Detection of the Tulip Breaking Virus (TBV) in tulip using spectral and vision sensors

摘要

Experiments are described for detection of Tulip Breaking Virus in tulip plants. Four optical techniques were investigated and compared with visual assessment by crop experts as well as with Elisa (Enzyme ImmunoAssay) analysis of the same plants. The optical sensor techniques used were: an RGB Color camera, a spectrophotometer ranging from 400-2400 nm, a spectral imaging camera covering a spectral range from 400-900 nm and a chlorophyll fluorescence imaging system which measures the photosynthetic activity. Linear discriminant classification was used to compare the results of these optical techniques with the Elisa score and the visual assessment. A good correlation was found when the results of the spectral camera and the results of the visual examinations of the crop expert were compared. These are very promising results which are translated in a follow up with as final goal an autonomous robot for detection and removal of diseased tulip plants in the open field.