摘要
为科学地利用工业CT扫描的沥青混合料断面图像,通过比较图像分割学科的阈值法及模式识别领域的混合高斯模型及模糊C均值聚类三种方法,采取一系列二分类的策略较好地将二维切片图像区分为几类物质,并实现了自动批处理大量图片的过程,最后采用马歇尔试件进行了验证。
For the scientific use of industa4al CT sectional images of the asphalt mixture, the threshold segmentation and pattern recognition methods of Gaussian mixture model and fuzzy C means clustering were compared, a series of 2 - classificatin strategy were adopted to divided the image into different substances, and a large number of images automatically batch process were implemented , and finally the different methods were Marshall specimens.
出处
《广东公路交通》
2010年第3期5-8,14,共5页
Guangdong Highway Communications
关键词
道路工程
沥青混合料
混合高斯模型
road engineering
asphalt mixture
gaussian -mixture model compared by the