摘要
利用傅里叶变换技术将空间域的纹理图像变换到频率域中,采用频谱图对机织物纹理信息进行分析。通过傅里叶变换技术得到机织物图像的频谱图,利用阈值分割曲线选取合适的阈值对频谱图进行分割,然后用邻域最大值的方法找到织物纹理的特征坐标点。根据坐标点与实际频率的关系计算机织物纹理的频率范围。实验结果表明,随着密度的增大织物纹理的频率增加,斜纹和缎纹的频率范围比平纹织物的大,常见机织物的纹理为中低频。
In order to analyze the texture information of woven fabric, the woven fabric image was converted from the spatial domain to the frequency domain using the Fourier transform. The spectrum of the woven fabric was obtained by using the Fourier Transform method. The appropriate threshold was selected from the threshold segmentation curve to segment the spectrum and the feature points containing the fabric texture were found by the maximum of neighborhood method. Based on the relationship between the feature points and the real frequency, the frequency range of woven fabric texture was obtained. The experiment result shows that the frequency of fabric texture increase with the increase of density and the frequency range of twill and satin is larger than the plain weave and the frequency of common woven fabric texture is low and medium.
出处
《纺织学报》
EI
CAS
CSCD
北大核心
2015年第7期55-60,共6页
Journal of Textile Research
基金
国家自然科学基金资助项目(61271006
61379011)
中央高校基本科研业务费专项资金资助(14D110130)
关键词
图像处理
傅里叶变换
织物纹理
频率
image processing
Fourier transform
fabric texture
frequency