Leaf Identification Using HOG, KNN, and Neural Networks

Sharma, Prerna; Aggarwal, Aastha; Gupta, Apoorva; Garg, Akshit

doi:10.1007/978-981-13-2354-6_10

Cited by 15 publications

(9 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…• K-nearest neighbours (KNN), support vector machine (SVM), random forest and convolution neural network (CNN; Sharma, Aggarwal, Gupta, & Garg, 2018) have been used, and their results have been compared with the ones obtained by IAPO.…”

mentioning

confidence: 99%

Artificial plant optimization algorithm to detect infected leaves using machine learning

et al. 2020

Self Cite

View full text Add to dashboard Cite

Plant leaves play an important role in the diagnosis of plant diseases. Losses from such diseases can have a significant economic as well as environmental impact. Thus, examination of leaves into a healthy or infected carries substantial importance. An improved artificial plant optimization (IAPO) algorithm using machine learning has been introduced that identifies the plant diseases and categorize the leaves into healthy and infected on a private dataset of 236 images. Features are extracted from the images using histogram of oriented gradients (descriptor). The concepts of artificial plant optimization are then applied to study the features of healthy leaves using IAPO. A machine learning algorithm has been created to make the model adaptive with varied datasets. The degree of infection is eventually computed, and the leaves with infection greater than a certain calculated threshold are classified as infected leaves. The results show that IAPO can be used for classification of infected and healthy leaves and this algorithm can be generalized to solve problems in other domains as well. The proposed IAPO is also compared with other classification algorithms including k-nearest neighbours, support vector machine, random forest and convolution neural network that show accuracies of 78.24%, 83.48%, 87.83%, and 91.26%, respectively, whereas IAPO shows quite accurate results in classification of leaves with an accuracy of 97.45% on training set and 95.0% accuracy on test set. K E Y W O R D Sconvolutional neural network, histogram of oriented gradients, improved artificial plant optimization, machine learning 1 | INTRODUCTION Data explosion is observed everywhere these days as data created in terms of volume in the past few years are colossal than in the history of human race. Thus, it has become important to find optimal solutions that can help in managing these large amounts of data. Intelligent bio-inspired algorithms are introduced to deal with incompleteness and imperfectness of dynamic problems (Fister, Brest, Fister, & Yang, 2015). These algorithms are called intelligent as they can acclimatize and behave like real world organisms. But most of the bio-inspired algorithms are getting ignored due to rigorous development in this field.The following are some important and prevalent pieces of work done in the field of bio-inspired algorithms. Artificial plant optimization algorithm (APOA) was used by Ziqiang, Zhihua and Jianchao (Zhao, Cui, Zeng, & Yue, 2011) to solve constrained optimization problem in which they proposed a shrinkage coefficient for checking the feasible region of a particle.

show abstract

mentioning

confidence: 99%

Artificial plant optimization algorithm to detect infected leaves using machine learning

et al. 2020

Self Cite

View full text Add to dashboard Cite

show abstract

“…As its name, this algorithm uses information regarding an example's k-nearest neighbors to classify unlabeled examples and assign one of them to the most common class among the k-nearest neighbors [70]. KNN has been widely used in statistical applications and it has been one of the most successful supervised classification algorithms, especially for the task of multi-class classification [31,71].…”

Section: Discussionmentioning

confidence: 99%

Discrimination of Gentiana and Its Related Species Using IR Spectroscopy Combined with Feature Selection and Stacked Generalization

