In this article, by presenting a new description of the characteristic of the central axis points of tubular structures, a method to strengthen these structures is proposed. In this method, in a multi-scale framework and using special vectors of the Hessian matrix of th
More
In this article, by presenting a new description of the characteristic of the central axis points of tubular structures, a method to strengthen these structures is proposed. In this method, in a multi-scale framework and using special vectors of the Hessian matrix of the image points, we obtain the distance of each point from the edges of the image. For the points located on the central axis, this distance from the bisector of any arbitrary direction is symmetrical. In this step, by sampling the distance of each point from the edges of the image in different directions, we assign a greater value to the points that have more symmetry. In the next step, we use a filter based on the Pock method to strengthen the central axis of the tubes. The evaluation of the proposed method has been done using two-dimensional and three-dimensional phantom images and medical data qualitatively and quantitatively with the criteria of maximum error in determining the central axis and detection rate, which shows the advantage of this method over the existing methods.
Manuscript profile
