Computer Vision-Based Smart Agriculture Storage with Quality and Quantity Analysis and Recipe Suggestion

A low-cost Computer Vision-based crop yield classifier for the crop stored inside the container is proposed with this experimentation. This system includes options for analyzing quantity and quality. In quantity analyses the crops such as wheat and rice are classified using a camera and computer vision algorithms and the data is saved in the firebase cloud. In quality analysis, the quality of fruits and vegetables is assessed using the TensorFlow object detection API, and the results are stored in the cloud alongside recipe ideas. Intelli-container also offers a function called monitoring mode for security purposes, in which content inside the system is periodically examined and the user is notified via a web application if there is any theft or missing objects. The web app was designed using HTML and Bootstrap. It displays the real-time updates and suggests a recipe based on the vegetables and eatables present inside the container with a help of the Computer Vision approach. The proposed system contains raspberry pi as the main unit and peripheral sensors like loadcell, HX711 module, and camera module. The system uses TensorFlow modules for classification and object detection using python. With this paper, we are proposing a new term for our implemented system as Intelli-Container (Intelligent +Container). This system is useful for machine learning-based smart agricultural purposes for quality, quantity, and security. As our system is capable of quality and quantity analysis, Also, our proposed system is useful for paying the minimum Support Price (MSP) directly to farmers without intervention middlemen, Thus, this paper has social application in good governance. Another application of our prototype is for giving recommendations for recipes using food in this Inteli_container.