Iterators （迭代器）are specialized(专门) classes that are built to go over（遍历） each element of a collection（集合的每个元素）, This application of the information-hiding principle makes scanning a collection easier and safer.

　　An iterator object for a cv::Mat instance(实例) can be obtained by first creating a cv::MatIterator_ object. As is the case with cv::Mat_, the underscore(下划线) indicates that this is a template subclass(模板子类). since(因为) image iterators are used to access the image elements, the return type(类型) must be known at the time of compilation(编译). The iterator is then declared as follows:

　　cv::MatIterator_<cv::Vec3b> it;

　　Alternatively(另外), you can also use the iterator type defined inside the Mat_ template class as follows:

　　cv::Mat_<cv::Vec3b>::iterator it;

　　Consequently(现在), our color reduction function is now written as follows:

//Scanning an image with iteratorsvoid colorReduce4(cv::Mat &image, int div = 64) {    // obtain iterator at initial position    cv::Mat_
     
      ::iterator it = image.begin
      
       ();    // obtain end position    cv::Mat_
       
        ::iterator itend = image.end
        
         ();    // loop over(遍历)all pixels    for (; it != itend; ++it) {        (*it)[0] = (*it)[0] / div * div + div / 2;        (*it)[1] = (*it)[1] / div * div + div / 2;        (*it)[2] = (*it)[2] / div * div + div / 2;    }}//Remember that the iterator here returns a cv::Vec3b instance because we are processing a color image.
        
       
      
     

　　用begin方法，在开始位置（本例中为图像的左上角）初始化迭代器。对于cv::Mat实例，可以使用image.begin<cv::Vec3b>()。还可以在迭代器上使用数学计算，例如若要从图像的第二行开始，可以用image.begin<cv::Vec3b>()+image.cols初始化cv::Mat迭代器。可以用运算符++来移动到下一个元素，也可以指定更大的步幅。例如用it+=10，对每10个像素处理一次。