Tomato bacterial wilt severity identification dataset

Published: 8 August 2023| Version 1 | DOI: 10.17632/skg3w6yxhc.1
Xin Wang


Hyperspectral dataset of tomato plants at different bacterial wilt levels. The hyperspectral images were acquired by using a Specim IQ camera (Specim Ltd., Oulu, Finland) Tomato bacterial wilt severity was usually assessed by visual inspection of the percentage of wilted leaves. In this experiment, it was classified as: level 0, healthy; level 1, infected but no obvious symptoms of wilting; level 2, infected and 1-33% of leaves wilting; level 3, infected and 34-66% of leaves wilting; level 4, infected and 67-100% of leaves wilting. The infected tomato plants were artificially inoculated in the greenhouse. Two hyperspectral images are given for each bacterial wilt level along with their corresponding RGB images. In addition, the hyperspectral images are given in .tif format, which can be processed directly in matlab.



Agricultural Engineering