YUAN Hui-jiuan, SUN Qiao, YU Xiao-yang. LVQ Neural Network Apply in Three Dimensional Object Recognition[J]. 2011, 16(5): 97-100.DOI:
LVQ神经网络在三维物体识别中的应用
摘要
三维物体识别已经广泛应用于工业、医疗及军事等领域
是近年来的研究热点之一.传统的识别方法是根据二维图像来识别三维物体.但摄像机从不同角度拍摄三维物体
引起三维物体的图像变化
使图像特征发生改变
这会严重影响三维物体识别的准确性.针对这个问题
本文采用矩不变量作为图像的特征向量
并且应用LVQ神经网络进行训练
实现识别三维物体.结果表明
这种方法能够从不同角度的图像识别三维物体.
Abstract
3D object recognition has become one research focus recently
and widely used in industrial
medical and military fields.The traditional identification method is based on 2D images to identify the 3D objects.But the images and the images’ characteristics will change
when the camera obtains images from different angles of 3D object.So the accuracy of 3D object recognition would be seriously affected.To resolve this problem
3D object recognition is realized by using moment invariants as characterist...