Soil Erosion Image Segmentation Based on Improved K-means clustering method

DOI

10.2991/icseee-16.2016.163How to use a DOI?

Keywords

soil erosion; images segmentation; improved K-means clustering method

Abstract

A method was developed for soil erosion image segmentation based on improved K- means clustering in order to solve some problem with the traditional K-means clustering method. First, the peaks was detected in the smoothed histogram of a gray-scale image, and sort peaks in descending order; then the number of clusters K is determined according to the number of main peaks in the smoothed histogram, while the grey value of a peak is selected as a cluster center; finally, the weighted Euclidean distance was applied to measure the similarity instead of simple Euclidean distance. Experiment results show that the improved K-means clustering method can not only shorten the process of convergence to the object but also get a more reasonable clustering result. It is effective and suitable for soil erosion images segmentation.

Copyright

© 2016, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Xuanzhang Song
AU  - Jianqiang Liu
AU  - Qiongyan Li
PY  - 2016/12
DA  - 2016/12
TI  - Soil Erosion Image Segmentation Based on Improved K-means clustering method
BT  - Proceedings of the 2016 5th International Conference on Sustainable Energy and Environment Engineering (ICSEEE 2016)
PB  - Atlantis Press
SP  - 911
EP  - 916
SN  - 2352-5401
UR  - https://doi.org/10.2991/icseee-16.2016.163
DO  - 10.2991/icseee-16.2016.163
ID  - Song2016/12
ER  -