Pith Estimation on Tree Log End Images

摘要

In this paper, we present an algorithm for pith estimation from digital images of wood cross-sections. The method is based on a probabilistic approach, namely ant colony optimization (ACO). After introducing the approach, we describe the implementation and the reproduction of the method linking to an online demonstration. Results show that the approach performs as well as state-of-the-art methods. The estimated pith is below 5mm from the ground truth. It is a fast method that could be used in real-time environment. This paper also gives the details about the intern parameter choice and shows how to use the C++ source code for testing, as well as provides limit cases of the proposed method and future improvements.