Investigations of plant infections by thermal vision and NIR imaging
W.B. Herppich 2
,  
H. Beuche 1
,  
K.-H. Dammer 3
,  
M. Linke 2
,  
 
 
More details
Hide details
1
Department of Technology Assessment and Substance Cycles, Leibniz-Institute of Agricultural Engineering Potsdam-Bornim (ATB), Max-Eyth-Allee 100, D-14469 Potsdam, Germany
2
Department of Horticultural Engineering, Leibniz-Institute of Agricultural Engineering Potsdam-Bornim (ATB), Max-Eyth-Allee 100, D-14469 Potsdam, Germany
3
Department of Engineering for Crop Production, Leibniz-Institute of Agricultural Engineering Potsdam-Bornim (ATB), Max-Eyth-Allee 100, D-14469 Potsdam, Germany
4
Institute for Plant Protection in Field Crops and Grassland, Federal Authority and Federal Research Centre for Agriculture and Forestry, Station Kleinmachnow, Stahnsdorfer Damm 81, D-14532 Kleinmachnow, Germany
 
Int. Agrophys. 2006, 20(1): 1–10
KEYWORDS
ABSTRACT
The successful applicability of thermal vision in horticulture was the starting point to launch a project in plant protection. Earlier studies proved the applicability of thermal vision for the assessment of freshness status and of microbial infestation by local temperature differences of plant parts. In recent laboratory experiments, studies were performed on the development of infestation of wheat plants which were infected by powdery mildew and by stripe rust. Fungi infections, which may cause variations of the surface temperature of plants, could be recognised by infrared cameras in the thermal range under laboratory conditions. Field studies revealed that pronounced natural temperature variations of several Kelvin within the crop canopy prevented the recognition of infected plants. Additionally, the comparatively low resolution of commercial thermal vision systems limited the detection chances of fungi-caused temperature variations within the plant canopy. Near infrared cameras fitted with band-pass filters showed different intensity distributions of the reflected radiation. The evaluation of the spectral intensity relations improved the differentiation.
eISSN:2300-8725
ISSN:0236-8722