Vessel Image Enhancement Method Using Nonlinear Transformation Based on Guided Filtering
Qi Tang, Wang Wen-Tian, Zhou Ke-xin, Sun Yi-Xuan · 2021
Aiming at the problems of blurred contrast and difficult diagnosis of vascular tissue in medical images, A nonlinear transform blood vessel enhancement method based on guided filtering is proposed. In this paper, a hybrid filter is obtained by fusing guided filtering and Laplacian filtering to complete the preprocessing of image vessel detail enhancement, then the green component is stretched by nonlinear mapping function in RGB color space, the red component and blue component are adjusted according to the stretching degree. Finally, the saturation component is nonlinear contrast stretched in HSV color space to achieve vascular image enhancement. The paper uses objective evaluation methods such as image DV-BV, information entropy and average gradient to compare the performance of existing image enhancement algorithms. The results show that the proposed method has excellent performance.