Abstract
Quality agriculture production is the essential trait for any nation’s economic growth.The agricultural sector has been facing great challenges to feed the increasing number of population living in the world. In the future, it will be very difficult to rely on traditional farming to produce food. In plants, pepper is used as a major source of nutrients throughout the world. However, peppers diseases badly affect the production and quality of pepper plants. Image processing and machine learning techniques have been widely used in the agriculture for detection and classification of diseases in plants. In this thesis, we propose a method for classification of diseases in peppers plants. The proposed method consists of several stages, which are include, the image acquisition stage, image pre-processing, features extraction, features normalization, and the classification stage. In the pre-processing stage, The proposed technique is tested on the new dataset for pepper plant images, the dataset named (DiyalaPepper) and contains (244) samples of healthy and unhealthy images for both peppers fruits and leaves collectively. The total number of leaves pepper images in dataset equal to (166) and (78) images for pepper fruits. Accuracy rate obtained from the peppers fruits image was 81.82 % , and 94.11% for peppers leaves image.
