Ensemble Technique of Deep Learning Model for Identifying Tomato Leaf Diseases Based on Choquet Fuzzy Integral

• DOI: 10.1080/01969722.2024.2343993
• Publication Date: 2024-05-08T16:37:00Z
• Authors: Yan Hang; Xiangyan Meng; Rui Wang; Xiaole Tian; Junqi Cui
• Keywords: Choquet integral; Ensemble Learning

As tomato leaves are often attacked by various microorganisms, pests and bacterial diseases, the yield of is seriously reduced. Accurate timely identification leaf diseases great significance to reduce farmers' economic losses. Ensemble learning as a combinatorial optimization method, which can improve generalization ability model stability, widely used in field plant disease identification. However, commonly ensemble methods such majority voting, weighted averaging, etc. do not consider interaction between inputs when aggregating multiple models, that they produce representative outputs. To solve this problem, paper adds fuzzy algorithms i.e., five pre-trained deep namely VGG16, VGG19, Xception InceptionV3 InceptionResnet V2, integrated using Choquet integral integrals for four classes classification. The experimental results show proposed method achieves encouraging results, with best single achieving 98.63% accuracy on PlantVillage dataset 99.80% accuracy. For natural scenarios, 97% effectively identify diseases.