匹配算法作为一种经典的图像识别算法,在计算机底层视觉处理中占有重要地位。在理论界,对该算法的讨论和研究由来已久,也提出了基于像素灰度、基于图像特征等的匹配算法。本文所提出的匹配算法主要基于模板特征,通过结合主分量分析(PCA)、分块快速傅里叶变换(FFT)等技术实现了高效快速的图像匹配算法。在对具有大量相似目标的图像进行检测的应用中,该算法取得了良好的效果。

As one of the classic image recognition algorithms, image matching algorithm plays an important role in computer low-level vision technology. In theoretical field, discussions and studies of the algorithm have a long history, and some algorithms based on image pixel gray level or image features have been proposed. The matching algorithm proposed in this paper is mainly based on template features and by using the combination of technologies including Principal Component Analysis（PCA） and blocking Fast Fourier Transform（FFT） to implement fast algorithm. In the application of detection of large amount of similar objects in an image, this algorithm has achieved good results.