Prediction of plant pest detection using improved mask FRCNN in cloud environment. (December 2022)

• Record Type: Journal Article
• Authors: Deepika, P.; Arthi, B.
• Is Part Of: Measurement. Volume 24(2022)
• Publication Date: 2022-12
• Subjects: Crop plants -- Artificial intelligence -- IMFRCNN -- MobileNet; Detectors -- Periodicals; Measurement -- Periodicals; 530.7
• DOI: 10.1016/j.measen.2022.100549
• Languages: English
• ISSNs: 2665-9174

Abstract

Abstract: According to the Indian Food and Agriculture Organizations (FAO), agricultural plants can disrupt the prospect of a worldwide productivity loss of about 20–40%. To avoid these overwhelming insect plants and increase crop yield, smart farming for farmers was developed using modern technology with Artificial Intelligence (AI).The foremost objective of this paper is to apprehend the advanced application for operational agricultural plantonline recognition on plant open fields such as greenhouses and large farms. This paper proposes Deep Learning (DL) for experts and agriculturalists with a help of an innovative mobile application to control the insects which affect plants. Improved MaskFaster Region-Based Convolutional Neural Network (IMFR-CNN) method proposed for developing technologically advanced application exploits to undertake the recognize exhaust of insect plants in Cloud Computing (CC). Additionally, for mentoring the farmers, the planticides database was maintained to suggest theplanticides andprovide the details of the spotted crop plants to the farmers. Four groups of plantsGram pod borer, Tobacco caterpillar, Whitefly, Spider mites, and Root-knot nematode are considered in this research.The proposed IMFRCNN model evaluated and produces the maximum accuracy of 99.0% for every confirmed plant image for detecting the plant disease at an earlier stage than the existing methods such as Back Propagation (BP) and Single Shot Multi-Box Detector (SSD) DL method.