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
. 2023 Jul 12;23(14):6331.
doi: 10.3390/s23146331.

A Novel Method for Filled/Unfilled Grain Classification Based on Structured Light Imaging and Improved PointNet+

Shihao Huang  1   2   3 , Zhihao Lu  1 , Yuxuan Shi  1 , Jiale Dong  1 , Lin Hu  1 , Wanneng Yang  4 , Chenglong Huang  1   2   3
Affiliations

A Novel Method for Filled/Unfilled Grain Classification Based on Structured Light Imaging and Improved PointNet+

Shihao Huang et al. Sensors (Basel). .

Abstract

China is the largest producer and consumer of rice, and the classification of filled/unfilled rice grains is of great significance for rice breeding and genetic analysis. The traditional method for filled/unfilled rice grain identification was generally manual, which had the disadvantages of low efficiency, poor repeatability, and low precision. In this study, we have proposed a novel method for filled/unfilled grain classification based on structured light imaging and Improved PointNet++. Firstly, the 3D point cloud data of rice grains were obtained by structured light imaging. And then the specified processing algorithms were developed for the single grain segmentation, and data enhancement with normal vector. Finally, the PointNet++ network was improved by adding an additional Set Abstraction layer and combining the maximum pooling of normal vectors to realize filled/unfilled rice grain point cloud classification. To verify the model performance, the Improved PointNet++ was compared with six machine learning methods, PointNet and PointConv. The results showed that the optimal machine learning model is XGboost, with a classification accuracy of 91.99%, while the classification accuracy of Improved PointNet++ was 98.50% outperforming the PointNet 93.75% and PointConv 92.25%. In conclusion, this study has demonstrated a novel and effective method for filled/unfilled grain recognition.

Keywords: 3D structured light; data enhancement; deep learning; grain classification; normal vector; point cloud segmentation.

PubMed Disclaimer

Conflict of interest statement

The authors declare having no competing interest.

Figures

Figure 1
Figure 1
Schematic diagram of rice grain scanning based on structured light imaging.
Figure 2
Figure 2
Rice grain point cloud acquisition based on structured light imaging.
Figure 3
Figure 3
Schematic diagram of grain segmentation and data enhancement, (a) Original point cloud data, (b1) Sample plane point cloud data, (b2) Original rice grains point cloud data, (c) Single particle rice grain point cloud data, (d) Rice grain point cloud data with up-sampling, (e) Rice grain point cloud data with normal vector fusion.
Figure 4
Figure 4
Schematic diagram of grain parameter extraction based on 3D point clouds.
Figure 5
Figure 5
Improved PointNet++ network structure diagram. The red rectangles indicated the improved modules.
Figure 6
Figure 6
The sampling and grouping for set abstraction module.
Figure 7
Figure 7
Multi-scale sampling module for rice grain point cloud grouping in radius of 0.1, 0.2 and 0.4 mm.
Figure 8
Figure 8
Confusion matrix of filled/unfilled rice grains of Improved PointNet++, (a) PointNet++ without data augmentation, (b) Improved PointNet++ without data augmentation, (c) Improved PointNet++ with data augmentation.
Figure 9
Figure 9
The filled/unfilled rice grain visual prediction for Zhonghua 11 and 9311 rice varieties.
See this image and copyright information in PMC

References

    1. Chen K., Horton R.M., Bader D.A., Lesk C., Jiang L., Jones B., Zhou L., Chen X., Bi J., Kinney P.L. Impact of climate change on heat-related mortality in Jiangsu Province, China. Environ. Pollut. 2017;224:317–325. doi: 10.1016/j.envpol.2017.02.011. - DOI - PMC - PubMed
    1. Kumar A., Sengar R.S., Pathak R.K., Singh A.K. Integrated Approaches to Develop Drought-Tolerant Rice: Demand of Era for Global Food Security. J. Plant Growth Regul. 2022;42:96–120. doi: 10.1007/s00344-021-10561-6. - DOI
    1. Kater M. New plant breeding technologies for a sustainable agriculture. Ist. Lomb. Accad. Sci. E Lett. Rend. Sci. 2020;151:31–35. doi: 10.4081/scie.2017.640. - DOI
    1. McBratney A.B., Field D.J., Morgan C.L.S., Jarrett L.E. Soil health considerations for global food security. Agron. J. 2021;113:4581–4589.
    1. Castanho R.B., Souto T.S. The Importance of Rice Production in the Construction of Geographical Space: Evolution and Dynamics of Rice Production and the Insertion of New Crops in Ituiutaba (Minas Gerais—MG, Brasil) between 1930 and 2010. Cuad. Geogr. Rev. Colomb. Geogr. 2014;23:93–107. doi: 10.15446/rcdg.v23n1.32465. - DOI

LinkOut - more resources