24 October 2014 | Lei Li, Qin Zhang, Danfeng Huang
This paper reviews various imaging techniques used in plant phenotyping, which is crucial for understanding and improving plant traits related to growth, yield, and stress tolerance. The review covers visible imaging, imaging spectroscopy, thermal infrared imaging, fluorescence imaging, 3D imaging, and tomographic imaging (MRI, PET, and CT). Each technique is described in terms of its principles, current applications, and advantages and limitations. Visible imaging is simple and cost-effective but limited by the complexity of plant structures. Imaging spectroscopy, particularly near-infrared spectroscopy, is promising for non-destructive measurements but faces challenges with data volume and cost. Thermal imaging is useful for detecting stress responses but is influenced by environmental factors. Fluorescence imaging is effective for early stress detection but has limitations in large-scale applications. 3D imaging techniques like LIDAR and stereo vision provide detailed plant structures but are costly and complex. Tomographic imaging techniques like MRI, PET, and CT offer high-resolution 3D data but are low-throughput. The paper also discusses the integration of multiple sensors for more accurate discrimination of stress types and the development of phenotyping platforms that combine sensing, control, and computing technologies.This paper reviews various imaging techniques used in plant phenotyping, which is crucial for understanding and improving plant traits related to growth, yield, and stress tolerance. The review covers visible imaging, imaging spectroscopy, thermal infrared imaging, fluorescence imaging, 3D imaging, and tomographic imaging (MRI, PET, and CT). Each technique is described in terms of its principles, current applications, and advantages and limitations. Visible imaging is simple and cost-effective but limited by the complexity of plant structures. Imaging spectroscopy, particularly near-infrared spectroscopy, is promising for non-destructive measurements but faces challenges with data volume and cost. Thermal imaging is useful for detecting stress responses but is influenced by environmental factors. Fluorescence imaging is effective for early stress detection but has limitations in large-scale applications. 3D imaging techniques like LIDAR and stereo vision provide detailed plant structures but are costly and complex. Tomographic imaging techniques like MRI, PET, and CT offer high-resolution 3D data but are low-throughput. The paper also discusses the integration of multiple sensors for more accurate discrimination of stress types and the development of phenotyping platforms that combine sensing, control, and computing technologies.