Naman-Gayakwad/Water-Issues-Prediction-Model
Domaine:
environment and energyType de record:
modelsoftware
Créateur:
Nam
Hôte:
# Water Issues Prediction Model
This repository contains a **machine learning model** designed to predict water-related issues such as **urban floods, drainage blockages, and sanitation problems** using image analysis.
## 📁 Repository Structure
```
Water-Issues-Prediction-Model/
│-- convert_model.py # Converts Keras model to TFLite format
│-- infer.py # Runs inference using the trained model
│-- predict_tflite.py # Runs inference using the TFLite model
│
├── Model/
│ │-- model_tf.keras # Trained Keras model
│ │-- model.tflite # Converted TFLite model
│ │-- sklearn_model.p # Scikit-learn trained model
│
├── Prediction Using Model/
│ │-- Sample Images/
│ │ │-- dirtylaketest.jpg
│ │ │-- drainagehole.jpg
│ │ │-- drainagetest.jpeg
│ │ │-- flood_test.jpg
│
├── Training/
│ │-- data/
│ │ │-- train/ # Training dataset
│ │ │-- val/ # Validation dataset
│ │-- sklearn_trained_model.py # Sklearn training script
│ │-- tflitemodel.py # Converts the trained model to TFLite
```
## 🚀 Features
- Predicts water-related issues from images.
- Uses **Scikit-Learn, Keras**, and **TFLite** for efficient model deployment.
- Includes both **standard and lightweight models (TFLite)** for mobile-friendly inference.
- Provides scripts for training, conversion, and prediction.
## 📌 Installation
To set up the environment, install the required dependencies:
```sh
pip install -r requirements.txt
```
## 📊 Model Training
To train the model using **Scikit-Learn**:
```sh
python Training/sklearn_trained_model.py
```
To train and convert the model to **TFLite**:
```sh
python Training/tflitemodel.py
```
## 🔄 Model Conversion
To convert the Keras model to **TFLite** format:
```sh
python convert_model.py
```
## 🏗️ Running Inference
Run inference using the standard model:
```sh
python infer.py --image_path Sample Images/d …