# Plant Recognition Using Morphological Feature Extraction and Transfer Learning over SVM and AdaBoost

^{1}

^{2}

^{3}

^{*}

## Abstract

**:**

## 1. Introduction

- Proposing a robust, precise, and fast plant-species recognition model;
- Making use of morphology-based features from leaf images with low dimensionality;
- Using features by transfer learning from low-dimensional ConvNet architecture;
- Evaluation of different classifiers using controlled comparison;
- Enhancing the classification results via adaptive boosting.

#### 1.1. Literature Review

## 2. Proposed Methodology

#### 2.1. Support Vector Machine (SVM)

#### 2.2. Adaptive Boosting (AdaBoost)

#### 2.3. Feature Extraction and Selection

#### 2.4. Model Highlights

## 3. Data and Training

## 4. Evaluation and Results

## 5. Credibility of the Methodology

## 6. Conclusions

## Author Contributions

## Funding

## Institutional Review Board Statement

## Informed Consent Statement

## Data Availability Statement

## Acknowledgments

## Conflicts of Interest

## References

- Sun, R.; Zhang, M.; Yang, K.; Liu, J. Data enhancement for plant disease classification using generated lesions. Appl. Sci.
**2020**, 10, 466. [Google Scholar] - Yang, K.; Zhong, W.; Li, F. Leaf segmentation and classification with a complicated background using deep learning. Agronomy
**2020**, 10, 1721. [Google Scholar] [CrossRef] - Zhang, Y.; Cui, J.; Wang, Z.; Kang, J.; Min, Y. Leaf Image Recognition Based on Bag of Features. Appl. Sci.
**2020**, 10, 5177. [Google Scholar] [CrossRef] - Azlah, M.A.; Chua, L.S.; Rahmad, F.R.; Abdullah, F.I.; Wan Alwi, S.R. Review on techniques for plant leaf classification and recognition. Computers
**2019**, 8, 77. [Google Scholar] [CrossRef] [Green Version] - Kumar, M.; Gupta, S.; Gao, X.Z.; Singh, A. Plant Species Recognition Using Morphological Features and Adaptive Boosting Methodology. IEEE Access
**2019**, 7, 163912–163918. [Google Scholar] [CrossRef] - Jeon, W.S.; Rhee, S.Y. Plant leaf recognition using a convolution neural network. Int. J. Fuzzy Log. Intell. Syst.
**2017**, 17, 26–34. [Google Scholar] - Pankaja, K.; Suma, V. Leaf Recognition and Classification Using GLCM and Hierarchical Centroid Based Technique. In Proceedings of the 2018 International Conference on Inventive Research in Computing Applications (ICIRCA), Coimbatore, India, 11–12 July 2018; IEEE: Piscataway, NJ, USA; pp. 1190–1194. [Google Scholar]
- Kumar, S. Leaf Color, Area and Edge features-based approach for Identification of Indian Medicinal Plants. Int. J. Comput. Sci. Eng.
**2012**, 3, 436–442. [Google Scholar] - Kadir, A.; Nugroho, L.E.; Susanto, A.; Santosa, P.I. Performance improvement of leaf identification system using principal component analysis. Int. J. Adv. Sci. Technol.
**2012**, 44, 113–124. [Google Scholar] - Priya, C.A.; Balasaravanan, T.; Thanamani, A.S. An efficient leaf recognition algorithm for plant classification using support vector machine. In Proceedings of the InInternational conference on pattern recognition, informatics and medical engineering (PRIME-2012), Salem, India, 21–23 March 2012; IEEE: Piscataway, NJ, USA; pp. 428–432. [Google Scholar]
- Anami, B.S.; Nandyal, S.S.; Govardhan, A. A combined color, texture and edge features-based approach for identification and classification of indian medicinal plants. Int. J. Comput. Appl.
**2010**, 6, 45–51. [Google Scholar] [CrossRef] - Sambhaji, E.S.; Andore, D.B. Leaf recognition algorithm using neural network-based image processing. Asian J. Eng. Technol. Innov.
**2014**, 2, 10–16. [Google Scholar] - Sun, Y.; Liu, Y.; Wang, G.; Zhang, H. Deep learning for plant identification in natural environment. Comput. Intell. Neurosci.
**2017**, 22, 2017. [Google Scholar] [CrossRef] [PubMed] - Cortes, C.; Vapnik, V. Support-vector networks. Mach. Learn.
**1995**, 20, 273–297. [Google Scholar] [CrossRef] - Duan, K.B.; Keerthi, S.S. Which is the best multiclass SVM method? An empirical study. In International Workshop on Multiple Classifier Systems, Proceedings of the Multiple Classifier Systems; Springer: Berlin/Heidelberg, Germany, 2005; pp. 278–285. [Google Scholar]
- Hastie, T.; Tibshirani, R. Classification by pairwise coupling. In Advances in Neural Information Processing Systems; MIT Press: Cambridge, MA, USA, 1998; pp. 507–513. [Google Scholar]
- Freund, Y.; Schapire, R.E. A decision-theoretic generalization of on-line learning and an application to boosting. J. Comput. Syst. Sci.
**1997**, 55, 119–139. [Google Scholar] [CrossRef] [Green Version] - Boosting Algorithms: AdaBoost, Gradient Boosting and XGBoost. 5 May 2018. Available online: hackernoon.com (accessed on 4 January 2020).
- Hastie, T.; Rosset, S.; Zhu, J.; Zou, H. Multi-class adaboost. Stat. Its Interface
**2009**, 2, 349–360. [Google Scholar] [CrossRef] [Green Version] - Kim, T.H.; Park, D.C.; Woo, D.M.; Jeong, T.; Min, S.Y. Multi-class classifier-based adaboost algorithm. In International Conference on Intelligent Science and Intelligent Data Engineering, Proceedings of the Intelligent Science and Intelligent Data Engineering; Springer: Berlin/Heidelberg, Germany, 2011; pp. 122–127. [Google Scholar]
- CS231n: Convolutional Neural Networks for Visual Recognition. Available online: http://cs231n.github.io/transfer-learning/ (accessed on 10 January 2021).
- Yosinski, J.; Clune, J.; Bengio, Y.; Lipson, H. How transferable are features in deep neural networks? arXiv
**2014**, arXiv:1411.1792. [Google Scholar] - Plant Leaf Recognition. Available online: http://cs229.stanford.edu/proj2016/report/LiuHuang-PlantLeafRecognition-report.pdf (accessed on 5 January 2021).
- He, K.; Zhang, X.; Ren, S.; Sun, J. Deep residual learning for image recognition. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Boston, MA, USA, 7–12 June 2015. [Google Scholar]
- Li, X.; Wang, L.; Sung, E. AdaBoost with SVM-based component classifiers. Eng. Appl. Artif. Intell.
**2008**, 21, 785–795. [Google Scholar] [CrossRef] [Green Version] - Govindaraj, D. Can Boosting with SVM as Week Learners Help? arXiv
**2016**, preprint. arXiv:1604.05242. [Google Scholar] - García, E.; Lozano, F. Boosting Support Vector Machines. MLDM Posters. 2007, pp. 153–167. Available online: https://www.sciencedirect.com/science/article/abs/pii/S0957417420301457 (accessed on 10 January 2020).
- Flavia (at a Glance). Available online: http://flavia.sourceforge.net/ (accessed on 2 January 2021).
- Sangeetha, R.; Kalpana, D.B. Identifying efficient kernel function in multiclass support vector machines. Int. J. Comput. Appl.
**2011**, 28, 18–23. [Google Scholar] [CrossRef]

