Fiber Image Analgsis. Part III: A New Segmentation Algorithm for Autonomous Separation of Fiber Cross-sections

摘要

One of the challenging tasks in fiber image analysis is to separate touchin cross-sections of fibers before extracting fiber geometrical features. This paper reports a skeleton-based segmentation algorithm that deals with this porblem. A skeleton of touching objects in an image is the meidal axis of the objects, adn can be obtained with a sequential erosion procedure. A skeleton reflects the overall longitudinal shape of an object. When the pixels on a skeleton are weighted by the distances to the object edge, the skeleton also contains information about the transverse shape of the object. The pixels where the distances are minimal indicate possible connecting points of two adjacent objecs. A line perpendicular to the skelecton can be drawn across the touching objects at an identified connecting pixel to separate the objects. Inappropriate separations caused by small convcave noises on the edge can be avoided by using pre-specified thresholds when checking the skeleton distance, and the time-efficiency can be greatly improved by eliminating iterative thickening operations.