Classification Of Rhizoma Zingiberaceae Spice Plant Types Using CNN and VGG Methods 19

Abstract

The need for identification of spice plant species is very important to achieve accuracy levels accurately and efficiently. Previous researchers have demonstrated the success of this CNN method in classifying various spice plant species. However, only three species of Zingiberaceae (also known as ginger) spice plants were studied in this research: ginger, turmeric, and galangal. There has not been much previous research on these plant species. To ensure label accuracy, this study compares the performance of two popular CNN optimizers, Adam and SGD. A dataset of spice plant images obtained from Internet websites was then diagnosed by experts. To prepare for training the CNN model with VGG19, the image data is pre-processed. The pre-trained VGG19 architecture is used as the basis for spice plant classification. The classification accuracy is used to evaluate the performance of the model. The results of the study show that in the classification of spice plants, the use of the pre-trained VGG19 architecture is used, providing research results that also show that the architectural CNN method successfully classifies Zingiberaceae spice plant species. Consistently, the use of Adam's optimizer resulted in higher accuracy than SGD. This suggests that Adam's optimizer may be more effective in optimizing VGG19 model parameters for spice plant classification

References

Badan Pusat Statistik (BPS), “Statistik Indonesia-Statistical Yearbook of Indonesia 2023”.

Li J., “Geographic Distribution and Environmenttal Drivers of Global Amphibian Richness,” Diversity and Distribution, vol. 26, no. 7, hlm. 784–796, 2020.

Hariyadi H., “Biodiversity of Plants in Indonesia: Science and Conservation,” Indonesian Journal of Conversation, vol. 3, no. 2, hlm. 67–82, 2019.

FMIPA, “Sebanyak 80 Persen Tanaman Obat Dunia Ada di Indonesia,” IPB University, Bogor, 27 Juli 2020.

Syahril M., “Permasalahan klasifikasi Zingiberaceae di Indonesia: Sebuah tinjauan,” Jurnal Biodiversitas, vol. 21, no. 1, hlm. 1–10, 2020.

Ardiyanti M., Widodo E. S., dan Santosa E., “Integrating Taxonomy, Phylogeny, and Conservation of Zingiberaceae in Indonesia: A Case Study,” Biodiversitas, vol. 20, no. 6, hlm. 1688–1695, 2019.

Kessler P. W. F. B. dan Haynes E. R. S., “The Economic Botany of the Zingiberaceae,” Economy Botany, vol. 73, no. 4, hlm. 375–393, 2019.

Utami N. F., Putri D. W. A., dan Syahril M., “Classification of Zingiberaceae Species in West Java Using Convolutional Neural Network (CNN),” dalam IOP Conference Series: Earth and Environmental Science, IOP Publishing, 2022, hlm. 012058.

Putri D. W. A., Syahril M., dan Utami N. F., “Classification of Zingiberaceae Species in Indonesia Using Convolutional Neural Network (CNN),” dalam Journal of Physics, IOP Publishing, 2023, hlm. 012053.

Wulandari C., Handayani D., dan Supriyadi S., “Klasifikasi Bumbu Dapur dan Rempah Berdasarkan Warna dan Tekstur Menggunakan Convolutional Neural Network (CNN),” Jurnal Sains dan Teknologi Universitas Islam Indonesia, vol. 21, no. 1, hlm. 78–85, 2020.

LeCun Y., Bengio Y., dan Hinton G., “Deep learning for computer vision: A survey,” In arXiv preprint arXiv, hlm. 14–15, 2015.

Lester Mackey, Volodymyr Mnih, dan Geoffrey Hinton, “On the Variance of Stochastic Gradient Descent,” Journal of Machine Learning Research, no. 1, hlm. 1–22, 2019.

Kaiming He, Xiangyu Zhang, Shaoqing Ren, dan Jian Sun, “A Comparative Analysis of Optimization Algorithms for Deep Learning,” IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), vol. 37, no. 5, hlm. 1026–1044, 2016.

Z. Zhang, “Improved Adam Optimizer for Deep Neural Networks,” dalam 2018 IEEE/ACM 26th International Symposium on Quality of Service (IWQoS), IEEE, Jun 2018, hlm. 1–2. doi: 10.1109/IWQoS.2018.8624183.

F. Chollet, “Keras: The Python Deep Learning Library,” Keras.Io, 2015.

Verma V. K., Thakur A., dan Kumar M., “Sign Language Recognition Using Convolution Neural Networkd.,” International Journal Computer Applications, vol. 134, no. 14, hlm. 31–35, 2016.