纺织学报 ›› 2011, Vol. 32 ›› Issue (4): 52-56.

• 纺织工程 • 上一篇    下一篇

基于嗅觉神经网络的织物组织识别

包晓敏,曹作宝,汪亚明,周砚江,朱寒宇   

  1. 浙江理工大学
  • 收稿日期:2010-05-31 修回日期:2010-12-09 出版日期:2011-04-15 发布日期:2011-04-15
  • 通讯作者: 曹作宝

Study of fabric pattern recognition based on olfactory neural network

  • Received:2010-05-31 Revised:2010-12-09 Online:2011-04-15 Published:2011-04-15

摘要: 提出了一种基于嗅觉神经网络的织物组织识别的新方法。在生物嗅觉神经网络建模技术的基础上,根据激励响应关系来调节KIII模型输入电压。然后将采集的织物图像在x和y方向投影,计算其组织点大小,然后重排组织点,将重排结果运用上述经过参数调节的KIII模型识别。样本数据验证表明:所建的模型对平纹组织,缎纹组织识别率都高于未调节参数的KIII模型,而对斜纹组织在采用新的组织点提取模型后,识别率也有明显的提高。

Abstract: According to the stimulus-response relationship regulate the input voltage of the KIII model, based on the biological olfactory neural network model technology. And then project the collected fabric image in the direction of x and y, calculated the size of the fabric point, and then rearrangement fabric point. Images are trained by the neural network after processing above-mentioned. The experiment results show that the improved network is effective, and the method can accurately extract the fabric points and it is better than the traditional methods. A new method to extract the fabric points is presented for the twill stitch, the improved network also is effective.

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!