Dataset for "Predicting the growth trajectory and yield of greenhouse strawberries based on knowledge-guided computer vision"

Overall

A strawberry dataset for the paper "Qi Yang, Licheng Liu, Junxiong Zhou, Mary Rogers, Zhenong Jin, 2024. Predicting the growth trajectory and yield of greenhouse strawberries based on knowledge-guided computer vision, Computers and Electronics in Agriculture, 220, 108911. https://doi.org/10.1016/j.compag.2024.108911"

Plant traits measurements

The folder "measurement.zip" includes treatment-level and fruit-level ground truth data. 

Treatment-level

data_dryMatter_2022.csv
data_dryMatter_2023.csv
data_freshMatter_2022.csv
data_freshMatter_2023.csv
data_fruitNumber_2022.csv
data_fruitNumber_2023.csv
data_plantBiomass_2022.csv
data_plantBiomass_2023.csv

Fruit-level

Fruit conditions with five classes, 1-5 represent Normal, Wizened, Malformed, Wizened & Malformed, and Overripe, respectively.

data_size_freshWeight_condition_2022_0N.csv
data_size_freshWeight_condition_2022_50N.csv
data_size_freshWeight_condition_2022_100N.csv
data_size_freshWeight_condition_2022_150N.csv

Fruit size for tagged fruits

data_taggedFruit_diameter_2022.csv
data_taggedFruit_diameter_2023.csv
data_taggedFruit_length_2022.csv
data_taggedFruit_length_2023.csv

Fresh yield and lifespan for tagged fruits (only available in experiment 2023)

data_taggedFruit_freshMatter_2023.csv
data_taggedFruit_lifespan_2023.csv

Weather data

weather_daily_2022.csv
weather_daily_2023.csv

Image data with label

Object and phenology detection

The folder "strawberry_img_random.zip" contains images and the corresponding JSON labels for object and phenological stages detection.

Fruit size and decimal phenological stage

The folder "strawberry_img_tagged.zip" contains images and the corresponding JSON labels for fruit size and decimal phenological stages detection.

For example,
"label": "small g, 8.84, 7.62, 0.4",
This label means the fruit has an 8.84mm diameter and 7.62mm length,
with the main stage being small green and the decimal stage being DS-4

Merge and split Data

A Python script, "datasetProcessing.py", can be used to merge and split the image data into training and testing set.

Data collector: Dr. Qi Yang, University of Minnesota, USA. Email: [e-mail]

All the files belong to Prof. Zhenong Jin, University of Minnesota, USA. Email: [e-mail]