Automated glaucoma screening system using visual field based on virtual reality and fundus image based on deep learning
Principal Investigator: Assoc. Prof. Dr. Chaturong Tantibundhit, Thammasat University
Glaucoma is the second leading cause of vision loss after cataract disease. Due to the shortage of ophthalmologists in Thailand, an automated glaucoma screening will help lessen the burden of ophthalmologists and allow patients to get access to diagnosis and treatment in a timely manner, thus preventing vision loss. This project proposes to develop an automated screening system for glaucoma powered by artificial intelligence. The proposed system employs a virtual reality device to screen for visual field abnormalities and uses deep learning to detect retinal abnormalities from retinal images. This screening system has a potential to save THB 14 million on imported technology and the cost of screening.
Development of a Bangkok Cooperative Store Management System
Principal Investigator: Asst. Prof. Dr. Pongsarun Bonyopakorn, King Mongkut’s University of Technology North Bangkok
Bangkok Cooperative Store operates five store branches and has more than 140,000 members. Advancement in digital technology and change in lifestyle have driven customers to engage in online shopping. To stay competitive, Bangkok Cooperative Store plans to open an online shop. This project aims to develop Bangkok Cooperative Store Management System to enable Bangkok Cooperative Store to increase sales volume from an online channel while avoiding the duplicate work within the cooperative and enhancing overall operational efficiency. Digital tools and services will be introduced to the operation, such as using cloud system as a data center and offering new payment methods such as QR code and mobile banking to increase customers’ convenience.
Development of a deep learning system for smart robot bin-picking
Principal Investigator: Assoc. Prof. Dr. Siridech Boonsang, King Mongkut’s Institute of Technology Ladkrabang
Bin picking robots combines industrial robots with 3D cameras, thus allowing them to operate with 3D vision and accurately pick-up objects. Thanks to this capability, bin picking robots are increasingly used in manufacturing to improve productivity, reduce human errors and replace human in labor-intensive or dangerous tasks. This project aims to develop a deep learning system to enhance object recognition and position estimation capability of bin picking robots. The smart bin picking robot developed from this project will be put into use and reduce the dependence on foreign technology.
