Disease identification in grapevine leaf images using fuzzy-PNN

Reliable and accurate identification of disease is required for protecting the plant from pathogens and obviating the yield loss. The advent of computer vision and image processing techniques has encouraged contribution in disease identification systems in plants. This paper proposes a fuzzy feature extraction technique and Probabilistic Neural Network (PNN) for the identification of grapevine diseases using leaf images. The color features are extracted using fuzzy color histogram. Then, the extracted features are fed to a PNN classifier for grapevine disease classification. The proposed technique achieves a maximum recognition accuracy of 95.54% on the test dataset. On comparing the proposed system with upcoming deep learning techniques, the former is found to be more efficient for small training data.