MuSSE: A Tool to Extract Meta-Data from Game Sprite Sheets Using Blob Detection Algorithm

• Published in: Brazilian Symposium on Games and Digital Entertainment (Print) pp. 61 - 69
• Main Authors: De Barros Barbosa, Marcelo, De Barros Barbosa, Cecilia, Freitas Barbosa, Andre
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.11.2015
• Subjects: blob detection; Detection algorithms; development tool; Engines; Feature extraction; Games; Image edge detection; Labeling; sprite sheet; Sprites (computer)
• ISSN: 2159-6662
• DOI: 10.1109/SBGames.2015.22

This paper introduces a tool to automatically generate meta-data from game sprite sheets. MuSSE is a tool developed to extract XML data from sprite sheet images with non-uniform - multi-sized - sprites. MuSSE (Multi-sized Sprite Sheet meta-data Exporter) is based on a Blob detection algorithm that incorporates a connected-component labeling system. Hence, blobs of arbitrary size can be extracted by adjusting component connectivity parameters. This image detection algorithm defines boundary blobs for each individual sprite in a sprite sheet. Every specific blob defines a sprite characteristic within the sheet: position, name and size, which allows for subsequent data specification for each blob/image. Several examples on real images illustrate the performance of the proposed algorithm and working tool.