The target audience of the CarbScan application is patients with diabetes, or others who are partaking in diets that restrict their carbohydrate intake. The application allows a user to take a photograph of a plate of food that they wish to eat.

The application then uses image recognition to detect what type of food is contained in the photo. The amount of carbohydrates contained in the plate of food is then calculated using this information. This user can then take the correct amount of insulin for that meal.

The main issue that the application tackles is that of carbohydrate counting, which is a large part of living with type 1 diabetes. Carb counting is the process of calculating the amount of carbohydrates contained within a meal. Diabetics must perform this at every meal in order to know how much insulin they must take for that meal.

The process can be time consuming and an annoyance to people trying to live with diabetes. CarbScan aims to reduce the time of this carb counting process, whilst still allowing its user to accurately estimate the carbs in their meal. Integration with diabetes tracking systems such as the Freestyle Libre is planned for a future development of the project, allowing insights into the changes of blood sugar after meals.