论文部分内容阅读
人工神经网络的研究与应用正引起国际学术界的广泛关注.本文从计算机视觉的角度出发,以姿态测定和不变性识别为应用背景,对人工神经元网络的技术与应用进行了有益的搜索和研究. 首先,本文详细地研究和分析了最典型的人工神经网络——BP网络,对其学习算法、结构设计及其参数影响等问题给出了实验结果和理论说明,并在此基础上提出了BP网络的一般设计原则.
The research and application of artificial neural network are aroused widespread concern in the international academic community.In this paper, from the perspective of computer vision, attitude determination and invariance identification as the application background, the technology and application of artificial neural network have been useful search and Research.Firstly, this paper studies and analyzes in detail the most typical artificial neural network - BP network, gives its experimental results and theoretical explanations for its learning algorithm, structural design and its parameters, and on this basis, puts forward The general design principles of BP network.