协同神经网络立体图像识别方法

Heasure of Synergetic Neural Network Stereo Image Recognition

基于前人在协同神经网络的相关研究,本文提出了立体图像的原型向量的构造方法——立体像素平铺法和投影截面法,并给出了投影截面法的优化方法。

Based on previous research in synergetic neural network, the construction methods of threedimensional image＇s prototype vector, three dimensional pixels spread out method and Projection section method are proposed Tn the paper. And the optimization of ProJection section method is given also.