Open AccessCombination of saliency histogram equalisation and seam carving for image resizing
Author(s) -
Hsin HsiChin
Publication year - 2017
Publication title -
the journal of engineering
Language(s) - English
Resource type - Journals
ISSN - 2051-3305
DOI - 10.1049/joe.2017.0003
Subject(s) - seam carving , histogram , image (mathematics) , carving , computer science , artificial intelligence , computer vision , domain (mathematical analysis) , histogram matching , computer graphics (images) , simple (philosophy) , mathematics , engineering , mechanical engineering , mathematical analysis , philosophy , epistemology
This study presents a novel approach to content‐aware image sizing by combining the continuous approach using saliency histogram equalisation (SHE) and the discrete approach using seam carving (SC). It constructs a non‐uniform quad mesh to represent an image by equalising its saliency histogram (SH), and changes the mesh size to represent the resized image by carving seams of insignificant quads in the SH domain. As SHE is simple to implement, and SC can be obtained via dynamic programming, the proposed SHE combined with SC (SHE‐SC) algorithm is computationally efficient. Experimental results show that SHE‐SC is preferable to other existing algorithms.