一种基于图像灰度的快速匹配算法





在图像模板匹配问题中,基于像素灰度值的相关算法尽管已经十分普遍,并得到广泛的应用,但目前此类算法都还存在有时间复杂度高、对图像亮度与尺寸变化敏感等缺点.为了克服这些缺点,提出一种新的基于图像灰度值的编码表示方法.这种方法将图像分割为一定大小的方块(称为R-块),计算每个R-块图像的总灰度值,并根据它与相邻R-块灰度值的排序关系进行编码.然后通过各个R-块编码值的比较,实现图像与模板的匹配新算法中各个R-块编码的计算十分简单 匹配过程只要对编码值进行相等比较,而且可以采用快速的比较算法新算法对像素灰度的变化与噪声具有鲁棒

-Based on grayscale images fast matching algorithm in the image template matching, based on the pixel gray value correlation algorithm despite widespread and widely used, but there are still such algorithms have high time complexity, the image brightness and sensitive to changes in size and other shortcomings. to overcome these drawbacks, a new representation of the encoded image gray value based method. this method divides the image into a certain size of the box (called R- blocks), calculated each block R- total gray value images, and sort it according to the relationship between the gray values of the neighboring blocks R- encoded and then by comparing the value of each R- block coding, the new algorithm to realize image matching and template each R- block coding calculation is very simple　the matching process as long as the encoded value equality comparison, and can quickly compare the new algorithm algorithm changes and pixel gray noise robust