Detecting and counting coin using opencv and watershed algorithm

Money is an object that is accepted by the general public as a medium of exchange in economic activities. According to the material, money is divided into two types, coin and bank note. Coins are generally used in small number of transactions, but in some cases used for large transactions. Indonesia as a developing country has very varied goods prices to meet all the needs of various groups, so they often use coins as the main transaction tool, especially for low-priced goods. This creates problems in counting coins in terms of time and energy. This paper will discuss how computer vision can be used to count coins using watershed algorithm. We conduct simulations using images of coins of various sizes. We found the detection of coins of various sizes was successful. In addition, the calculation of the number of coin objects is roughly in accordance with the number of coins in the picture. This finding reinforces that the combination of OpenCV and the watershed algorithm is very suitable for counting the number of objects contained in the image.