Planti

A small but mighty plant identifier

Learn More

Automatic database updates

750+ species

1500+ refinements

Up to 93 years on a single charge

Virtual herbarium

Syncing with MehtorAPI

Cross-platform access

Automatic database updates

to eliminate the need for cumbersome cables and external computers, allowing users to effortlessly refresh botanical keys and taxonomy files while out in the field

Wireless Over-the-Air (OTA) Technology

The device features built-in Wi-Fi modules that allow it to receive firmware and database packages wirelessly

Local File Rewriting

When an update is triggered, the system directly edits and overwrites the structured text files stored within the microcontroller's internal LittleFS or SPIFFS file system

No External Tools Required

This architecture means users can effortlessly refresh botanical keys and taxonomic data in the field without ever needing to plug the device into a computer
750+ species
to ensure high field utility by packing a massive, ecologically diverse, and deterministic taxonomic dataset directly onto an offline wearable device
  • Extensive Paper Guide Digitization: The database contains comprehensive data for $751$ unique plant species, completely adapted from M.I. Neishtadt’s classic botanical identification guide
  • Ecological Diversity: The dataset covers a wide array of taxonomic groups, including aquatic plants (hydrophytes), desert/steppe flora (xerophytes), ferns, trees, shrubs, and herbaceous plants
  • High Field Utility: By packing this entire diverse floristic complex locally, the device offers highly reliable, deterministic taxonomic diagnosis across various regional ecosystems

1500+ refinements

to transform messy, paper-based botanical guides into highly accurate digital data, achieving a 95.45% extraction accuracy rate for flawless algorithm navigation

AI-Backed Text Processing

Converting a traditional paper reference book required running pages through Optical Character Recognition (OCR) and using AI models to clean up fragmented text, typos, and ink smudges

Structural Data Optimization

We refined the non-linear, multi-step branching paths of the original text to map cleanly into an optimized binary decision tree

Experimental Accuracy Peak

These extensive refinements resulted in an optimal text preprocessing zoom factor of 4.8, achieving a remarkable text data extraction accuracy of 95.45%
Up to 93 years on a single charge
to showcase the extreme energy efficiency of the device's microcontrollers compared to power-hungry smartphones, promising long-lasting reliability on remote expeditions
  • The Real Specs: The device draws .017 mA in deep sleep mode, 115–120 mA while idle, and up to 130 mA during active plant identification
  • True Field Longevity: The final prototype's 2,500 mAh battery delivers an impressive 100+ hours of continuous active field use, which easily blows standard smartphones out of the water

Virtual herbarium

to let users catalog, track, and review their botanical discoveries entirely offline, acting as a personal digital ledger of their wilderness field research

Local Data Logging

This built-in feature allows users to save their deterministic plant identification results straight to the microcontroller's flash memory with a single click

Interactive Collection Management

Users can open the dedicated "Library" screen to view an organized list of saved plants alongside a counter showing how many of each species they have collected

Instant Descriptions

Clicking on any plant entry within this local database instantly pulls up its full scientific description, characteristics, and habitat details on the screen
Syncing with MehtorAPI
to provide a reliable backup pipeline, ensuring that discoveries logged in remote, signal-free environments are securely uploaded to the web once connectivity is found
  • Data Synchronization Module: The final hardware prototype (TYPE A) is integrated with a functional synchronization system to bridge local and external data
  • Ecosystem Integration: This feature links the offline device's localized file system directly with the cloud backend of the custom "Mehtor" ecosystem
  • Reliable Cloud Backups: It ensures that botanical collections logged completely offline during deep wilderness expeditions can be safely backed up to the web once Wi-Fi or network access becomes available
Planti

Cross-platform access

To expand accessibility, we built a matching Telegram MiniApp that perfectly mirrors the graphical interface and dichotomous tree logic of the physical device

Check it out


Inclusive Educational Tool

This dual approach ensures that students, researchers, and hobbyists can utilize the exact same powerful identification algorithm on smartphones, tablets, or PCs without even needing the physical hardware

Full description of

the scientific work

English

Full description of

the scientific work

Russian(original)

© 2021-2026 Mehtor, All rights reserved.

‌

HTML Website Maker