An Implementation of YOLOv5 and Flutter Framework for Fruit and Vegetable Object Detection

Fruits and vegetables are recognized as highly important sources of vitamins and minerals for humans. When the body lacks vitamins and minerals, the immune system weakens. To maintain health and immunity, a variety of nutrients found in fruits and vegetables are required. According to the World Health Organization (WHO), it is recommended to consume at least 400 grams or five servings of fruits and vegetables every day to obtain sufficient dietary fiber and reduce the risk of non-communicable diseases. However, many people still do not adhere to these recommendations. Based on a survey of 100 respondents, only 19% consume fruits every day, and only 44% consume vegetables every day. The aim of this research is to develop a mobile application using machine learning technology and the YOLOv5 architecture that can recognize fruits and vegetables, as well as provide information and education about the benefits and nutritional content of fruits and vegetables. This is intended to raise awareness among the public about the importance of consuming fruits and vegetables. The development method used for the application is the Mobile Application Development Lifecycle (MADLC), which involves stages of identification, design, development, prototyping, testing, deployment, and maintenance. The results of the confusion matrix testing show a precision of 0.9, an average recall of 0.92, and an average F1-Score of 0.9, indicating that the application is able to identify fruits and vegetables effectively.