Abstract
The morphology of retinal blood vessels is important for the diagnosis of various ophthalmic diseases and physical diseases. Automatic segmentation of retinal blood vessels by using scientific methods is a key step. This paper proposes a new supervised method for vessel segmentation in retinal images. Firstly, the pre-processed retinal image is cropped into an 48×48 image patch, which not only increases the training sample, but also reduces the scale of the neural network. Secondly, a deep fully convolutional neural network is used to extract the depth features of the retinal image. The features extracted by the high-level convolutional layers and the low-level convolutional layers are combine to develop much more accurate segmentation. Finally, the trained model is tested on the dataset DRIVE, the accuracy, sensitivity, specificity and AUC are 0.9512, 0.7256, 0.9856 and 0.9863, respectively. It shows that the proposed method has great values in computer aided diagnosis of ophthalmic diseases.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.