An Improved Method of Blood Vessel Enhancement Based on Hessian Matrix

Abstract

The analysis of blood vessel images is of great significance to the diagnosis and treatment of vascular diseases. By performing blood vessel enhancement operations on blood vessel images, it can help clinicians perform diagnoses to reduce workload and time. In the blood vessel enhancement algorithm based on the Hessian matrix, to save time when calculating the Hessian matrix, usually use LoG operation for approximate replacement, resulting in the loss of some details. In this paper, we propose a new calculation method that uses successive small size convolution kernels to convolve the filtering results of the previous scale when calculating the Hessian matrix, thus avoiding excessive calculation time and not losing the details of blood vessel enhancement. We evaluate our method on real image data sets, and the results verify that our method has better performance.