Edge Vision for Plant Disease Monitoring: Lightweight Detection from Field Images to Farm Decisions

Abstract

Edge Smart Agriculture poses a recurring problem of coordinating performance with evidence quality, resource limits, and transfer across settings. The present synthesis investigates connecting compact visual models with sensing conditions, deployment limits, and agronomic action. The corpus joins 1 focal paper with 11 independently retrieved publications confirmed at bibliographic registration or publisher level. The analysis is organized around model compression, domain shift, label quality, device latency, and decision support. To avoid reading reported outcomes as directly interchangeable, the review compares research framing, method assumptions, and test envelope. Across the literature, the literature consistently implies that advances in edge smart agriculture become credible when representation, objective, and evaluation protocol are evaluated together and when uncertainty about distribution shift is reported explicitly. The synthesis ties method selection to operational consequence while identifying external-validity hazards, and proposes a research agenda centered on well-specified controls, robustness tests, and reproducible workflows.