Skip to content

README.md

Latest commit

 

History

History
94 lines (72 loc) · 3.34 KB

README.md

File metadata and controls

94 lines (72 loc) · 3.34 KB

Ordy

A React Native mobile application that analyzes your photo gallery to detect monochromatic images and find similar/duplicate photos using advanced image processing algorithms.

Gallery View Scan Results Scanning

Technical Architecture

Core Components

  • PhotoGrid: Renders photo collection in a customizable grid with efficient memory management
  • ScanTab: Manages scanning workflows and result visualization
  • BatchProcessor: Handles concurrent image analysis operations

Key Technologies

  • React Native + TypeScript
  • Expo Media Library for asset access
  • SHA-256 perceptual hashing for image similarity detection
  • Parallel processing architecture for analyzing large photo libraries

Features

Gallery Management

  • Paginated loading of device media
  • Responsive grid display with date metadata
  • Modal view for photo inspection
  • Pull-to-refresh and infinite scroll implementations

Photo Analysis

  • Monochrome Detection: Identifies images with color uniformity above 92%
  • Similarity Detection: Finds similar photos with >87% hash similarity
  • Duplicate Detection: Identifies near-identical photos with >98% hash similarity
  • Real-time progress tracking with dual-phase progress display

Performance Optimizations

  • Batch processing with configurable chunk sizes
  • In-memory caching of analysis results
  • Limited parallel operations to prevent device overload
  • Progressive asset loading for large libraries

Implementation Details

Image Analysis Pipeline

  1. Asset Retrieval: Efficiently loads photo metadata in configurable batches
  2. Perceptual Hashing: Generates compact signatures for similarity comparison
    • Images are resized to 16×16px
    • Color data is transformed into cryptographic hash
  3. Color Analysis: Identifies monochromatic images
    • Samples color distribution in 32×32px downsamples
    • Calculates variance across color spectrum
    • Applies 92% uniformity threshold for classification
  4. Similarity Comparison: Compares hash signatures using Hamming distance algorithm

Configuration Constants

  • HASH_SIZE: 16px (hash generation dimensions)
  • COLOR_ANALYSIS_SIZE: 32px (color analysis dimensions)
  • MONOCHROME_THRESHOLD: 0.92 (92% color uniformity)
  • SIMILARITY_THRESHOLD: 0.87 (87% hash similarity)
  • BATCH_SIZE: 100 (photos processed per batch)
  • MAX_PARALLEL_OPERATIONS: 25 (concurrent operations)

Getting Started

Prerequisites

  • Node.js 12+
  • React Native environment
  • Expo CLI

Installation

# Clone the repository
git clone https://github.com/yourusername/ordy.git

# Install dependencies
cd ordy
npm install

# Start development server
expo start

Performance Considerations

The application implements several optimizations for processing large photo libraries:

  • Memory-efficient processing: Photos are processed in configurable batches
  • Operation throttling: Parallel operations are limited to prevent device lockup
  • Progressive loading: Two-phase loading strategy for improved UX
  • Hash caching: Previously processed images are cached to avoid redundant processing