Abstract

The DualModal2019 is a dual-modal fundus image dataset. It is for vessel, arteriole, and venule segmentation tasks. The dataset consists of five types of images: RGB color images, the 570 nm and 610 nm monochromic images, and the corresponding manually annotated ground truth images of the arterioles and venules.

Instructions:

The DualModal2019 dataset has 30 sets of images with a resolution of 1024×1024 pixels. Each set of images consists of five images: RGB color image, the 570 nm and 610 nm monochromic images, and the corresponding manually annotated ground truth images of the arterioles and venules. The directory names of the images represent the images type.

Comments

The dual-modal fundus images were acquired using a dual-modal fundus imaging system (OT-110M, Hefei Orbis Biotech LTD., China). The system provides a field of view of 45 degrees using a xenon lamp as the light source; it captures images of 570 nm, 610 nm and RGB channels simultaneously.

Submitted by Mingzhai Sun on Fri, 02/22/2019 - 01:14

thanks

Submitted by Priyank Saxena on Thu, 12/23/2021 - 02:32

hard to download ...

Submitted by Lei Wang on Fri, 01/20/2023 - 01:53

thanks

Submitted by Li Yang on Sat, 05/04/2024 - 06:51

Dataset Files

The dual-modal fundus image dataset for vessel, arteriole and venule segmentation tasks DualModal2019.zip (50.44 MB)

Datasets

Standard Dataset

DualModal2019 dataset

Abstract

Comments

Dataset Files

QUESTIONS?