Open Source Colour Sensor
Sorting based on colour, shape, and other quality metrics is a persistent problem in agriculture and the food industry. Easy-to-implement technologies are needed to sort fruits and vegetables, with the objective of saving time, reducing waste, and increasing profitability based on sorting according to different quality parameters. The main idea of our work was inspired by the need of the agricultural industries to classify the fresh produce by colour in an inexpensive and efficient way.
Currently there are a number of colour sensors on the market, with prices ranging from low cost RGB chips to sophisticated and very expensive spectrophotometers.
We developed a sensor for precise colour measurements using only off-the-shelf components only, but with the additional advantage of being calibrated against a spectrophotometer. The calibration corrects for the unavoidable nonlinearity of the response of the RGB sensor and enable ones to produce accurate results in the L*a*b* colour scale or equivalents scales.
The colour sensor illuminates a sample using in-built red, green and blue (RGB) LED light sources and records the quantity of light reflected back from the object in the three different colour channels.
The sensor requires the sample to be in direct contact with the object for an optimum reading without interference from external ambient or sunlight.
The sensor comes with a dedicated Android/iOS app, BLE connectivity and WiFi/IoT upgrade option.