An automatic method of tree-rings boundary detection on wood micro-images

The research of tree-rings helps us to understand the local climate change, pollution, etc. in that time, but also to identify the wood species. The detection of tree-rings boundary is a prerequisite for the computer to automatically obtain information about tree-rings. A new approach based on mathematical morphology to achieve tree-rings boundary from micro-images of wood on cross-section is presented in this paper. This method finds the regions of tree-rings, and isolates the tree-rings from the complex background firstly; eliminates burrs by image processing, and fills breakpoints. Then a new and efficient search algorithm, called PFS, is used to track the tree-rings boundary. The experimental results show that the method is efficient for tree-rings boundary detection on cross-section images, especially can get satisfied outcomes for the condition of many cracks around the tree-rings.