一种基于链码向量的图像匹配算法

基于特征的图像匹配算法被广泛应用于图像处理和模式识别领域中,图像特征提取以及采用的匹配算法并直接决定图像匹配的效果。为了尽可能准确的实现图像匹配,提出了一种基于链码向量的边缘特征匹配算法。首先通过改进的Laplace边缘检测算子提取图像中的边缘信息,提高了边缘检测的可靠性;然后,将提取到的边缘信息由边界链码描述,并将边界链码构造成向量,利用数学向量相似度原则进行图像匹配。实验结果表明,该匹配算法简单快速,匹配准确率高,具有较高的实用价值。

An image matching algorithm based on chain code vector

The image matching algorithm based on feature is popular in image processing and pattern recognition.The image feature extraction and matching algorithm which we select directly determine the effect of the image matching.In order to achieve accurately image matching as far as possible,we have put forward an edge feature matching algorithm which based on chain code vector.First of all,we extract the edge information of image by an improved Laplace edge detection operator which can improve the reliability of edge detection.Secondly,the edge information which has been extracted can be described by chain code.Thirdly,we structure the chain code into a vector and make use of similarity principle of mathematics vector to carry on image matching.Experimental results show that the method is simple and fast,and can gain a high precision of matching.So,the method has a high practical value.