Real-time Dragonfruit's Ripeness Classification System with Edge Computing Based on Convolution Neural Network

In recent years, planted area and production of dragonfruits in Taiwan increased three times in the past decade. As the improvement of agricultural research, it also develops lots of brand-new species. However, grading on dragonfruits still needs to improve. Our research use the appearance of dragonfruit to predict its ripeness by using Convolution Neural Network(CNN) model, which helps to reduce labor and time costs and increase the profit for farmers. In addition, we also develop a real-time ripeness classification system, which combine our research with fruit gravity classifier in the dragonfruit farm. First, we record dragonfruits' videos by using IP camera for data collection. Then, use object detection methods and data preprocessing to process our target frames which would be trained by CNN model. During the prediction stage, use IP camera to capture image and put it into real-time ripeness classification system for edge computing. Then, transmit the predicted result to fruit gravity classifier for grading. Our model predict ripeness of dragonfruits to three classes. Overall accuracy of our model is 94.1%, and f1-score of each class is over than 90%.