ParamTunerCV 0.1.2

Parameter tuning tool for computer vision

ParamTunerCV

Professional interactive image viewer and processing application for computer vision research and real-time parameter tuning.

---

📋 Table of Contents

• Quick Start
• Features
• Examples
• Controls
• Interface Gallery
• Architecture
• API Reference
• Contributing
• License

🚀 Quick Start

  # Option 1: Install from PyPI (recommended for users)
  pip install ParamTunerCV

  # Option 2: Install with uv (fastest)
  uv add ParamTunerCV

  # Option 3: Development setup with uv (recommended for contributors)
  git clone https://github.com/harsh194/ParamTunerCV.git
  cd ParamTunerCV
  uv sync

  # Option 4: Development setup with pip
  git clone https://github.com/harsh194/ParamTunerCV.git
  cd ParamTunerCV
  pip install -e .

  # Test the installation
  python examples/01_basic_usage.py

Requirements: Python 3.8+, OpenCV 4.10.0, NumPy 2.2.6, Matplotlib 3.10.3

✨ Features

• 50+ Pre-configured Trackbars - Real-time parameter tuning for morphology, edge detection, filtering, thresholding
• 15+ Specialized Factory Viewers - Ready-to-use viewers for Canny, Hough transforms, corner detection
• Interactive ROI System - Rectangle, polygon, and line drawing with visual feedback
• Multi-Space Thresholding - Support for 7 color spaces (BGR, HSV, Lab, etc.) with binary, adaptive, Otsu methods
• Professional Analysis Suite - Multi-channel histograms, pixel intensity profiles, JSON/CSV export
• Dual Operation Modes - Full GUI for development, headless for production automation

📖 Examples

Start with these progressive examples in the examples/ folder:

Example	Description
01_basic_usage.py	Core workflow and main loop pattern
02_window_control.py	Window management and UI controls
03_trackbar_definitions.py	Custom trackbars and parameter control
04_app_debug_mode.py	GUI vs headless operation modes
05_viewer_factory_usage.py	Pre-built viewers for common CV tasks

Quick Integration

from ParamTunerCV import ImageViewer, ViewerConfig
import numpy as np
import cv2

config = ViewerConfig()

trackbar_definitions = [
    {"name": "Show Image", "param_name": "show", "max_value": "num_images-1", "initial_value": 0},
    {"name": "Count", "param_name": "count", "max_value": 50, "initial_value": 10}, 
    {"name": "Gauss Size", "param_name": "GaussianSize", "max_value": 31, "callback": "odd", "initial_value": 5}, 
    {"name": "Thresh Val", "param_name": "thresh_val", "max_value": 255, "initial_value": 128}, 
]

IMG_HEIGHT, IMG_WIDTH = 600, 800
viewer = ImageViewer(config, trackbar_definitions)

while viewer.should_loop_continue():
    params = viewer.trackbar.parameters
    
    current_thresh = params.get("thresh_val") 
    current_gaussian_size = params.get("GaussianSize", 5)
    block_count = params.get("count", 10)

    base_color_image = np.full((IMG_HEIGHT, IMG_WIDTH, 3), (block_count * 5, 0, 0), dtype=np.uint8)
    cv2.rectangle(base_color_image, (50, 50), (IMG_WIDTH - 50, IMG_HEIGHT - 50), (0, 255, 0), 3)

    gray_image = cv2.cvtColor(base_color_image, cv2.COLOR_BGR2GRAY)

    gauss_image = cv2.GaussianBlur(gray_image, (current_gaussian_size, current_gaussian_size), 0)

    _, thresh_image = cv2.threshold(gauss_image, current_thresh, 255, cv2.THRESH_BINARY)
    
    viewer.display_images = [
        (base_color_image, "Color"),
        (gray_image, "Grayscale Image"),
        (gauss_image, "Gaussian Blur)"),
        (thresh_image, "Threshold image")
    ]

viewer.cleanup_viewer()

🖱️ Controls

Mouse: Left drag (ROI), wheel (zoom), middle drag (pan), right click (remove selection)

Keys: R (rectangle), L (line), P (polygon), H (histogram), Q/ESC (quit)

📸 Interface Gallery

🎛️ Real-Time Parameter Control

50+ preconfigured trackbars for real-time OpenCV parameter tuning - adjust Gaussian blur, edge detection, morphological operations, and thresholding with immediate visual feedback

📊 Professional Analysis Suite

Comprehensive analysis control panel with ROI management, drawing tools, histogram generation, pixel intensity profiling, and integrated data export functionality

🎨 Advanced Multi-Space Thresholding

Sophisticated thresholding interface supporting 7 color spaces (BGR, HSV, Lab, etc.) with binary, adaptive, Otsu, and range-based methods for precise image segmentation

💾 Multi-Format Data Export

Professional export system for analysis results - save histograms, pixel profiles, and geometric data in JSON/CSV formats with configurable options

🏗️ Architecture

src/ParamTunerCV/ 
├── core/           # ImageViewer main orchestrator
├── config/         # ViewerConfig management  
├── controls/       # TrackbarManager for real-time controls
├── events/         # MouseHandler for interactions
├── gui/            # WindowManager and UI components
├── analysis/       # Analysis modules (plotting, export, threshold)
└── utils/          # Factory methods and utilities

Design: Factory pattern for viewers, Observer pattern for callbacks, Fluent interface for configuration

📚 API Reference

Testing

python examples/01_basic_usage.py  # Primary test

🤝 Contributing

1. Fork the repository
2. Create feature branch: git checkout -b feature/amazing-feature
3. Test changes: python examples/01_basic_usage.py
4. Submit pull request

Coding Standards: PEP 8, docstrings, type hints

📄 License

MIT License © 2025 Harsh Ranjan
Contact: [e-mail]

---

⬆️ Back to Top