Classification of Agriculture Crops Using Transfer Learning

Abstract

Deep learning is a relatively new, cutting-edge method of image processing and data analysis with a lot of potential. Deep learning has lately entered the field of agriculture as a result of its success in other fields. In this study, we do an assessment of various research projects that apply deep learning techniques, applied to various agricultural and food production difficulties. We look at the specific agricultural issues being studied, the models and frameworks used, sources, types, and pre-processing of the data used, as well as the overall success attained according to the metrics employed at each study effort. In addition, we investigate the performance differences between different deep learning techniques and classifiers in classification. Our results show that deep learning gives great accuracy, exceeding previous extensively used image processing approaches. The result obtained was from VGG19 that is 98.5% with Logistic regression classifier.