imvf 0.1.3

Image Visualization and Feature Extraction Library

ImgVisFeat

Category	Badges
Project
Package
Development
Community

ImgVisFeat is a Python library for image visualization and feature extraction, providing a comprehensive set of tools for analyzing and visualizing various image features.

Features

ImgVisFeat provides the following visualization and feature extraction capabilities:

• Color Channel Visualization: Extract and visualize individual RGB color channels
• Gradient Visualization: Compute and visualize image gradients in X, Y, and combined XY directions
  • ColorGradientVisualizer: For color images
  • GrayGradientVisualizer: For grayscale images
• HoG (Histogram of Oriented Gradients): Visualize feature descriptors for object detection
• LBP (Local Binary Patterns): Extract texture descriptors for texture classification
• Keypoint Detection: Detect and visualize keypoints using SIFT, AKAZE, or ORB algorithms
• Power Spectrum Analysis: Analyze frequency domain characteristics of images
• CLI Tool: Command-line interface for quick visualizations

Requirements

• Python >= 3.10

Installation

Install from PyPI

# Using pip
pip install imvf

# Using uv (faster alternative)
uv add imvf

Development Installation

For development, we recommend using uv:

git clone https://github.com/chatflip/ImgVisFeat.git
cd ImgVisFeat
uv sync

# Install pre-commit hooks
pre-commit install

Development

This project uses make commands for common development tasks:

# Show available commands
make help

# Run tests
make test

# Format code
make format

# Run linting
make lint

# Build documentation
make builddocs

For more details on development workflows, see CLAUDE.md.

Verify Installation

import imvf
print(imvf.__version__)

Quick Start

Using the All-in-One Visualizer

The Visualizer class provides a convenient way to apply all visualization methods at once:

import imvf

# Create visualizer instance
visualizer = imvf.Visualizer()

# Visualize all features and save results to a directory named after the image
visualizer.visualize("path/to/image.jpg")

This will display all visualizations in OpenCV windows and save the results to a directory named path/to/image/.

Using Individual Visualizers

You can also use individual visualizers for specific analyses:

import cv2
import imvf

# Load image
image = cv2.imread("path/to/image.jpg")

# Color channel visualization
color_channel = imvf.ColorChannelVisualizer()
result = color_channel(image)
cv2.imshow("Blue Channel", result.blue)
cv2.imshow("Green Channel", result.green)
cv2.imshow("Red Channel", result.red)

# Gradient visualization (for color images)
gradient = imvf.ColorGradientVisualizer()
result = gradient(image)
cv2.imshow("Gradient X", result.gradient_x)
cv2.imshow("Gradient Y", result.gradient_y)
cv2.imshow("Gradient XY", result.gradient_xy)

# HoG visualization
hog = imvf.HoGVisualizer()
result = hog(image)
cv2.imshow("HoG", result.hog)

# Keypoint detection
keypoint = imvf.KeypointVisualizer(algorithm="SIFT")  # or "AKAZE", "ORB"
result = keypoint(image)
cv2.imshow("Keypoints", result.keypoint)
cv2.imshow("Rich Keypoints", result.rich_keypoint)

Using the CLI

ImgVisFeat provides a command-line interface for quick visualizations:

# Visualize all features
imvf path/to/image.jpg

# Visualize specific feature (currently all methods are applied)
imvf path/to/image.jpg --method all

Available methods:

• all: All visualization methods (default)
• color_channel: Color channel visualization
• gradient: Gradient visualization
• hog: HoG visualization
• lbp: LBP visualization
• keypoint: Keypoint detection
• power_spectrum: Power spectrum analysis

Documentation

For full documentation, including API reference and tutorials, please visit our documentation site.

Project Status

ImgVisFeat is a personal project created for learning and experimentation. While it's open-source and you're welcome to use and learn from it, please note that it may not be actively maintained or updated regularly.

Feedback and Questions

This is a practice repository, but I'm always eager to learn. If you have any questions about the project or suggestions for improvement, feel free to open an issue for discussion. Please understand that responses may not be immediate.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Author

chatflip