Effective Detection Method for Fruit Recognition with Deep Learning
This paper portrays a methodology of making a framework recognizing foods grown from the ground in the retail advertise utilizing pictures caught with a camcorder connected to the framework. The framework causes the clients to mark wanted products of the fruit with a value as per its weight. The reason for the framework is to limit the quantity of human resource, speed up the distinguishing process and improve the ease of use of the graphical UI contrasted with existing manual frameworks. The equipment of the framework is established by a Tensor Flow Lite, camera, trained dataset. Faster R-CNN (Region based Convolutional Neural Network) and YOLO ( You Only Look Once) are used to improve the efficiency of identifying object with better recognition system for reducing computational complexity. Here we proposed our recognition system with better accuracy and faster performance for fruit recognition.
Keyword: Faster R-CNN, YOLO, Tensor Flow Lite, Convolutional Neural Network, Fruit Recognition.