Audio Spectrum Processing for Sound-based Nondestructive Methods Used in Fruit Quality Determination

—Despite significant advancements in science and technology, assessing the maturity of certain fruits in large quantities, such as durian, pineapple, watermelon, coconut, and walnut, remains a challenging task. To address this, non-destructive methods are commonly employed to determine fruit maturity. One such approach involves studying the sound obtained from tapping on the fruits' husks. Depending on the presence of ambient noise during knocking, the need for noise filtering becomes apparent. In quiet environments with minimal noise, noise filtering may not be necessary. However, in noisy environments, a noise reduction algorithm becomes crucial to enhance the intrinsic features of the fruit's sound while attenuating external contaminating components. In this research, we propose a noise reduction algorithm for the aforementioned purpose.