Arbitrary Screen-Aware Manga Reading Framework with Parameter-Optimized Panel Extraction
While reading Manga on smart devices, it is inconvenient to drag and zoom through the contents. Therefore, this work proposes a user-friendly and intuitive Manga reading framework for smart devices of a limited screen size and resolution and arbitrary aspect ratio and orientation. Our framework creates a reading flow by breaking pages into fundamental contextual panels, organizing panels as display units based on commonly accepted Manga reading rules, grouping units as a display list according to the physical and reading conditions, and playing the list with designed transitions to simulate eye movement of Manga reading. The core is parameter-optimized panel extraction which uses the occupied area of extracted panels and the ratio of border lines parallel to each other to automate panel extraction of matching up identified candidate panel corners and borders. While applying to 70 Manga chapters of 1272 pages from 17 titles, our extraction algorithm performs comparatively better against two state-of-the-art algorithms and successfully locates break-out panels which others generally fail to identify. Furthermore, we conduct a usability test on three chapters of One Piece, Kuroko No Basketball, and God Only Knows. The result shows that our system can provide an effective and intuitive Manga reading mechanism with the reading flow and panel presentation for arbitrary scenarios while preserving contextual ideas through panel arrangement.