Embedded Color Segregation using Arduino

Abstract

This research study proposes an embedded color segregation system using the multi-rate sensor data and color identification. There are plenty of applications for color segregation. The most prominent uses are for waste management and fruit and veg packing. In waste management, clutter is identified based on its size, shape and color. Sensor enabled color segregation helps to segregate the unwanted items with ease of use. Another application is segregating the available fruit and veg from the agricultural produce. One of best approaches to achieve this is based on its color, which is the most economical and fast method. The idea of this color segregation is to extend the work further to develop an autonomous waste management system. The proposed prototype segregates color category based on sensor measurements collected from RGB sensor, TCS34725. Segregating color is very simple to human eyes, but there are a lot of background tasks for a sensor to detect the actual given color. TCS34725 detects the color and makes human life easier by providing the exact RGB values, which cannot be identified by the naked eye. A test methodology has been followed to validate the proposed segregation approach. To perform this, a real time prototype has been developed and measured around 10k samples under different conditions. Results indicate that the proposed approach has achieved significant benefits, i.e., accuracy is above 95%, response time less than 3ms.