Main Article Content
In vision tasks like image retrieval, colour is one of the most essential properties used to characterise images. The process of recognizing the name of any colour is known as colour detection. The way we perceive and understand things is influenced by colour. It's much easier to be fed values without having to go through the trouble of locating someone who understands colours. This study proposes that a computer be taught to recognise and define colours accurately enough to be helpful. The polite way to extracting the concerns using CNN algorithms is presented in this study. The foundations of computer vision are used to track three different colours: red, green, and blue. Each colour value in a computer is defined as a number between 0 and 255. We use a dataset that includes RGB values along with their names. If we want to increase the precision of existing colour detection algorithms based on convolutional neural networks the properties of the attention mechanism are used to model colour priors in this paper. In order to increase the accuracy of the system, we also look at several colour detection models and compare them. The research also looks at how colour detection is used in real-world applications like object detection.