Using Deep Learning Model to Identify Iron Chlorosis in Plants

Abstract

Iron deficiency in plants causes iron chlorosis which frequently occurs in soils that are alkaline (pH greater than 7.0) and that contain lime. This deficiency turns affected plant leaves to yellow, or with brown edges in advanced stages. The goal of this research is to use the deep learning model to identify a nutrient deficiency in plant leaves and perform soil analysis to identify the cause of the deficiency. Two pre-trained deep learning models, Single Shot Detector (SSD) MobileNet v2 and EfficientDet D0, are used to complete this task via transfer learning. This research also contrasts the architecture and performance of the models at each stage and freezes the models for future use. Classification accuracy ranged from 93% to 98% for the SSD Mobilenet v2 model. Although this model took less time to process, its accuracy level was lower. While the EfficientDet D0 model required more processing time, it provided very high classification accuracy for the photos, ranging from 87% to 98.4%. These findings lead to the conclusion that both models are useful for real-time classifications, however, the EfficientDet D0 model may perform significantly better.