CNN-based Plant Disease Identification in Crops from Multilabel Images using Contextual Regularization

Main Article Content

R. Srinivasan, C. Santhanakrishnan, S. Iniyan, R. Subash, Pradeep Sudhakaran


Image processing is a technique of enhancing or extracting several information from an image using various algorithms like contrast enhancement, resizing, color conversion, etc. Due to the high performance in recent years, Deep Learning (DL) algorithms have grown more popular for image categorization. The purpose of this research work to identify the healthy leaf and unhealthy leaf from the Multilabel image dataset. The next step is to identify the type of disease affect in the unhealthy leaf. The dataset used in the proposed work comprises of 13 crops leaf images that includes apple, blueberry, cherry, tomato, grapes, orange, peach, pepper, potato, raspberry, soybean, squash, strawberry, and corn. This research work classifies the plant leaf into health or unhealthy using Convolutional Neural Network (CNN) technique. CNN proves that 98.75% of accuracy obtained to the plant village dataset.

Article Details