**Figure 1.**Representations of two AdaBoost Methods as algorithms (

**a**) Shown algorithm of AdaBoost, (

**b**) Shows Multiclass AdaBoost [19].

**Figure 2.**Shows code snippet implemented on MATLAB for the processing of images before feeding into the model.

**Figure 3.**Leaf sample image in different stages of pre-processing. (

**a**) Original RGB Image (Size: [1200, 1600]). (

**b**) Grayscale Converted Image (Size: [1200, 1600]). (

**c**) Binary Converted and Resized Image (Size: [256, 256]). (

**d**) Centroid location of the leaf sample

**Figure 4.**Shows samples from the FLAVIA leaf image dataset (

**a**) Image of species, label Podocarpus macrophyllus (Thunb.) Sweet. (

**b**) Image of species, label Prunus serrulata Lindl. var. lannesiana auct. (

**c**) Image of species, label Cinnamomum japonicum Sieb. (

**d**) Image species, label Kalopanax s. Koidz.

**Figure 8.**Model’s output samples of four different species of leaves. (

**a**) Output of Podocarpus macrophyllus (Thunb.) Sweet class, (

**b**) Output of Prunus s. Lindl var. l. auct., (

**c**) Cinnamomum japonicum Sieb. (

**d**) Kalopanax s. Koidz.

**Figure 9.**All four models’ performance as root mean squared error (RMSE) vs partitioning strategy of the proposed systems.

