Glaucoma is an irreversible, chronic eye disease for which there is no effective treatment. It is caused by elevated intraocular pressure that damages the optic nerve. There are no symptoms in the early stages of glaucoma, so early diagnosis is crucial to prevent blindness. In this paper, an accurate, reliable, and rapid glaucoma diagnosis method based on computer-aided detection (CAD) technology is presented. The effective use of CAD can significantly reduce the workload and burden of clinical doctors. The CAD system designed in this paper is successfully deployed on the Android platform and provides a user-friendly interface for clinical use. Experimental results demonstrate that the proposed strategy may precisely segment the optic cup and disc in retinal fundus images. Additionally, this research establishes the viability of using deep learning models on mobile devices to partition the optic cup and disc in fundus images.