Convolutional neural networks for predicting moisture ratio from images of beetroot cubes under different drying pretreatments

The dataset consists of 140 observations derived from RGB images collected during the drying experiments under different pretreatments and temperatures. Each record corresponds to a single image and includes associated experimental conditions and moisture-ratio values. The variables contained in the dataset are:

filename — the name of the image file (e.g., bc6001.jpg), where the prefix identifies the pretreatment:
bc (control), be (ethanol), bcu (ultrasound), and beu (ultrasound + ethanol).

treatment — the specific pretreatment applied to the samples prior to drying (control, ethanol, ultrasound, or ultrasound + ethanol).

temperature — drying temperature in degrees Celsius (60, 70, or 80 °C).

time — drying time (min) at which the corresponding image was captured.

rm1 — experimental moisture ratio calculated from gravimetric measurements.

rm2 (two columns) — predicted moisture ratio values associated with image-based modeling; duplicated entries originate from the raw export format.

Together, these variables represent paired experimental and image-based drying information, enabling the evaluation of moisture-ratio prediction models under controlled conditions. The dataset includes all combinations of pretreatment, temperature, and drying time used in the study.