Multilayer Perceptron Approach to Non-Rigid Image Matching

摘要：

The problem of the nonlinear variational image matching is solved by approximating the displacement field via multilayer perception network. The unknown weights are found by minimizing the summed square error between the target and template images and also their corresponding gradients. As a smoothness constraint we consider general form of the non-isotropic elastic deformation potential and cost functions introduced in variational grid generation. An example is presented where we solve Static Navier equations imposing Bayesian Navier prior in template based recognition of the bitmaps of handwritten characters.

作者: Ram(u)nas Girdziu(s)as Jorma Laaksonen

作者单位: Laboratory of Computer and Information Science Helsinki University of Technology P.O.Box 5400, FIN-02015 HUT, Espoo, Finland

会议类型: 国际会议

会议名称: 8th International Conference on Neural Information Processing(ICONIP 2001)(第八届国际神经信息处理大会)

会议地点: 上海

会议语种:英文

页码: 529-534

在线出版日期: 2001-11-14（万方平台首次上网日期，不代表论文的发表时间）