USEQ: Ultra-Fast Superpixel Extraction via Quantization


  • Author(s):

Chun-Rong Huang, Wei-An Wang, Szu-Yu Lin, and Yen-Yu Lin

 


Introduction

    We propose a novel superpixel extraction method named USEQ to generate regular and compact superpixels. To reduce the computational burden of iterative optimization procedures used in most recent approaches, the spatial and color quantizations are performed in advance to represent pixels and superpixels. Maximum a posteriori estimation in both pixel and region levels is then adopted to aggregate pixels into spatially and visually coherent superpixels. The resultant superpixels are extremely efficient to generate and can more precisely adhere to object boundaries. Compared to the state-of-the-art approaches to superpixel extraction, USEQ can achieve better or competitive performance in terms of boundary recall, undersegmentation error and achievable segmentation accuracy, and is significantly faster than these approaches.


Software Download

Download USEQ (ver 1.0)

 

Please refer to main.cpp for the sample code of USEQ.

 


Copyrights

    This software of USEQ is available under the (GNU GPL or FIPL) for personal research usage only. Any commercial use or any redistribution of this software requires a license from the National Chung Hsing University. This license file must be retained with all copies of the software, including any modified or derivative versions. For further details on obtaining a commercial license, please contact Chun-Rong Huang (crhuang@nchu.edu.tw). THE NATIONAL CHUNG HSING UNIVERSITY AND THE AUTHORS MAKE NO REPRESENTATIONS OR WARRANTIES OF ANY KIND CONCERNING THIS SOFTWARE.


Release Notes

 

  • V1.0 First release of USEQ.


Feedback

Chun-Rong Huang crhuang at nchu.edu.tw


Publications

[1] Chun-Rong Huang, Wei-An Wang, Szu-Yu Lin, and Yen-Yu Lin, "USEQ: Ultra-Fast Superpixel Extraction via Quantization," in Proc. International Conference on Pattern Recognition, 2016.