A multi-instance learning based approach for whitefly pest detection

Agriculture constantly faces various challenges including attacks from new pests and insects. With large farm sizes plummeting manpower in the agricultural sector, it becomes challenging to continuously monitor crops for pest infestation. In this research paper, a specific type of attack known as white fly has been investigated which affects variety crops. This paper presents four different approaches automated classification whiteflies are Bayesian network, convolution neural network (CNN), ResNet multi-instance learning-CNN. A comparative analysis with conventional machine learning deep techniques also presented. The performance proposed technique evaluated terms accuracy. experimental results obtained show that attains accuracy 95.53%, 96.9%, 97.6% 98.13% models respectively. classificaiton, existing shows outperforms baseline identifying whitefly