Comprehensive Performance Analysis of Multiple Mango Leaf Diseases Detection Using Deep Learning
摘要
In the agriculture field, detecting new leaf diseases is becoming increasingly difficult because of environmental changes. The disease occurs when living plants are infected. Fungi are the most common pests that cause plant diseases. Plant diseases reduce product quality or crop sustainability, decrease the nutritional value of fruits and vegetables, reduce yields, and render some crops unmarketable. Every year, diseases cause great economic losses for the industry. 20 to 40% of global agricultural production is destroyed by pests. However, in this article, mango leaf diseases are classified into eight different classes. The finetuned Inception V3 model uses the “Mango Leaf Disease Dataset” and achieves comparable accuracy from binary class classification to multi-class classification in disease detection. The comparative result analysis gradually decreases in accuracy as the number of classifications increases.