基于小波分形的图像分割算法

皮鞋内腔的自动测量是当今国内外都没有解决的难题。国外通常采用接触式测量,用传感器与皮鞋内腔接触获得数据。这种测量方法存在精度不高的缺点。重庆大学ICT中心在多年ICT研究的基础上提出了用工业CT来解决长期困扰皮鞋工业的难题。皮鞋内腔CT测量仪获得皮鞋的断面图像以后,由于工业CT通过断层扫描获得的图像通常会产生伪影见图1,它是图像中与被检物的物理参数分布没有对应关系的部分,因此图像质量不是非常理想,要对重建出来的图像进行准确的分割,才能够根据分割后的图像进行测量,进而根据多个断面的

Wavelet Fractal-Based Image Segment Algorithm

The image of shoe leather lumen is not very satisfaction because of technology of CT. The smart imagesegment is the base of getting smart measurement data. An algorithm of image segment based on wavelet and fractalhas been proposed after analyzing the specialty of images. The image is decomposed through wavelet multi-resolutiondecomposition , and the fractal dimension is calculated by the decomposed image. This approach is more satisfied thangeneral method in image segment of shoe leather lumen image by CT. This algorithm can segment the edge of shoe lu-men exactly. The experimentations prove the approach is rational.