Back to Projects
Completed Project

AI Livestock Diagnostic Mobile App

Cross-platform cattle health monitoring app with biometric muzzle-print ID, CNN skin disease diagnosis, MFCC acoustic health screening, real-time health records, and backend API synchronization. Built from scratch using Flutter, Dart, and Kotlin.

|
AI Livestock Diagnostic Mobile App
RoleLead Mobile & Embedded AI Engineer
TimelineFeb 2026 - May 2026
TeamInternship Project
Tech Stack8 Technologies

Mission Brief

A cutting-edge on-device AI mobile diagnostic application for cattle and livestock health management. Incorporates biometric muzzle-print computer vision for unique cattle identification (analogous to human fingerprints), Convolutional Neural Networks (CNN) for photographic skin disease detection (e.g. Lumpy Skin Disease), and Mel-Frequency Cepstral Coefficients (MFCC) audio signal processing for respiratory cough screening. Designed with offline-first synchronization to operate reliably in remote rural farm environments.

Key Features

Biometric & Vision AI

  • Muzzle-Print Identification: OpenCV edge contour and feature descriptor matching for cattle biometric recognition
  • CNN Dermatological Screening: On-device TensorFlow Lite model diagnosing lesions and lumpy skin infections from camera feeds
  • Real-Time Bounding Boxes: Instant visual confidence overlays and severity scoring

Acoustic Respiratory Analysis

  • MFCC Audio Extraction: High-frequency audio sampling converting cough recordings into Mel-frequency spectrograms
  • Respiratory Pathology Classifier: Differentiates healthy bovine sounds from pneumonia and respiratory distress
  • Noise-Reduction Filters: Native Kotlin audio pre-processing isolating animal vocalizations from tractor and ambient farm noise

Mobile Architecture & Sync

  • Offline SQLite Engine: Full local database allowing field vets to log medical histories without network connectivity
  • Background Sync: Automatic batch synchronization with cloud FastAPI backend once network connectivity resumes
  • Multi-Language UI: Localized interfaces tailored for grassroots dairy farmers and field technicians

Engineering Chronicles

Running real-time image recognition and MFCC signal extraction on low-cost Android smartphones caused memory throttling and frame drops.

SolutionQuantized TensorFlow Lite models to 8-bit integers (INT8) and offloaded audio processing into native C++/Kotlin Android NDK isolates, achieving steady 60 FPS UI performance.

Unpredictable cellular reception in rural farmlands caused data loss when veterinarians submitted field diagnosis reports.

SolutionBuilt an offline-first transactional SQLite write-ahead queue with idempotent UUID records and automatic retry mechanisms on network reconnect.

Installation Instructions

1. Prerequisites

Ensure Flutter 3.x, Dart 3.x, and Android SDK (API 34+) are installed and configured in your PATH.

2. Clone & Install Dependencies
$git clone https://github.com/KartikSharma4448/Pranag-fluter.git
$cd Pranag-fluter
$flutter pub get
3. Run on Connected Device
$flutter run --release

Project Access

Live DemoSource Code

Technologies

Flutter
Dart
Kotlin
OpenCV
TensorFlow Lite
Audio MFCC
FastAPI
SQLite

Table of Contents

  • • Mission Brief
  • • Key Features
  • • Engineering Chronicles
  • • Visual Gallery
  • • Installation Instructions
Back to Projects

More Projects

PRANAG AI – Web Platform
Completed

PRANAG AI – Web Platform

React.js • AI & Machine Learning

VCC ERP – Coaching Institute Management
Completed

VCC ERP – Coaching Institute Management

Flutter 3.41 • Full Stack & ERP Ecosystem

Rajasthali – Travel & Fleet Management System
Completed

Rajasthali – Travel & Fleet Management System

Flutter 3.7+ • Fleet & Mobile Logistics

CVCraft v2 – AI-Powered ATS Resume Builder
Completed

CVCraft v2 – AI-Powered ATS Resume Builder

FastAPI • AI & Full Stack Web App

Aegis Care – Blockchain Healthcare System
Completed

Aegis Care – Blockchain Healthcare System

Algorand • Web3 & Blockchain