Feature | Explanation | Formula |
---|---|---|

Solidity | Area fraction of the region compared to its convex hull, or the extent of the pixels in the convex hull that is additionally in the area. | $S=\frac{Area}{ConvexArea}$ |

Centroid | Center of mass of a region or area. | ${C}_{x}=\frac{\sum {C}_{ix}{A}_{i}}{{A}_{i}};{C}_{y}=\frac{\sum {C}_{iy}{A}_{i}}{{A}_{i}}$ $A=\frac{1}{2}{\displaystyle {\displaystyle \sum}_{i=1}^{N-1}}{x}_{i}{y}_{i+1}+{x}_{i+1}{y}_{i}$ |

Perimeter | Path length of the shape externally. | $P=2L+2P$ |

Major axis length | Line connecting one end, called a base point, to the tip of the leaf. | $MajorAxisLength=\frac{{\left({x}_{1}-{x}_{c}\right)}^{2}}{r{x}^{2}}+\frac{{\left({y}_{1}-{y}_{c}\right)}^{2}}{r{y}^{2}}$ |

Minor axis length | Line drawn perpendicular to the major axis. | ${\left(\u0131\left(\frac{{\left({x}_{1}-{x}_{c}\right)}^{2}}{r{x}^{2}}+\frac{{\left({y}_{1}-{y}_{c}\right)}^{2}}{r{y}^{2}}\right)\right)}^{\perp}$ |

Orientation | Angle between the x-axis and the major axis of the ellipse. | $O=\mathrm{cos}{m}_{j}+\mathrm{sin}{m}_{j}+\sqrt{\mathrm{tan}\left({m}_{j}\times {m}_{i}\right)}$ |

ConvNet via transfer learning | Features extracted from ResNet50 via transfer learning | $\mathrm{N}/\mathrm{A}$ |

Element | Value/Detail |
---|---|

Fundamental algorithm | Support vector machine (polynomial kernel) |

Secondary algorithm | Multiclass adaptive boosting |

Dataset | FLAVIA (32 classes distributed among 1907 instances) |

Features | Morphological and from transfer learning (ResNet50) |

Partitioning strategy | 80:20 split, five-fold CV and 10-fold CV (best: 10-fold CV) |

Fundamental model | Support vector machine and adaptive boosting |

Serial | Research Study | Accuracy (%) |
---|---|---|

1 | CNN-based [6] | 94.00 |

2 | CNN-based [13] | 91.78 |

3 | Bag of features-based [3] | 94.22 |

4 | Our study (hybrid) | 94.72 |

Serial | Class | Precision | Class | Precision |
---|---|---|---|---|

1 | Phyllostachys edulis (Carr.) Houz. | 1.0000 | Cedrus deodara (Roxb.) G. Don | 0.93750 |

2 | Aesculus chinensis | 0.9000 | Ginkgo biloba L. | 0.92860 |

3 | Berberis anhweiensis Ahrendt | 1.0000 | Lagerstroemia indica (L.) Pers. | 0.91660 |

4 | Cercis chinensis | 0.9090 | Nerium oleander L. | 0.92860 |

5 | Indigofera tinctoria L. | 0.9000 | Podocarpus macrophyllus (Thunb.) Sweet | 0.92300 |

6 | Acer palmatum | 1.0000 | Prunus s. Lindl var. l. auct. | 0.92857 |

7 | Phoebe nanmu (Oliv.) Gamble | 1.0000 | Ligustrum lucidum Ait. f. | 1.00000 |

8 | Kalopanax s. Koidz. | 1.0000 | Tonna sinensis M. Roem. | 1.00000 |

9 | Cinnamomum japonicum Sieb. | 1.0000 | Prunus persica (L.) Batsch | 1.000000 |

10 | Koelreuteria paniculata Laxm. | 0.9166 | Manglietia fordiana Oliv. | 0.90000 |

11 | Ilex macrocarpa Oliv. | 0.9166 | Acer buergerianum Miq. | 1.00000 |

12 | Pittosporum tobira Ait. f. | 1.0000 | Mahonia bealei (Fortune) Carr. | 0.90900 |

13 | Chimonanthus praecox L. | 0.9090 | Magnolia grandiflora L. | 1.00000 |

14 | Cinnamomum camphora (L.) J. Presl | 0.9231 | Populus× canadensis Moench | 0.92307 |

15 | Viburnum awabuki K. Koch | 1.0000 | Liriodendron chinense Sarg. | 1.00000 |

16 | Osmanthus fragrans Lour. | 1.0000 | Citrus reticulata Blanco | 1.00000 |

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |

© 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Mahajan, S.; Raina, A.; Gao, X.-Z.; Kant Pandit, A.
Plant Recognition Using Morphological Feature Extraction and Transfer Learning over SVM and AdaBoost. *Symmetry* **2021**, *13*, 356.
https://doi.org/10.3390/sym13020356

**AMA Style**

Mahajan S, Raina A, Gao X-Z, Kant Pandit A.
Plant Recognition Using Morphological Feature Extraction and Transfer Learning over SVM and AdaBoost. *Symmetry*. 2021; 13(2):356.
https://doi.org/10.3390/sym13020356

**Chicago/Turabian Style**

Mahajan, Shubham, Akshay Raina, Xiao-Zhi Gao, and Amit Kant Pandit.
2021. "Plant Recognition Using Morphological Feature Extraction and Transfer Learning over SVM and AdaBoost" *Symmetry* 13, no. 2: 356.
https://doi.org/10.3390/sym13020356