An Optimal Weighted Segmentation with Hough Transform based Feature Extraction and Classification Model for Citrus Disease

In the agricultural sector, plant diseases are chiefly answerable for the decrease in production which have resulted in economic loss. The citrus plant is employed as a foremost source of nutrients such as vitamin C all over the globe. But, citrus disease greatly affects the productivity and quality of the citrus fruits. At present times, computer vision-based models are generally employed to detect and classify plant diseases effectively. This study devises a proficient citrus disease recognition and classifier approach to identify the presence of disease in the citrus fruit. The proposed method involves a set of different processes namely Bilateral filtering based preprocessing, optimal weighted segmentation (OWS), Hough Transform (HT) based feature extraction, and rough fuzzy artificial neural network (RFANN) based classification. The presented model called HT-RFANN is tested using a Citrus Image Gallery dataset and the extensive experimentation strongly highlighted the effective outcome of the HT-RFANN technique on the test images.