In this paper, the problem of detecting blood vessels in retinal images with early proliferative retinopathy is faced by highlighting vessels both in retina background and in the optic disc. For this purpose, a Combined Method for preprocessing fundus oculi images is developed. In detail, each retinal vessel image is enhanced via a contrast-limited adaptive histogram equalization after applying a suitable operator for feature extraction. Then, the proposed Combined Preprocessing Method is synthesized to modify each contrast-enhanced retinal image using both a Two-dimensional Matched filtering and a 2D Gabor Wavelet Transformation for vessel highlighting. Combination and segmentation of preprocessed images are subsequently performed and binary maps of retinal vessels are finally derived. The effectiveness of the proposed method is evaluated on obtained outcomes and results are compared to those obtained with available techniques.
A Combined Preprocessing Method for Retinal Vessel Detection Towards Proliferative Diabetic Retinopathy Screening / Carnimeo, Leonarda; Altomare, Annamaria Roberta; Nitti, Rosamaria (COMMUNICATIONS IN COMPUTER AND INFORMATION SCIENCE). - In: Advances in Artificial Life, Evolutionary Computation and Systems Chemistry: 10th Italian Workshop, WIVACE 2015, Bari, Italy, September 22-25, 2015. Revised Selected Papers / [a cura di] Federico Rossi, Fabio Mavelli, Pasquale Stano, Danilo Caivano. - STAMPA. - Cham, CH : Springer, 2016. - ISBN 978-3-319-32694-8. - pp. 106-116 [10.1007/978-3-319-32695-5_10]
A Combined Preprocessing Method for Retinal Vessel Detection Towards Proliferative Diabetic Retinopathy Screening
Carnimeo, Leonarda;
2016-01-01
Abstract
In this paper, the problem of detecting blood vessels in retinal images with early proliferative retinopathy is faced by highlighting vessels both in retina background and in the optic disc. For this purpose, a Combined Method for preprocessing fundus oculi images is developed. In detail, each retinal vessel image is enhanced via a contrast-limited adaptive histogram equalization after applying a suitable operator for feature extraction. Then, the proposed Combined Preprocessing Method is synthesized to modify each contrast-enhanced retinal image using both a Two-dimensional Matched filtering and a 2D Gabor Wavelet Transformation for vessel highlighting. Combination and segmentation of preprocessed images are subsequently performed and binary maps of retinal vessels are finally derived. The effectiveness of the proposed method is evaluated on obtained outcomes and results are compared to those obtained with available techniques.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
