Watermelon Ripeness Detector Using Signal Processing

Abstract

The gifts from nature always help those who are suffering from the sweltering heat and glaring sunlight. Watermelon is one of the summer's most wanted fruits, but we fail to judge the ripeness level. The present work aims at categorizing the state of ripeness of watermelons using recorded tapping sounds and photographed visuals. This prevents farmers from picking immature fruit. By manually hitting the watermelon and recording the sound, the sound file dataset is produced. In the case of image processing technology, a digital camera is used to capture the textures on the watermelon's exterior layers. These images have been augmented. The data gathered will assist in assessing the watermelon's ripeness. The experiments demonstrate acoustic signal processing and image processing techniques. The watermelon datasets have been divided into ripe and unripe categories with greater accuracy of 89 percent out of 336 audio samples and 93 percent out of 4864 image samples respectively.