https://github.com/jannatul-fredaues/CropLens-AI

CorpLens is an AI-powered flower recognition system that identifies flower species from images using deep learning and computer vision techniques.
https://github.com/jannatul-fredaues/CropLens-AI

artificial-intelligence-projects computer-vision deep-learning flower-classification image-recognition keras machine-learning python tensorflow

Last synced: about 24 hours ago
JSON representation

CorpLens is an AI-powered flower recognition system that identifies flower species from images using deep learning and computer vision techniques.

• Host: GitHub
• URL: https://github.com/jannatul-fredaues/CropLens-AI
• Owner: jannatul-fredaues
• Created: 2026-02-06T12:46:08.000Z (5 months ago)
• Default Branch: main
• Last Pushed: 2026-06-14T04:23:30.000Z (29 days ago)
• Last Synced: 2026-06-14T06:22:48.358Z (29 days ago)
• Topics: artificial-intelligence-projects, computer-vision, deep-learning, flower-classification, image-recognition, keras, machine-learning, python, tensorflow
• Language: Jupyter Notebook
• Homepage: https://crop-lens-ai.vercel.app
• Size: 1.87 MB
• Stars: 8
• Watchers: 0
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

• 100-AI-Machine-learning-Deep-learning-Computer-vision-NLP - 👆

README

          
# CropLens - AI

## Flower Classification & Dataset Pipeline (3-Class)

An end-to-end Computer Vision project for flower classification and dataset engineering, designed for research and real-world AI applications.

## Overview

This project focuses on building a robust flower dataset (3 classes) and applying deep learning techniques for classification and preprocessing.

### It includes:

1.Dataset collection & cleaning;

2.Background removal pipeline;

3.Data augmentation;

4.Model training & evaluation.

## Objectives:

1.Build a high-quality flower dataset;

2.Remove noisy backgrounds for better feature learning;

3.Train a high-performance classification model;

4.Provide a reproducible pipeline for research use.

## Preview

![Home](assets/home.png)

## Methodology:

### 1. Data Preprocessing

Background removal using OpenCV / rembg;

Image resizing (e.g., 224x224);

Noise filtering.

### 2. Data Augmentation

Rotation;

Flipping;

Brightness adjustment.

### 3. Model Architecture

CNN / Transfer Learning;

Loss: CrossEntropyLoss;

Optimizer: Adam.