Publication

Abstract : This paper presents a colour detection application that utilises a deep learning model to predict the name of a colour based on its HSV value. The use of the HSV values rather than the RGB values makes for a more intuitive representation of colours. The application supports two modes of operation, real-time mode and static mode, which cater to the needs of visually impaired individuals and content designers in the varying scenarios that they may find themselves in. This paper presents a design that leverages the advantages of deep learning and the stability of the HSV colour scheme to provide a model with good accuracy in detecting colours and labelling them accordingly. The application’s integration with the Flutter Framework and Firebase database provides a scalable and robust platform for both managing the gathered data and collecting more relevant data for the improving the model further.