Mean Shift 这个概念最早是由Fukunaga等人[1]于1975年在一篇关于概率密度梯度函数的估计中提出来的,其最初含义正如其名,就是偏移的均值向量,在这里Mean Shift是一个名词,它指代的是一个向量,但随着Mean Shift理论的发展,Mean Shift的含义也发生了变化,如果我们说Mean Shift算法,一般是指一个迭代的步骤,即先算出当前点的偏移均值,移动该点到其偏移均值,然后以此为新的起始点,继续移动,直到满足一定的条件结束.-Mean Shift concept first introduced by Fukunaga et al. [1] in 1975 in an article about the probability density function of the gradient estimate put forward, its original meaning, as its name, is the shift of the mean vector, where is the Mean Shift a noun, it refers to the generation of a vector, but with the development of the theory Mean Shift, Mean Shift of meaning has changed, if we say that Mean Shift algorithm, generally refers to an iterative step, which is to calculate the current point offset mean, move the shift points to its mean, and then as a new starting point, continue to move until the end to meet certain conditions.