基于BP神经网络的复合材料预制件表面纹理特征分析

Texture features analysis of composite preform surface based on BP neural network

讨论了基于神经网络的复合材料预制件的表面纹理分类 ,提出了 BP神经网络在预制件图像边缘增强、特征提取的基础上加感知器而实现纹理有效识别的应用方法 .实验证明 ,该方法具有一定的可靠性 。

The paper discusses the texture classification of the composite surface based on the neural network.It shows the application way will be recognized through image edge enhancing,abstracting image features and then using detection.The experiments prove that it has good steady performance.It will create favourable conditions for composite material mechanical analysis.