Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2024 Jan 15:14:1342970.
doi: 10.3389/fpls.2023.1342970. eCollection 2023.

Non-destructive identification of Pseudostellaria heterophylla from different geographical origins by Vis/NIR and SWIR hyperspectral imaging techniques

Tingting Zhang #  1   2 , Long Lu #  1   2 , Yihu Song #  1 , Minyu Yang  1 , Jing Li  1 , Jiduan Yuan  3 , Yuquan Lin  4 , Xingren Shi  4 , Mingjie Li  1   2 , Xiaotan Yuan  3 , Zhongyi Zhang  1   2 , Rensen Zeng  1   2 , Yuanyuan Song  1   2 , Li Gu  1   2
Affiliations

Non-destructive identification of Pseudostellaria heterophylla from different geographical origins by Vis/NIR and SWIR hyperspectral imaging techniques

Tingting Zhang et al. Front Plant Sci. .

Abstract

The composition of Pseudostellaria heterophylla (Tai-Zi-Shen, TZS) is greatly influenced by the growing area of the plants, making it significant to distinguish the origins of TZS. However, traditional methods for TZS origin identification are time-consuming, laborious, and destructive. To address this, two or three TZS accessions were selected from four different regions of China, with each of these resources including distinct quality grades of TZS samples. The visible near-infrared (Vis/NIR) and short-wave infrared (SWIR) hyperspectral information from these samples were then collected. Fast and high-precision methods to identify the origins of TZS were developed by combining various preprocessing algorithms, feature band extraction algorithms (CARS and SPA), traditional two-stage machine learning classifiers (PLS-DA, SVM, and RF), and an end-to-end deep learning classifier (DCNN). Specifically, SWIR hyperspectral information outperformed Vis/NIR hyperspectral information in detecting geographic origins of TZS. The SPA algorithm proved particularly effective in extracting SWIR information that was highly correlated with the origins of TZS. The corresponding FD-SPA-SVM model reduced the number of bands by 77.2% and improved the model accuracy from 97.6% to 98.1% compared to the full-band FD-SVM model. Overall, two sets of fast and high-precision models, SWIR-FD-SPA-SVM and SWIR-FD-DCNN, were established, achieving accuracies of 98.1% and 98.7% respectively. This work provides a potentially efficient alternative for rapidly detecting the origins of TZS during actual production.

Keywords: Pseudostellaria heterophylla; deep learning; geographical origin; hyperspectral imaging; machine learning.

PubMed Disclaimer

Conflict of interest statement

Authors YL and XS were employed by the company Huzhou Wuxing Jinnong Ecological Agriculture Development Co., Ltd. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Figures

Figure 1
Figure 1
Typical TZS samples from Guizhou (GZ), Jiangsu (JS), Anhui (AH) and Fujian (FJ) Provinces. The numbers “1, 2, 3” on the left represent the different quality levels of TZS.
Figure 2
Figure 2
Raw and average spectra of TZS samples in the range of Vis/NIR and SWIR. (A) Raw spectra and (C) average spectra of TZS samples in the range of Vis/NIR; (B) Raw spectra and (D) average spectra of TZS samples in the range of SWIR.
Figure 3
Figure 3
Scores scatter plots of Vis/NIR and SWIR spectra of TZS from four geographical origins. (A) Vis/NIR spectra; (B) SWIR spectra.
Figure 4
Figure 4
The loss and accuracy curves of the FD-DCNN model based on the SWIR.
Figure 5
Figure 5
The confusion matrices of the PLS-DA, SVM, RF and DCNN models on the prediction set using different preprocessed SWIR spectra.
Figure 6
Figure 6
The process of extracting EWs with CARS and SPA. (A) Number of preferred EWs with CARS; (B) The root mean square error of cross-validation variation with CARS; (C) Regression coefficient path map with CARS; (D) Extraction of EWs with SPA.
Figure 7
Figure 7
The confusion matrices of the simplified PLS-DA, SVM and RF models on the prediction set.
Figure 8
Figure 8
Detection visualization of TZS samples from Guizhou (GZ), Jiangsu (JS), Anhui (AH) and Fujian (FJ) Provinces.
See this image and copyright information in PMC

References

    1. Alchanatis V., Schmilovitch Z., Meron M. (2005). In-field assessment of single leaf nitrogen status by spectral reflectance measurements. Precis. Agric. 6, 25–39. doi: 10.1007/s11119-005-0682-7 - DOI
    1. Barnes R., Dhanoa M. S., Lister S. J. (1989). Standard normal variate transformation and de-trending of near-infrared diffuse reflectance spectra. Appl. Spectrosc. 43, 772–777. doi: 10.1366/0003702894202201 - DOI
    1. Breiman L. (2001). Random forests. Mach. Learn. 45, 5–32. doi: 10.1023/A:1010933404324 - DOI
    1. Burges C. J. (1998). A tutorial on support vector machines for pattern recognition. Data Min. Knowl. Discovery 2, 121–167. doi: 10.1023/A:1009715923555 - DOI
    1. Delwiche S. R., Kim M. S. (2000). Hyperspectral imaging for detection of scab in wheat. Biol. Qual. Precis. Agric. II. 4203, 13–20. doi: 10.1117/12.411752 - DOI