Real Time Riped Fruit Detection using Faster R-CNN Deep Neural Network Models

Computer Vision and its associated emerging technology prove high potential in advanced agricultural applications. Recently Deep Learning algorithms are much more efficient in producing solution to Computer Vision Problems. In this paper, a literature review on deep learning algorithms and its uses in agriculture is done. Also, this paper reports the research work on autonomous harvesting system using IOT Technology. Fruit detection and plucking mechanism using Deep Learning Faster RCNN method is proposed. The accuracy rate of fruit recognition using VCG16, ResNet50 and InceptionV2 architectures are discussed. The mean average precision of fruit is 0.869. At the end, several problems related to recognition and localization are addressed.