One DAG to Rule Them All

In this paper, we present novel strategies for optimizing the performance of many binary image processing algorithms. These strategies are collected in an open-source framework, GRAPHGEN, that is able to automatically generate optimized C++ source code implementing the desired optimizations. Simply starting from a set of rules, the algorithms introduced with the GRAPHGEN framework can generate decision trees with minimum average path-length, possibly considering image pattern frequencies, apply state prediction and code compression by the use of directed rooted acyclic graphs (DRAGS). Moreover, the proposed algorithmic solutions allow to combine different optimization techniques and significantly improve performance. Our proposal is showcased on three classical and widely employed algorithms (namely Connected Components Labeling, Thinning, and Contour Tracing). When compared to existing approaches -in 2D and 3D-, implementations using the generated optimal DRAGs perform significantly better than previous state-of-the-art algorithms, both on CPU and GPU.

Automated summary

This paper presents novel strategies for optimizing the performance of binary image processing algorithms. These strategies are integrated into an open-source framework, GRAPHGEN, which automatically generates optimized C++ source code. The proposed algorithms generate decision trees with minimum average path-length, consider image pattern frequencies, and utilize DRAGs for state prediction and code compression. The experimental results demonstrate that the proposed approach significantly improves performance compared to existing algorithms on both CPU and GPU.