Artificial Intelligence based on Deep Learning in Ophthalmic Imaging

Abstract

The quantitative study of lesion area is of great significance in ophthalmic clinic. Therefore, it is necessary to segment the focus from the ophthalmic image. The existing segmentation algorithms do not make full use of the overall information of ophthalmic images and are not suitable for corneal lesions segmentation. Therefore, this paper proposes a new algorithm based on artificial intelligence and deep learning. Firstly, the improved mean algorithm is used to cluster the chroma and brightness, and then the corneal lesions are segmented by mathematical morphology operation. Experimental results show that the algorithm can effectively segment the lesions in color ophthalmic images.