Hyperspectral Imaging for Identification of an Invasive Plant Mikania micrantha Kunth.

Yiqi Huang,Rui Yang,Fanghao Wan,Xi Qiao,Yanzhou Li,Shuo Zhang,Fukuan Wang,Wanqiang Qian,Jie Li

doi:10.3389/fpls.2021.626516

Abstract

Mile-a-minute weed (Mikania micrantha Kunth) is considered as one of top 100 most dangerous invasive species in the world. A fast and accurate detection technology will be needed to identify M. micrantha. It will help to mitigate the extensive ecologic and economic damage on our ecosystems caused by this alien plant. Hyperspectral technology fulfills the above requirement. However, when working with hyperspectral images, preprocessing, dimension reduction, and classifier are fundamental to achieving reliable recognition accuracy and efficiency. The spectral data of M. micrantha were collected using hyperspectral imaging in the spectral range of 450–998 nm. A different combination of preprocessing methods, principal component analysis (for dimension reduction), and three classifiers were used to analyze the collected hyperspectral images. The results showed that a combination of Savitzky-Golay (SG) smoothing, principal component analysis (PCA), and random forest (RF) achieved an accuracy (A) of 88.71%, an average accuracy (AA) of 88.68%, and a Kappa of 0.7740 with an execution time of 9.647 ms. In contrast, the combination of SG, PCA and a support vector machine (SVM) resulted in a weaker performance in terms of A (84.68%), AA(84.66%), and Kappa (0.6934), but with less execution time (1.318 ms). According to the requirements for specific identification accuracy and time cost, SG-PCA-RF and SG-PCA-SVM might represent two promising methods for recognizing M. micrantha in the wild.

Highlights

Mikania micrantha Kunth (M. micrantha), known as “mile-a-minute,” is one of the world’s 100 most dangerous invasive species (Khadka, 2017)
Where TP is the number of samples correctly predicted to be M. micrantha, TN is the number of samples correctly predicted as the background, FP is the number of background samples incorrectly predicted as M. micrantha, and FN is number of M. micrantha samples incorrectly predicted to be the background
The raw spectral data distribution of M. micrantha in the remaining spectral range is almost the same as the background. It indicates that the intra-class differences were more than interclass differences of M. micrantha and background, and it is a challenging work for M. micrantha identification

Summary

Introduction

Mikania micrantha Kunth (M. micrantha), known as “mile-a-minute,” is one of the world’s 100 most dangerous invasive species (Khadka, 2017). It is estimated that M. micrantha can produced between 90,000 and 210,000 seeds/m2 (Macanawai et al, 2012; Day et al, 2016). The seeds are dispersed by wind, animals, and humans (Yang et al, 2005; Day et al, 2016). Hyperspectral Recognition of Invasive Plant has been influenced by this weed (Shen et al, 2017). The yield losses of banana (Musa spp.), Citrus spp., and sugarcane (Saccharum officinarum L.) infested with M. micrantha ranged from 60 to 70% due to the twining which would block out sunlight (Shen et al, 2013). The economic losses were estimated at US$650,000–1.6 M/year on Neilingding Island (about 554 ha; Zhong et al, 2004). Identifying and monitoring M. micrantha are urgent, which would allow the plant to be controlled by providing accurate information about its geographical distribution (Tesfamichael et al, 2018)

Methods

Results

Discussion

Conclusion