基于模板匹配和不变矩的浮游动物快速测量方法

徐帅; 杨俊毅; 郑旻辉; 谢尚微

doi:10.3969/j.issn.1001-909X.2017.04.007

PDF(1731 KB)

海洋学研究 ›› 2017, Vol. 35 ›› Issue (4) : 69-75. DOI: 10.3969/j.issn.1001-909X.2017.04.007

研究论文

基于模板匹配和不变矩的浮游动物快速测量方法

徐帅^1,2, 杨俊毅^1,2, 郑旻辉^1,2, 谢尚微^1,2

作者信息 +

The method of rapid measurement of marine zooplankton based on template matching and invariant moment

XU Shuai^1,2, YANG Jun-yi^1,2, ZHENG Min-hui^1,2, XIE Shang-wei^1,2

Author information +

文章历史 +

摘要

近年来基于图像处理方法的浮游动物分类技术逐渐用于海洋生态系统的研究中,相关检测仪器也从实验室处理向原位现场分析发展,这对检测算法的效率和处理速度提出了更高要求。本文根据海洋浮游动物显微图像的形状特点,提出将模板匹配方法与不变中心矩描述方法相结合,先利用模板匹配限定目标初步范围,再比较不变矩确定最终目标动物,并统计动物数量和尺寸。本文方法不受动物方位旋转和尺寸缩放的影响。实验验证结果表明,该方法与传统识别方法相比简单高效,处理速度快,误差范围小,适用于浮游动物的实时原位观测。

Abstract

Zooplankton is widely distributed in the marine ecosystem, which plays an important role in global climate change. Zooplankton analysis based on image processing needs rapid development. According to the shape of marine zooplankton microscopic image characteristics, a hybrid method based on template matching and invariant central moment description method was proposed in this study. Template matching was used to define the target initial range, and the final target organisms were determined by comparing the moment invariants. The amount and size of the target were calculated. This method was not affected by the rotation of biological azimuth and the scaling of size. The verified experiments show that the proposed method is fast and efficient, which is suitable for situ zooplankton monitoring in real time.

导出引用

徐帅, 杨俊毅, 郑旻辉, 谢尚微. 基于模板匹配和不变矩的浮游动物快速测量方法[J]. 海洋学研究. 2017, 35(4): 69-75 https://doi.org/10.3969/j.issn.1001-909X.2017.04.007

XU Shuai, YANG Jun-yi, ZHENG Min-hui, XIE Shang-wei. The method of rapid measurement of marine zooplankton based on template matching and invariant moment[J]. Journal of Marine Sciences. 2017, 35(4): 69-75 https://doi.org/10.3969/j.issn.1001-909X.2017.04.007

中图分类号： Q178.53

参考文献

[1] CULVERHOUSE P F, REGUERA B, GONZALEZ G S, et al. Do experts make mistakes[J]. Marine Ecology Progress,2003,247(247):17-25.
[2] LOKE R E, BUF J M H D, BAYER M M, et al. Diatom classification in ecological applications[J]. Pattern Recognition,2004,37(6):1 283-1 285.
[3] HU Q,DAVIS C. Accurate automatic quantification of taxa-specific plankton abundance using dual classification with correction[J]. Marine Ecology Progress,2006,306(8):51-61.
[4] TANG X, STEWART W K, HUANG H, et al. Automatic plankton image recognition[J]. Artificial Intelligence Review,1998,12(1):177-199.
[5] YANG R, ZHANG R, SUN S. Automated classification of zooplankton based on digital image processing[J]. Computer Simulation,2006,23(5):167-170.
[6] RAFAEL C G, RICHARD E W. Digital image processing[M]. Third edition. Beijing: Publishing House of Electronics Industry,2011.
[7] KENNETH R C.Digital image processing[M]. Beijing:Publishing House of Electronics Industry,2002.
[8] OTSU N A. Athreshold selection method from gray-level histograms[J]. IEEE Transactions System Man and Cybemectics,1979,9(1):62-66.
[9] CANNY J. A computational approach to edge detection[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence,1986,8(6):679-98.
[10] CARSTEN S,MARKUS U,CHRISTIAN W. Machine vision algorithms and application[M]. Beijing: Tsinghua University Press,2008.