ABSTRACT
Retinal vessel segmentation is a prime utensil in ophthalmology. From the last two decades, a massive number of studies have been carried out in the creation and development of automated vessel segmentation. A benefit of this procedure has been recognised in various retinal oddities. A keen mathematical tool that anticipates the degree of difficulty of disease is the fractal dimension. Typically, the fractal dimension depends on the complexity of the objects. In the case of vessel segmentation, we placed filters and contrast adjustments for obtaining a better configuration of the original images. As well as, precise visualisation, identification, prevention medication and surgery planning for particular diseases can be done in an effective manner by considering this segmentation procedure. This will enhance the feature of the skeletonized/segmented structure of the retinal vessels. Hence, focusing on the segmentation method is more important. As a consequence, determining an elite strategy for vessel segmentation remains a challenging task. Thus, this article mainly focuses on the segmentation procedures in four ways and it uses the local connected fractal dimension (for short, ). Also, a precise method for segmenting the source images has been fixed in such a way that manipulating the fractal dimension of the output images occurred in those four processes. Ultimately, it results in the best technique by focusing on the variation of the fractal dimension of the raw pictures, which yields that the U-Net segmentation is more effective.
Acknowledgement
The work of the authors is supported by UGC - SAP, Department of Mathematics, The Gandhigram Rural Institute (Deemed to be University) Gandhigram, Dindigul District, Tamil Nadu, India. Pincode: 624 302.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Code availability
MATLAB codes are available.
Data availability statement
STARE data and Matlab outputs.