在采用一个鲁棒性匹配误差映射函数的基础上，使用一种类牛顿法的迭代优化方法实现了基于仿射模型的图像目标跟踪算法。
With the use of a robust error mapping function, an affine model-based target tracking algorithm is formulated, in which a quasi-Newton's iterative method is implemented in the optimization progress.
为了克服这一困难，本文通过引进一个仿射变换矩阵，构造仿射投影既约预条件共轭梯度路径来搜索获得迭代方向。
In order to overcome the difficulties, we introduce an affine scaling matrix, and form the affine scaling reduced preconditional conjugate gradient path to obtain a new accepted step.
目前迭代函数系统中的变换多限于仿射变换，而仿射迭代函数系统对于区域的任意四边形不规则剖分情形是不能实现的。
At present, the transform in IFS is limit to affine transform. However, affine IFS is unable to do anything with the anomalous dissection on discretionary quadrangle area.
应用推荐