DEVELOPMENT OF TOMATO SORTING PROTOTYPE BASED ON ITS WEIGHT, COLOR, AND DIAMETER

• Main Author: Rahayuningtyas, Ari
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/39520
• Institution: Institut Teknologi Bandung

Tomato is a type of horticulture plant that has many benefits and also a multipurpose commodity. Post-harvest handling of tomato, particularly sorting, has an important role in determining the quality of tomatoes. Farmers are currently sorting manually so that the results are not the same weight, color, and diameter, it also time consuming. To overcome this problem, this study implements a weight sensor and image processing for tomato sorter device, so that it can detect the weight, the color, and the diameter of tomatoes automatically. This study uses a load cell and a strain gauge with 0.05% level of accuracy to measure the weight. Image processing is used for detecting the maturity level of tomatoes based on its color and diameter. The sorting target tomato weight between 50-100 grams, based on national standard of fresh tomatoes (SNI 01.3162.1992). The maturity level of tomatoes was divided into three groups based on its color which are green (raw), turning or yellowish (medium maturity), and red (mature). The sorting target color group is yellowish, based on farmers’ testimony that tomatoes will change color since being picked, sorted, to shipped to consumers. The target diameter of the tomato was chosen between 40-60 mm. Image processing used to detect the color and diameter of the tomato are RGB thresholding, RGB value comparison, and contour detection. System evaluation was done by evaluating the weight criteria as well as color and diameter criteria of the sample. The weight criteria tests were carried out three times, using 45 tomato samples weighting around 40-110 grams, showing an error of 7%, 4% and 7% for each test. The evaluation of color and diameter criteria were also done three times using 45 samples of tomatoes with various colors and diameter ranging from 30-70 mm. The evaluation of color and diameter criteria shows an error of 4%, 7% and 7% for every test conducted.