Two tree-based methods for the waterfall

作者：

Highlights：

• We present two new methods for the classical waterfall transform.

• Both methods are tree-based and easier to implement than the state-of-the-art.

• The Simplified Waterfall (SW) method is 32% faster than the state-of-the-art; the Balanced Waterfall (BW) method is 16% faster.

• The Balanced Waterfall method additionally handles non-minimal plateaux robustly.

摘要

Highlights•We present two new methods for the classical waterfall transform.•Both methods are tree-based and easier to implement than the state-of-the-art.•The Simplified Waterfall (SW) method is 32% faster than the state-of-the-art; the Balanced Waterfall (BW) method is 16% faster.•The Balanced Waterfall method additionally handles non-minimal plateaux robustly.

论文关键词：Image segmentation,Waterfall transform

论文评审过程：Received 4 September 2013, Revised 9 February 2014, Accepted 16 April 2014, Available online 28 April 2014.

论文官网地址：https://doi.org/10.1016/j.patcog.2014.04.017