Dataset: Electronic Nose for Quality Control of Colombian Coffee through the Detection of Defects in “Cup Tests”

Published: 11-03-2020| Version 1 | DOI: 10.17632/7spd6fpvyk.1
Juan Carlos Rodriguez Gamboa,


This dataset is related to the research paper "Electronic Nose for Quality Control of Colombian Coffee through the Detection of Defects in "Cup Tests" published in Sensors (Basel) journal (available online since December 24, 2009, Please cite the mentioned article if you found useful information or use the dataset. This dataset comprises time-series recorded by an electronic nose system (E-Nose) for an application of coffee quality control focused on the detection of defects. The database contains 58 measurements of coffee divided into three groups and labeled as high quality (HQ), average quality (AQ), and low quality (LQ). The time series acquired at 1 Hz of sampling frequency during 300 seconds correspond to 300 data points per sensor. Each file in the dataset has eight columns with the resistance readings in kΩ of the eight gas sensors: SP-12A, SP-31, TGS-813, TGS-842, SP-AQ3, TGS-823, ST-31, TGS-800. We organized the database into three folders: AQ_Coffee, HQ_Coffee, LQ_Coffee. Each folder contains text files that correspond to different measurements. The filename identifies the measurement as follows: the first two characters of the filename are an identifier of the quality coffee threshold (AQ: average-quality, HQ: high-quality, LQ: low-quality); and the last two characters indicate the measurement number. For example, file LQ_01 contains the time series recorded when the sample 01 of low-quality coffee was measured.