Efficient Binary and Run Length Morphology and its Application to Document Image Processing

TL;DR

This paper presents an open source library for efficient mathematical morphology operations on binary and run-length compressed images, optimized for document image processing, with benchmarks demonstrating performance improvements.

Contribution

It introduces a novel implementation of morphology operations on compressed images and provides comprehensive benchmarks against traditional bit-blit methods.

Findings

Significant performance gains over bit-blit implementations.

Effective processing of document images using compressed representations.

Open source library facilitates practical applications in document imaging.

Abstract

This paper describes the implementation and evaluation of an open source library for mathematical morphology based on packed binary and run-length compressed images for document imaging applications. Abstractions and patterns useful in the implementation of the interval operations are described. A number of benchmarks and comparisons to bit-blit based implementations on standard document images are provided.