Editorial: Machine vision and machine learning for plant phenotyping and precision agriculture

Abstract

Machine vision and machine learning for plant phenotyping and precision agriculturePlant phenotyping (PP) describes the physiological and biochemical properties of plants affected by both genotypes and environments.It is an emerging research field assisting the breeding and cultivation of new crop varieties to be more productive and resilient to challenging environments.Precision agriculture (PA) uses sensing technologies to observe crops and then manages them optimally to ensure that they grow in healthy conditions, have maximum productivity, and have minimal adverse effects on the environment.Traditionally, the observation of plant traits heavily relies on human experts, which is labour-intensive, time-consuming, and subjective.Although PP and PA are two different fields, they share similar sensing and data processing technologies in many respects.Recently, driven by computer and sensor technologies, machine vision (MV) and machine learning (ML) have contributed to accurate, high-throughput and nondestructive sensing and data processing technologies to PP and PA.However, these technologies are still in their infant stage, and many challenges and questions related to them still need to be addressed.This Research Topic aims to share the latest research results on applying MV and ML to PP and PA.It demonstrates cutting-edge technologies, bottle-necks and future research directions for MV and ML in crop breeding, crop cultivation, and disease or pest management.This Research Topic of Frontiers in Plant Sciences published a total of 28 peer-reviewed research articles, including one review paper for the phenotyping of Prunoideae fruits (Liu et al.).These articles reveal the latest research trends regarding different crop species, data types and algorithms.The summary of the published reports shows that cotton (Gossypium), canola or oilseed rape (Brassica napus), wheat (Triticum) and maize (Z.mays) are the most important crops for study in PP and PA (Figure 1A).Cotton stands out as the most frequently examined crop, with a total of five articles dedicated to it.Yan et al. developed a leaf segmentation method in the field environments.Tang et al. investigated early detection

Code and data availability

This is an editorial summarizing a Research Topic of 28 articles; it contains no paper-specific datasets, images, code, models, or supplements with availability statements, and no URLs are provided.

No evidence-backed public reproduction asset is currently recorded.