山羊绒和细支绵羊毛纤维的图像识别技术

Research on Image Recognition Technique for Cashmere and Fine Wool

提出一种对山羊绒和细支绵羊毛的SEM图像进行数字化处理并识别的方法。该方法由扫描电子显微镜成像,通过图像采集卡把实时图像传至计算机,经过灰度变换等图像预处理后,用最大类间方差法求得阈值,采用阈值分割法对图像分割处理,检测边界提取其特征值,然后对羊绒和细支绵羊毛进行分析和识别。

This paper described a kind of digital processing and automatic recognizing method for cashmere and fine wool images pattern. The fiber image (including cashmere and fine wool fiber) is magnified by scan electronic microscope to produce a digital image .The image taken by a camera is transported to a computer and displayed on the screen by an image-acquiring adapter. The image is processed according to the color value of their pixels. After the image enhancement, the two-valued image is obtained by the ideal threshold method. Detecting the edges of cashmere and fine wool, then the characteristic values are extracted. The experiments show that the proposed algorithm is efficient.