摘要
介绍了一种利用图像形态学来处理棉花异纤图像的方法。首先在HSV模型下,采用全局阈值分割和边缘阈值分割相结合的方法来提取异纤,极大地提高了异纤的识别精度。然后再用图像形态学中的全方位腐蚀与膨胀对分割后的图像进行处理,使得异纤图像连续清晰,最后对异纤定位并予以清除。
A method for processing color image of cotton foreign fiber based on image morphology is presented. Firstly, global threshold segmentation combined with marginal threshold segmentation is adopted to extract foreign fiber from natural cotton under HSV （Hue, Saturation, Value） model, which improve the accuracy of identification greatly. Secondly, in order to make a clear image, the all-round expansion and corrosion is implemented to process the image. Finally, positioning and removing of the foreign fiber is realized.
出处
《机械科学与技术》
CSCD
北大核心
2009年第1期121-123,共3页
Mechanical Science and Technology for Aerospace Engineering
基金
河南大学科技攻关项目(05YBGG003)资助
关键词
HSV模型
图像形态学
腐蚀
膨胀
HSV model
image morphology
image processing
cotton foreign fiber
