<![CDATA[Scales are needed by traders, including vegetable traders, but the scales created in the market can only determine the weight. That way traders need time to calculate the price based on the weight and type of vegetables. In previous research there has been research on smart scales that can calculate the total price based on the weight and type of vegetables being weighed, this study used the Raspberry Pi 3 Model B and the Convolutional Neural Network (CNN) as a method for the scales to be able to identify the types of vegetables that are on it. Along with the rapid development of technology, the price of the Raspberry Pi for all variants has increased in price. Therefore the need for research on smart scales with components that have relatively cheaper prices. In this study, researchers used the ESP32-CAM microcontroller, which is priced relatively cheaper than the Raspberry Pi 3 Model B. This research still uses the Convolutional Neural Network (CNN) method and a load cell equipped with the HX711 module as a sensor to obtain the weight value of an object. The dataset collected totaled 600 image data with 150 image data for each type of vegetable, classes in the training data consisted of tomatoes, cabbage, carrots, and potatoes. Smart scales using the ESP32-CAM get results of a classification accuracy of 90% and the average difference of the tools built is 0.8 grams compared to the SF-400 brand digital scales.]]>
Copyrights © 2023
