Prakruty Plant Diagnosis App
An on-device agronomist for every grower. Point your camera at any troubled leaf to identify 40+ pathologies in 2.4 seconds with safe, non-toxic bio-remedies — even with zero cellular signal in the field.
Four Core Capabilities Built for the Field
Instant 40+ Foliar Pathologies Detected
Trained on certified botanical pathology samples across tomato, paddy, chili, soybean, and pulses to differentiate fungal blights from bacterial spots and nutrient hunger.
Non-Toxic Organic Treatment Recipes
Clear formulation ratios for neem oil emulsions, copper soap biocides, Trichoderma bio-fungicides, and microbial compost teas. Zero groundwater contamination.
100% Offline Edge Inference (2.4 Seconds)
The entire 14MB quantized vision model runs within phone silicon memory. Works deep in valley fields, forest fringes, and remote Chhattisgarh blocks without mobile signal.
Crop Health History & Treatment Tracker
Saves past scans with date stamps. Compare leaf lesion spread over 5 to 7 days to verify whether bio-treatment is stopping spore propagation.
Prakruty Plantii
Local On-Device Engine
Tomato Late Blight
94.8% MatchPhytophthora infestans · High Severity
Copper octanoate soap or bio-fungicide (Trichoderma viride).
Dilute 15ml per 4.5L rainwater. Spray leaves early morning. Isolate infected vines.
How Scans Add to the Chhattisgarh Crop-Health Dataset (With Consent)
Every scan remains private to the handset by default. Here is how voluntary grower participation builds an early warning shield for farming communities across Chhattisgarh.
1. Private on the Handset
When you photograph a leaf, the inference model processes the image locally. No photo is automatically transmitted to any server or cloud database.
2. Explicit Grower Consent
After reading your diagnosis, a simple toggle allows you to contribute the anonymous leaf pathology tag to the Chhattisgarh regional dataset. You can revoke this anytime.
3. Village Early Warning
Consented scans are stripped of metadata and clustered at the district level (e.g., Durg or Dhamtari). When 15+ early blight spots appear in a block, neighboring growers receive proactive warning alerts.
Supported Crops & Disease Library
Calibrated on Chhattisgarh agricultural patterns and ICAR reference foliar pathologies.
Tomato Late Blight
94.8%Phytophthora infestans
Water-soaked dark lesions spreading rapidly across foliage.
Early Blight
91.2%Alternaria solani
Brown spots with concentric rings surrounded by yellow chlorosis.
Septoria Leaf Spot
88.6%Septoria lycopersici
Small circular spots with dark brown margins and grey centres.
Bacterial Spot
86.4%Xanthomonas campestris
Small greasy water-soaked spots turning black with yellow halo.
Try On-Device Diagnosis On Real Leaves
Tap different spots on the leaf to see how our offline model analyzes pathogens and suggests safe organic remedies.
Tap any pulsing dot on the leaf to run instant on-device diagnosis
Tomato Late Blight
Water-soaked dark lesions spreading rapidly across leaf tips during moist weather.
Apply copper soap or bio-fungicide (Trichoderma viride)
Mix 15ml copper octanoate soap in 4.5 litres of rainwater. Spray early morning on both sides of leaves. Prune infected stems with sterilized shears.
Frequently Asked Questions
Straight answers about app compatibility, offline storage, and closed beta access.
Yes. The plant vision model runs directly on your smartphone processor. Once downloaded during setup, you do not need 4G, 5G, or WiFi to photograph a leaf and see the diagnosis. It works deep inside rural fields, basements, or valleys.
Join the Closed Beta Waitlist
Free access for active growers in Chhattisgarh and Central India. Get early Android APK downloads and seasonal organic guidance.