测绘通报 ›› 2024, Vol. 0 ›› Issue (10): 84-90.doi: 10.13474/j.cnki.11-2246.2024.1014.

• 学术研究 • 上一篇    

基于Spark与优化分块的大幅面遥感影像SLIC分割方法

谢志伟1,2,4, 宋光明2, 张丰源3, 陈旻1,4, 彭博5   

  1. 1. 南京师范大学虚拟地理环境教育部重点实验室, 江苏 南京 210042;
    2. 沈阳建筑大学交通与测绘 工程学院, 辽宁 沈阳 110168;
    3. 南京师范大学环境学院, 江苏 南京 210042;
    4. 南京师范大学地理 科学学院, 江苏 南京 210042;
    5. 辽宁生态工程职业学院测绘工程学院, 辽宁 沈阳 110122
  • 收稿日期:2024-01-30 发布日期:2024-11-02
  • 通讯作者: 张丰源,E-mail:fengyuan.zhang@polyu.edu.hk.
  • 作者简介:谢志伟(1986—),男,博士,副教授,主要研究方向为遥感图像处理。E-mail:zwxrs@sjzu.edu.cn
  • 基金资助:
    国家自然科学基金(42101353);教育部人文社会科学研究基金(21YJC790129);辽宁省教育厅基本科研项目(LJKMZ20220946);辽宁省教育厅基本科研项目(LJKMZ20222128)

Large format remote sensing image segmentation method based on Spark with optimised chunking

XIE Zhiwei1,2,4, SONG Guangming2, ZHANG Fengyuan3, CHEN Min1,4, PENG Bo5   

  1. 1. Key Laboratory of Virtual Geographic Environment, Ministry of Education, Nanjing Normal University, Nanjing 210042, China;
    2. College of Transportation and Surveying Engineering, Shenyang Jianzhu University, Shenyang 110168, China;
    3. College of the Environment, Nanjing Normal University, Nanjing 210042, China;
    4. School of Geographic Sciences, Nanjing Normal University, Nanjing 210042, China;
    5. School of Surveying and Mapping Engineering, Liaoning Ecological Engineering Vocational College, Shenyang 110122, China
  • Received:2024-01-30 Published:2024-11-02

摘要: 针对大幅面遥感影像在分块边界特征不连续和分割效率不高等问题,本文提出了结合Spark平台及最优紧密度评估的简单线性迭代聚类超像素分割算法(SLIC)。首先,使用结合最优紧密度的SLIC超像素分割方法完成图像分块,解决分块边界精度低的问题;然后,利用Spark对分块数据并行SLIC分割算法,提高运算效率;最后,将WorldView-2卫星影像和GF-2号影像作为试验数据,利用比值植被指数结合最大类间方差法改进SLIC算法以提高超像素分割精度。结果表明,改进SLIC方法在运算效率上比原方法提高了约9倍,边缘拟合精度提高了1.5%,欠分割误差提高了8.2%,边缘召回率提高了0.2%。

关键词: 大幅面遥感影像, Spark平台, 改进SLIC算法, 并行计算, 最优参数评估

Abstract: Aiming at the large format remote sensing image in the chunk boundary feature discontinuity and segmentation efficiency is not high. In this paper, we propose a simple linear iterative clustering superpixel segmentation algorithm (SLIC) that combines the Spark platform and optimal compactness evaluation. Firstly, the SLIC superpixel segmentation method with optimal tightness is used to complete the image chunking, which solves the problem of low accuracy of the chunk boundary; then, the SLIC segmentation algorithm is used in parallel to the chunked data by using Spark to improve the computational efficiency; finally, the SLIC algorithm is improved by using the ratio of vegetation index combined with the method of maximum interclass variation to improve the accuracy of the superpixel segmentation.WorldView-2 Satellite Imagery and GF-2 images are used as experimental data. The experimental results show that the improved SLIC method improves about 9 times of the original method in terms of computing efficiency, 1.5% of the edge fitting precision, 8.2% of the under-segmentation error, and 0.2% of the edge recall.

Key words: large format remote sensing image, Spark platform, improved SLIC algorithm, parallel computing, optimal parameter evaluation

中图分类号: