DennisLiu1993/Fastest_Image_Pattern_Matching

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

1151 stars · 269 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

A C++ library implementing an accelerated Normalized Cross Correlation (NCC)-based template matching and image alignment algorithm, based on a ScienceDirect paper for real-time automatic optical inspection. It uses image pyramids, rotation invariance, and SIMD (including ARM Neon) optimizations, with Python bindings via Pybind11.

Use cases

• find a template image inside a larger image fast
• rotation-invariant template matching with OpenCV
• real-time automatic optical inspection pattern alignment
• locate objects in high-resolution images faster than cv2.matchTemplate
• match patterns at multiple angles and positions
• use fast NCC matching from Python
• industrial vision inspection alignment

When to choose

• you need faster-than-OpenCV NCC template matching, especially on large images
• you need rotation-invariant matching with sub-degree precision
• you are building AOI (automatic optical inspection) or machine vision alignment pipelines
• you want Python bindings for a fast C++ matching core

When to avoid

• you need deep-learning-based object detection robust to appearance changes
• you need matching under scale changes or perspective distortion
• you need a pure-Python or easily portable solution without C++ compilation
• your templates are tiny and OpenCV's matchTemplate is already fast enough

Facets

library · maturity active

image-processing computer-vision search-engine computer-vision image-processing machine-learning cpp python windows cross-platform template-matching ncc normalized-cross-correlation opencv image-alignment simd neon pybind11 automatic-optical-inspection pattern-matching linux macos

Member repositories

For agents

markdown · JSON · MCP: product_card(name="DennisLiu1993/Fastest_Image_Pattern_Matching")