Shen

Wang

2020

Molecules

View full text Add to dashboard Cite

Gentiana, which is one of the largest genera of Gentianoideae, most of which had potential pharmaceutical value, and applied to local traditional medical treatment. Because of the phytochemical diversity and difference of bioactive compounds among species, which makes it crucial to accurately identify authentic Gentiana species. In this paper, the feasibility of using the infrared spectroscopy technique combined with chemometrics analysis to identify Gentiana and its related species was studied. A total of 180 batches of raw spectral fingerprints were obtained from 18 species of Gentiana and Tripterospermum by near-infrared (NIR: 10,000–4000 cm−1) and Fourier transform mid-infrared (MIR: 4000–600 cm−1) spectrum. Firstly, principal component analysis (PCA) was utilized to explore the natural grouping of the 180 samples. Secondly, random forests (RF), support vector machine (SVM), and K-nearest neighbors (KNN) models were built while using full spectra (including 1487 NIR variables and 1214 FT-MIR variables, respectively). The MIR-SVM model had a higher classification accuracy rate than the other models that were based on the results of the calibration sets and prediction sets. The five feature selection strategies, VIP (variable importance in the projection), Boruta, GARF (genetic algorithm combined with random forest), GASVM (genetic algorithm combined with support vector machine), and Venn diagram calculation, were used to reduce the dimensions of the data variable in order to further reduce numbers of variables for modeling. Finally, 101 NIR and 73 FT-MIR bands were selected as the feature variables, respectively. Thirdly, stacking models were built based on the optimal spectral dataset. Most of the stacking models performed better than the full spectra-based models. RF and SVM (as base learners), combined with the SVM meta-classifier, was the optimal stacked generalization strategy. For the SG-Ven-MIR-SVM model, the accuracy (ACC) of the calibration set and validation set were both 100%. Sensitivity (SE), specificity (SP), efficiency (EFF), Matthews correlation coefficient (MCC), and Cohen’s kappa coefficient (K) were all 1, which showed that the model had the optimal authenticity identification performance. Those parameters indicated that stacked generalization combined with feature selection is probably an important technique for improving the classification model predictive accuracy and avoid overfitting. The study result can provide a valuable reference for the safety and effectiveness of the clinical application of medicinal Gentiana.

show abstract

“…Penelitian [3] melakukan penelitian menggunakan pendekatan KNN untuk mendeteksi dan mengklasifikasikan berbagai penyakit yang ada pada daun tanaman berdasarkan fitur tekstur. Dalam penelitian [4] yang dilakukan oleh Sharma dkk mengenai identifikasi daun dengan ekstraksi fitur HOG dengan membandingkan algoritma klasifikasi KNN dan Neural Network. R.A. Saputra dalam penelitiannya [5] membahas mengenai klasifikasi penyakit daun padi.…”

Section: Pendahuluanunclassified

Klasifikasi Daun Herbal Berdasarkan Fitur Bentuk dan Tekstur Menggunakan KNN

Meiriyama¹,

Devella²,

Adelfi³

2022

JATISI

View full text Add to dashboard Cite

Indonesia memiliki kekayaan dan keanekaragaman hayati yang sangat melimpah. Dari total sekitar 40.000 jenis tanaman herbal yang dikenal di dunia, terdapat kurang lebih 30.000 tanaman herbal disinyalir berada di Indonesia. Tanaman herbal merupakan tanaman yang umum digunakan masyarakat, khususnya di Indonesia yang memiliki keanekaragaman hayati sebagai bahan pembuatan obat herbal. Tanaman herbal tentu tidak mudah untuk dikenali meskipun sering tumbuh di sekitar lingkungan. Karena masih kurangnya pengetahuan masyarakat tentang tanaman herbal, maka tidak memungkinkan untuk memanfaatkan tanaman herbal tersebut. Penelitian ini bertujuan untuk mengklasifikasikan daun tanaman herbal menggunakan metode K-Nearest Neighbor (KNN) dengan nilai K = 3 dan ekstraksi ciri Histogram of Oriented Gradient (HOG) dan Local Binary Patterns (LBP). Penelitian dilakukan pada 15 jenis tanaman herbal. Akurasi metode HOG dengan KNN adalah 92,67%, Akurasi LBP dengan KNN adalah 88,67% dan akurasi kombinasi fitur HOG dan LBP dengan metode KNN adalah 92,67%. Berdasarkan ketiga skenario percobaan yang telah dilakukan menunjukkan bahwa kombinasi fitur HOG dan LBP tidak mempengaruhi akurasi klasifikasi daun tanaman herbal.

show abstract

Leaf Identification Using HOG, KNN, and Neural Networks

Cited by 15 publications

References 11 publications

Artificial plant optimization algorithm to detect infected leaves using machine learning

Artificial plant optimization algorithm to detect infected leaves using machine learning

Discrimination of Gentiana and Its Related Species Using IR Spectroscopy Combined with Feature Selection and Stacked Generalization

Klasifikasi Daun Herbal Berdasarkan Fitur Bentuk dan Tekstur Menggunakan KNN

Contact Info

Product

Resources

About