Fruit Freshness Grading Method Based on GhostNet

As living standards rise, people are placing greater emphasis on food quality and safety. Fruits are easy to rot and deteriorate in the process of preservation and transportation, which directly affects the nutritional value of fruits and the health and safety of consumers.Achieving precise grading of fruit freshness is of great significance for ensuring fruit quality and optimizing market circulation. However, traditional methods for detecting fruit freshness have limitations in terms of efficiency and accuracy. Based on this, this paper proposes an intelligent fruit freshness grading method based on the GhostNet lightweight convolutional neural network. This method fully utilizes the advantages of GhostNet as a lightweight network, which reduces the computational load through the Ghost module. Without compromising essential accuracy, it can effectively reduce model parameters and computational costs, making it highly suitable fruit freshness detection that may need to be run on mobile or embedded devices. The experimental results show that, on the premise of maintaining only 3.92M of model parameters, the accuracy rate of freshness grading for five types of fruits such as apples and bananas by this method reaches 97.487%.