Description of Invention
Coffee is one of the most renown beverage in the world. In the process of brewing a good cup of coffee, it is vital that the quality of coffee remains the upmost importance. There are many types of distinct coffee serve around the world which depends on the type of coffee bean. Such as light, medium and dark roast coffee bean. When roasting the coffee bean some of the coffee beans would be overly roasted or under roast due to the inconsistent heat distribution or some of the beans are defective which needs more heat in order to crack the coffee aroma out of the bean. The degree of roast can be seen from the shade of the coffee bean. There is also a possibility that the beans could be mixed up due to the not thorough cleaning of the roaster before another batch of roasting process is done. This causes the remnants of the previous roast to be mixed with the new set of coffee beans. However, for small and midsize enterprises (SME) there is a lack of implementing automated machines that does analytical methodologies for detection and quantification of roasted bean coffee of different class. The reason causing this is because of the bulky machinery and also the hefty price tag. Thus, in this project, an automated bean sorting system with vision inspection functions to sort light, medium and dark roasted coffee bean, aims to market for SMEs’ to classify the coffee beans without having to buy a big and expensive coffee sorter machine. The bean sorter system consists of two parts which are the mechanical system that sorts the beans and the vision system that analyse the colour of the bean. The coffee bean sorter also dynamically stores detected bean colours and maintaining an index for sorting, the system effectively creates a pipeline where detection, processing, and sorting are continuously happening at different stages for different beans. This improves the efficiency and throughput of the bean sorting system. The mechanical system consists of two motors which are stepper motor and servo motor which are controlled by the Arduino UNO. The vision system consists of a camera module and also PyCharm software program. With the combination of the mechanical and software system, it results in an efficient and accurate sorting system.