Prototype of Tomato Fruit Sorting Based on Color Based on Esp32-Cam with Fuzzy Logic Method

Abstract

The demand for fruits, especially tomatoes, continues to increase due to their widespread use in cooking, sauces, and juices. Before distribution, tomatoes must be sorted by weight, size, and color. The current sorting process is still done manually, making it less efficient and accurate. To overcome this, this study developed a prototype of an automatic sorting system using ESP32-CAM, HCSR04 distance sensor, and TCS34725 color sensor with fuzzy logic method. The HCSR04 sensor showed an average distance measurement error of 1.8%, indicating high accuracy in measuring tomato size. Color analysis with the TCS34725 sensor revealed different RGB value dominance patterns according to the ripeness of the tomatoes: green tomatoes are dominated by green values (G), yellow tomatoes are dominated by red values (R) that are close to green values, and ripe tomatoes have the most dominant red values compared to green and blue (B). With the application of fuzzy logic, the system is able to classify tomatoes into three categories: good, medium, and bad. Out of 12 trials, the system successfully classified correctly 11 times, resulting in an accuracy of 91.67% and an error rate of 8.33%. These results indicate that the fuzzy logic-based sorting system is effective in recognizing and classifying tomatoes based on color and size, although there is still room for further improvement to improve accuracy and reliability.