摘要
针对目前织物组织的识别主要依赖人工而导致效率低下的问题,研究并设计了一种基于Tamura方向度纹理特征的织物组织识别算法。该算法通过将Tamura纹理特征中的方向度、形态学运算以及高斯模糊有机结合,成功生成经纬组织点分布图以实现织物组织识别。通过计算织物组织图的Tamura方向度特征值提取了纹理信息,利用纹理信息结合数字图像处理技术获得组织点分布。实验证明,针对具有较强纹理信息的织物组织,特别是经纬纱线颜色相近的织物,该算法能够很好地实现识别效果。
Nowadays identification of fabric relies mainly on artificial, which leads to the problem of low efficiency. Recognition algorithm of woven fabric was studied and designed based on directionality of Tamura texture features. Through the combination of the directionality which belongs to six attributes of Tamura texture features, morphological expansion operation and method of connected domain area, the algorithm successfully generated distribution diagram of interlacing points. Texture information was extracted by calculating the feature value of Tamura directionality in the fabric structure chart. Distribution of interlacing points could be obtained by using the texture information combined with digital image processing technology. Experiments showed that this algorithm was able to achieve great recognition results in the fabric structure chart with strong texture information, especially the fabric structure chart with similar color of warp and weft yarn.
作者
梅军
张森林
樊臻
MEI Jun ZHANG Senlin FAN Zhen(College of Electrical Engineering, Zhejiang University, Hangzhou 310027, China)
出处
《轻工机械》
CAS
2017年第4期52-55,共4页
Light Industry Machinery
基金
浙江省科技计划项目(2015C31089)
关键词
织物组织
Tamura纹理特征
方向度
膨胀运算
连通域
fabric structure
Tamura texture feature
directionality
expansion algorithm
connected domain
