---

license: other language: - en task_categories: - image-classification tags: - plant-disease - agriculture - crops - leaf - banana - tomato - potato - lemon - tea - rice - watering - irrigation - treatment size_categories: - 100K<n<1M format: image modality: - image library_tags: - pytorch - tensorflow

CropHelth — Unified Crop Disease & Watering Dataset

~125,000 plant leaf images across 109 disease/healthy classes for 26 crops, plus a per-image provenance manifest, a treatment knowledge base, and per-crop watering recommendation rules.

Built for the CropHelth project: an image classifier that outputs code-based JSON (e.g. {"status": {"code": "potato_lb"}, "treatment": {"pill_code": ...}}) and an environment-based watering decision engine.

Layout

images/<plant>_<ill>/NNNNNN.jpg     # 109 classes, ~125k images (the training set)
_source_manifest.csv                 # per-image provenance: target_code, source, original_path
detection/                           # PlantDoc object-detection dataset (VOC: .jpg + .xml, TRAIN/TEST)
knowledge/
  names.json                         # class code -> human-readable message
  treatments.json                    # class code -> {treatment: {pill_code, pill_name}, agronomy}
  watering_rules.json                # per-crop thresholds for the watering decision engine
  watering_recommendations.csv       # human-readable watering reference table (ph, temp, moisture, practice)
labels.json                          # class list + code->index + per-class counts
README.md

Class code scheme

<plant>_<ill> — e.g. potato_late_blight, banana_sigatoka, tea_helopeltis, tomato_healthy.

• *_healthy = healthy class for that plant (the model's "healthy" output).
• Plant = first token of the code.
• 109 classes, one folder per class — drop images/ straight into any classification pipeline (ImageFolder / the CropHelth trainer both work as-is).

Sources (all open source)

Source	DOI / URL	Contribution
Mendeley tywbtsjrjv — "Data for: Identification of Plant Leaf Diseases Using a 9-layer Deep CNN" (PlantVillage)	https://data.mendeley.com/datasets/tywbtsjrjv/1	apple, blueberry, cherry, corn, grape, orange, peach, pepper, potato, raspberry, soybean, squash, strawberry, tomato (~52k)
Mendeley 9tb7k297ff — BananaLSD	https://data.mendeley.com/datasets/9tb7k297ff/1	banana (sigatoka, cordana, pestalotiopsis, healthy; Original set only — augmented set excluded to avoid synthetic-duplicate bias)
Mendeley 744vznw5k2 — teaLeafBD	https://data.mendeley.com/datasets/744vznw5k2/4	tea (7 classes)
Mendeley j32xdt2ff5 — Tea Sickness	https://data.mendeley.com/datasets/j32xdt2ff5/2	tea (8 classes, merged with teaLeafBD)
Mendeley 6243z8r6t6 — Multi-Crop Disease (roboflow)	https://data.mendeley.com/datasets/6243z8r6t6/1	banana, cauliflower, chilli, peanut, radish (~22k; class assigned via YOLO label files)
Mendeley 643f5bbc2t — BDLemonLeaf	https://data.mendeley.com/datasets/643f5bbc2t/6	lemon (13 classes, Raw Image set)
Mendeley 8d9fv6kpt3 — Large-Scale Lemon Leaf Disease & Pest	https://data.mendeley.com/datasets/8d9fv6kpt3/1	lemon (18 classes, merged with BDLemonLeaf)
Mendeley c5yvn32dzg — RoCoLe (robusta coffee)	https://data.mendeley.com/datasets/c5yvn32dzg/2	coffee healthy/unhealthy (labels from the shipped Labelbox CSV)
Mendeley 22p2vcbxfk — Groundnut (peanut)	https://data.mendeley.com/datasets/22p2vcbxfk/3	peanut (5 classes, Raw_Data)
Mendeley 7vpdrbdkd4 — Nutrient-deficient banana	https://data.mendeley.com/datasets/7vpdrbdkd4/2	banana nutrient deficiencies (7 classes, RAW set only)
Mendeley g7xnn2bm4g — Nitrogen deficiency in maize	https://data.mendeley.com/datasets/g7xnn2bm4g/1	maize N0/N7/NF (3 classes)
GitHub pratikkayal/PlantDoc-Dataset	https://github.com/pratikkayal/PlantDoc-Dataset	28 classes (train+test merged)
GitHub pratikkayal/PlantDoc-Object-Detection-Dataset	https://github.com/pratikkayal/PlantDoc-Object-Detection-Dataset	detection/ (VOC format, separate task)
Harvard Dataverse LQUWXW — Bananas TZ (Tanzania)	https://doi.org/10.7910/DVN/LQUWXW	banana black sigatoka, fusarium wilt, healthy (~15k large photos)

Excluded by design: Background_without_leaves (PlantVillage, not a plant class), augmented/synthetic image sets (our trainer does its own augmentation), and the Zenodo DiaMOS pear dataset (12.5 GB single archive — could not be processed in the build environment; add it with the same pipeline).

Note: the IEEE DataPort Paddy Doctor (rice) dataset requires a logged-in IEEE account to download — it could not be included automatically. Rice classes should be added the same way when you download it.

Watering decision data

knowledge/watering_rules.json drives a rule engine (water_now / hold_water / check_data):

• soil_moisture_pct < min → water_now
• soil_moisture_pct > max → hold_water (overwatering risk)
• no soil moisture given → check_data
• ph / air_temp_c / soil_temp_c outside range → extra reason codes (ph_low, ph_high, air_temp_high, soil_temp_high)

Per-crop values (moisture %, pH window, temperature stress limits) in watering_rules.json + the readable practice table in watering_recommendations.csv. Values are standard agronomic reference ranges — calibrate them with your own field sensors. Soil moisture is the primary signal; pH/temperature alone cannot decide irrigation. To replace rules with a learned model, log readings + actual watering actions and train a classifier (see the CropHelth trainer's train_watering.py).

Treatments

knowledge/treatments.json maps every non-healthy class to {treatment: {pill_code, pill_name}, agronomy}. These are standard extension-service style reference recommendations for training/testing — before any real-world use, verify products are registered in your region, follow label rates, rotate FRAC groups, and have a local agronomist review. The model predicts the class code only; treatment text always comes from this file (never model-generated).

Provenance & licenses

_source_manifest.csv maps every image back to its source dataset and original path. Each source keeps its own license (see detection/LICENSE.txt for PlantDoc; Mendeley datasets are typically CC-BY/CC0 — check each dataset page). If you redistribute, keep the source attribution above.

Quick usage (CropHelth trainer)

# point the trainer at this dataset
python scripts/build_labels.py --images-dir <this repo>/images
python scripts/split.py
python scripts/train.py
python scripts/predict.py --image some_leaf.jpg --meta
python scripts/predict_watering.py --json '{"crop":"potato","soil_moisture_pct":18,"ph":6.1}'