Space Apps Challenge 2018

NextGAR Team has built a medication pills recognition system which serve to enrich elderly and sick people’s lives and provide more insight in making decisions. the team used Matlab Software to build an algorithm that recognises medication pills. The team started with five medications, and now during the challenge they are working in expanding the dataset up to 12 undisgustible medications. This project is funded by KAUST and TAQADDAM funding grant (a joint award between KAUST and SABB Bank).

Goal:

The goal of this project is the ability to project on the smart devices display automatically generated results of detecting (i.e. localizing) and recognizing (i.e. labeling) objects in front of the user. The product creates device-agnostic software, which acquires and transfers still images, analyses them at the central processing unit, and conveys to the user the meta-data on the wearable/smart device display.

Customers:

• There is a potential customer which is The Saudi Food and Drug Authority. The team had a meeting with the Executive Vice-President for Medical Devices and Equipment at the Saudi Food and Drug Authority, Dr.Nazeeh Al-Othmany. As a result, NextGAR team has an agreement with the SFDA that they are willing to consider our product to be one of their services.
• E-Health care
• Hospitals and elderly centers
• Sensory impaired
• Dispensaries and pharmacies
• Insurance companies
• Clinical Studies and Research
• Pharmaceutical Companies

Callenges:

•The medications have similar edges/shapes, size, and colors

•The large size of the images which may slow down the system

•The difference between the ground truth and the actual image

Resources:

• Image and video understanding lab ( IVUL) at King Abdullah University of Science and Technology (KAUST).
• C. Lawrence, P. Dolĺar, “Edge Boxes: Locating Object Proposals from Edges”, Microsoft Research, 2015.