摘要
利用最大方差法与数学形态学对碳纤维木质复合板材SEM图像进行图像分割,成功获取碳纤维图像;以碳纤维特征数据与其宏观性能指标为基础,应用BP神经网络建立碳纤维木质复合板材均匀化模型。结果表明,碳纤维分割图像清晰、完整;所建立的BP神经网络模型准确、有效。
The carbon fiber wood composite plate SEM image image segmentation by using maximum variance method and mathematical morphology, success for carbon fiber. Take carbon fiber image and characteristic data and its macro performance as the foundation, the application of BP neural network to establish carbon fiber wood composite plate homogenization model. The experimental results showed that carbon fiber segmentation image clear and complete. The established BP neural network model was accurate and effective.
出处
《林业科技》
2013年第2期30-35,共6页
Forestry Science & Technology
基金
黑龙江省自然基金项目"基于有限元的SCFRW功能材料宏微观多尺度模型化分析研究"(C201230)
关键词
碳纤维木质复合材料
最大方差法
数学形态学
图像分割
神经网络建模
morphology
Carbon fiber wood composite materials
Maximum variance method
Mathematical Image segmentation
Neural network modeling