Fast Marker-based Algorithm for Exact Euclidean Distance Transformation

Euclidean distance transform is widely used in many applications of image analysis and pattern recognition. Traditional algorithms are time-consuming and complicated to realize. This paper proposes a fast marker-based distance transform algorithm. Firstly, mark the background pixels’ positions relative to the object pixels; secondly, scan the object area and define the first object pixel as the parent pixel, and calculate its distance transform information; then, according to the parent pixel’s information, the distance transform information of the object pixels (son pixels) which are 4-adjacent to the parent pixel is calculated successively; and finally, iteratively choose each son pixel as a parent pixel and manipulate its son pixels until all the object pixels have been resolved. Our algorithm has linear time complexity and is easy to implement. Experiments show that comparing to the existing boundary striping and contour tracking algorithm, our algorithm demonstrates a significant improvement in time and space complexity.