Faster Zhang-Suen Thinning Algorithm

This repository provides an optimized implementation of the Zhang-Suen thinning algorithm for skeleton extraction. It is adapted from the skimage.morphology module in the scikit-image library and is designed to be a high-performance alternative to the cv::ximgproc::thinning function in OpenCV. The implementation is tailored for seamless integration and ease of use in C++ projects.

Performance Comparison

Benchmark tests on an Intel i7-8700k processor:

• Faster Thinning: 0.08632 seconds per run
• OpenCV cv::ximgproc::thinning: 0.27614 seconds per run

Benchmark tests on an Intel i5-10400 processor:

• Faster Thinning: 0.02880 seconds per run
• OpenCV cv::ximgproc::thinning: 0.24943 seconds per run

Visualization Results

input : output :

Seamless Integration

This implementation is designed to be a drop-in replacement for cv::ximgproc::thinning. You can easily integrate it into your existing C++ projects without modifying your core logic. Simply import the provided code and use it as a direct substitute.

Key Features

• High Performance: Achieves faster processing times without compromising the quality of the skeleton extraction.
• Ease of Integration: Compatible with existing OpenCV-based projects, ensuring minimal disruption to your workflow.
• Reliability: Based on the well-established Zhang-Suen thinning algorithm, ensuring accurate and consistent results.
• Proven Implementation: Adapted from the widely-used and well-regarded scikit-image library.
• C++ Friendly: Specifically optimized for C++ usage, with clear and concise code structure.

License

This project is licensed under the MIT License.