Artificial Intelligence Model Implementation in Web-Based Application for Pineapple Object Counting

摘要

Indonesia's horticulture sector will gain the capacity to address traditional agricultural problems by adopting Agricultural 4.0 technology. As a pillar of Industrial Revolution 4.0, Artificial Intelligence (AI) can provide the benefits of time efficiency and minimize human error compared to traditional methods. The research used the crowd-sourcing method as a well-established approach in data gathering collaboration. Python and Flask also used in the research for web-based application development, Keras-RetinaNet for AI training models, and Web Server Gateway Interface (WSGI) for AI Model Deployment. This research resulted in a web-based AI application that can be used to count pineapple objects on a very wide area by using hundreds of aerial photography as training dataset. The accuracy of pineapple object counting with AI model utilization will optimized the use of resources such as water, fertilizer, insecticides, and packaging materials.