.png){kind=icon}
.png){kind=icon}
Return
A graph based superpixel generation algorithm
.png){kind=icon}
PRE
AI
DOI：10.1007/s10489-018-1223-1.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In recent years, superpixels have become a prevailing tool in computer vision and many methods have been proposed. However, due to the problems such as high time complexity, low object boundary adherence and irregular shape, only a few methods are widely used. To improve these issues, we propose a novel general superpixel segmentation method called minstpixel, which relies on energy functional minimization. Minstpixel introduces an energy functional based on minimal spanning tree and designs a strategy to gain the global optimum. It never needs sophisticated optimization scheme, complicated mathematical deduction or fussy iteration process. At the same time, the time complexity of minstpixel is approximately linear with respect to the number of image pixels. The benchmark on Berkeley segmentation database shows that minstpixel could rival state-of-the-art in every aspect.
Keywords:
Superpixels
Energy minimization
Minimization spanning tree
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.