Plant disease detection is one of the most active topics in the modern agriculture. The disease in plants are season-based which depends on the presence of the pathogen, crops, environmental conditions and varieties grown. The monitoring of leaf area is important in studying physiological capabilities associated with plant. This work makes use of image processing technique for the detection of disease and the use of Support Vector Machine for the classification of plant leaf disease. Plant Leaf disease detection and classification is performed, depending on various extracted features from plant leaves utilizing different image processing and deep learning techniques. Detection of plant leaf disease involves steps like data collection, image processing techniques like contrast enhancement, RGB to HSI, K-means clustering, feature extraction, segmentation and SVM based classification. This approach is useful when image dimensions are large and a reduced feature representation is required to efficiently complete tasks such as image matching and retrieval. The proposed work mainly concentrates on four major diseases that affect the plant leaf namely Alternaria alternata, Anthracnose, Bacterial blight and Cercospora leaf spot. The dataset considered for each disease is 22, 23, 20 and 20 respectively. The results of a test case for each of the four diseases are quantified and the percentage of disease affected area was observed to be 15.0013% in Alternaria alternata, 15.0015% in Anthracnose, 15.0142% in Bacterial blight and 23.0225% in Cercospora leaf spot.
Paper
Full text
Deep Learning Approach for the Detection of Plant Diseases
Semantic Scholar · Agricultural and Food Sciences · 2021
Abstract
Plant disease detection is one of the most active topics in the modern agriculture. The disease in plants are season-based which depends on the presence of the pathogen, crops, environmental conditions and varieties grown. The monitoring of leaf area is important in studying physiological capabilities associated with plant. This work makes use of image processing technique for the detection of disease and the use of Support Vector Machine for the classification of plant leaf disease. Plant Leaf disease detection and classification is performed, depending on various extracted features from plant leaves utilizing different image processing and deep learning techniques. Detection of plant leaf disease involves steps like data collection, image processing techniques like contrast enhancement, RGB to HSI, K-means clustering, feature extraction, segmentation and SVM based classification. This approach is useful when image dimensions are large and a reduced feature representation is required to efficiently complete tasks such as image matching and retrieval. The proposed work mainly concentrates on four major diseases that affect the plant leaf namely Alternaria alternata, Anthracnose, Bacterial blight and Cercospora leaf spot. The dataset considered for each disease is 22, 23, 20 and 20 respectively. The results of a test case for each of the four diseases are quantified and the percentage of disease affected area was observed to be 15.0013% in Alternaria alternata, 15.0015% in Anthracnose, 15.0142% in Bacterial blight and 23.0225% in Cercospora leaf spot.