Preeti
Banana exports in India during 2020–2021 exceeded INR 616 crores, with an increase in banana plants leading to higher demand. The focus is on cultivating disease-free banana plants, with authors emphasizing early detection and prediction of banana leaf diseases. A process involving image acquisition, pre-processing, segmentation, feature extraction, feature selection, and classification is outlined.
Image acquisition involves capturing high-quality images for effective feature extraction. Pre-processing includes resizing, formatting, and noise removal to prepare images for segmentation. The segmentation layer selects significant components of the image, making subsequent tasks simpler. Feature extraction involves studying pixel numbers and capturing information to differentiate between healthy and affected banana leaf plants.
Various methods like HOG features, Hu moments, Haralick texture, and color histograms are used for feature extraction. The feature selection layer reduces input variables, improving model performance. Classification is done using Support Vector Machines (SVM), Convolutional Neural Networks (CNN), and Artificial Neural Networks (ANN) to classify banana leaf images into diseased or healthy categories.
SVM classifies data points in an N-dimensional space, CNN is used for image recognition, and ANN is inspired by biological neural networks for pattern recognition. Each method has its advantages and limitations, with CNN being unable to handle rotation. Overall, these methods contribute to accurate classification of banana leaf images for disease management.
Source link
Source link: https://www.nature.com/articles/s41598-024-63930-y
GIPHY App Key not set. Please check settings