Sugar Cane - Spittle Bug

Published: 29-07-2020| Version 1 | DOI: 10.17632/pwprmck9h5.1
Carlos Vasquez,
Claudia Castiblanco


This data set provides 12,000 images for individual pest bugs in sugar cane crops, collected from pictures taken in yellow square traps on field, which are used to monitoring the pest presence and population. Data set is divided half for special bug of interest, spittle bug (Aeneolamia varia) which will be the True data set, and half for other insects in the trap which will be the False Data set. The data set is useful but not only, to train Convolutional Neural Networks – CNN, for binary classification (True/False). Data set images are in different sizes.


Steps to reproduce

Download both set of true and false spittle bug (6,000 each data set). Unzip the file to get the whole images and to create the personalized data sets for training, validation and testing depending of model to classify.