Identification and recognition of rice diseases and pests using convolutional neural networks

Rahman, Chowdhury Rafeed; Arko, Preetom Saha; Ali, Mohammed Eunus; Khan, Mohammad Ashik Iqbal; Apon, Sajid Hasan; Nowrin, Farzana; Wasif, Abu

doi:10.1016/j.biosystemseng.2020.03.020

Cited by 374 publications

(152 citation statements)

References 13 publications

Supporting

Mentioning

139

Contrasting

Unclassified

Order By: Relevance

“…Для задачи классификации в работе была использована концепция «transfer learning». Если современные нейронные сети используют анализ простейших графических примитивов на нижних слоях, то концепция «transfer learning» зарекомендовала себя уже на многих более сложных задачах [7,16]. Предобученная на большом количестве изображений нейронная сеть дает преимущество в стоимости и скорости обучения.…”

Section: рис 1 размеченные изображения из датасета (пример)unclassified

Определение Заболеваний Маниока Методами Компьютерного Зрения

Терещенко¹,

Perov²,

Osipov³

2021

SJLSA

View full text Add to dashboard Cite

Цель. Разработка модели сверточной нейронной сети для определения заболеваний маниока по фотографии с мобильного телефона. Методы и материалы исследования. Материалом для исследований послужили размеченные изображения с различными видами заболеваний маниока, опубликованные в открытом доступе платформы Kaggle. Методы исследования: теория проектирования и разработки информационных систем, программирования, методы аугментации и расширения датасетов для задач компьютерного зрения, методы настройки гиперпараметров обучения моделей нейронной сети. Результаты. Маниок съедобный – одна из ключевых культур для сельского хозяйства многих регионов мира. Одной из главных причин плохой урожайности является различного вида заболевания. Для профилактики и раннего предупреждения распространения заболевания растений необходим инструмент в виде модели нейронной сети, позволяющей определить наличия заболевания по фотографии со смартфона. В работе использовались методы глубинного обучения сверточных нейронных сетей, а также концепция «transfer learning». На базе сети ResNet 50 была обучена нейронная сеть, позволяющая с точностью 0,93 по метрике F1-score определять наличие заболевания у растения маниок съедобный по изображению. Заключение. Был подготовлен набор данных изображений маниоки, включающий пять классов, для эффективной классификации нейронной сетью. Четыре класса с признаками определенных заболеваний листьев маниоки и один класс для здоровых растений. Была построена и обучена модель для решения задачи классификации по обнаружению болезни листьев маниоки по изображениям со смартфона.

show abstract

Section: рис 1 размеченные изображения из датасета (пример)unclassified

Определение Заболеваний Маниока Методами Компьютерного Зрения

Терещенко¹,

Perov²,

Osipov³

2021

SJLSA

View full text Add to dashboard Cite

show abstract

“…Step 5: The cluster center corresponds to the target firefly, giving the highest fluorescence brightness. Calculate the Euclidean distance of the remaining sample points relative to each cluster center and assign different fluorescence brightness according to formula (14).…”

Section: • Basic Steps Of the Fcm-km Algorithmmentioning

confidence: 99%

“…Under the 10-fold cross-validation strategy, it was found that the average recognition rate of 10 common rice diseases was 95.48%. Literature [14] proposes a new stacked CNN architecture that uses two-stage training to significantly reduce the size of the model while maintaining high classification accuracy. Based on our experimental results, it was shown that compared to VGG16, by using stacked CNN, the test accuracy reached 95% while the model size was reduced by 98%.…”

Section: Introductionmentioning

confidence: 99%

Rapid Detection of Rice Disease Based on FCM-KM and Faster R-CNN Fusion

et al. 2019

View full text Add to dashboard Cite

In this paper, a method for detecting rapid rice disease based on FCM-KM and Faster R-CNN fusion is proposed to address various problems with the rice disease images, such as noise, blurred image edge, large background interference and low detection accuracy. Firstly, the method uses a twodimensional filtering mask combined with a weighted multilevel median filter (2DFM-AMMF) for noise reduction, and uses a faster two-dimensional Otsu threshold segmentation algorithm (Faster 2D-Otsu) to reduce the interference of complex background with the detection of target blade in the image. Then the dynamic population firefly algorithm based on the chaos theory as well as the maximum and minimum distance algorithm is applied for optimization of the K-Means clustering algorithm (FCM-KM) to determine the optimal clustering class k value while addressing the tendency of the algorithm to fall into the local optimum problem. Combined with the R-CNN algorithm for the identification of rice diseases, FCM-KM analysis is conducted to determine the different sizes of the Faster R-CNN target frame. As revealed by the application results of 3010 images, the accuracy and time required for detection of rice blast, bacterial blight and blight were 96.71%/0.65s, 97.53%/0.82s and 98.26%/0.53s, respectively, indicating clearly that the method is more capable of detecting rice diseases and improving the identification accuracy of Faster R-CNN algorithm, while reducing the time required for identification.

show abstract

“…Many researchers have worked on the automatic diagnosis of rice diseases through conventional means such as pattern recognition techniques ( Phadikar & Sil, 2008 ; Rahman et al, 2020 ), support vector machine ( Phadikar, Sil & Das, 2012 ; Prajapati, Shah & Dabhi, 2017 ), digital image processing techniques ( Arnal Barbedo, 2013 ; Zhou et al, 2013 ; Sanyal et al, 2008 ; Sanyal & Patel, 2008 ) and computer vision ( Asfarian et al, 2014 ) for enhancing the accuracy and rapidity of diagnosing the results. In an earlier study, Phadikar & Sil (2008) proposed a rice disease identification approach where the diseased rice images were classified utilizing Self Organizing Map (SOM) (via neural network) in which the train images were obtained by extracting the features of the infected parts of the leave while four different types of images were applied for testing purposes.…”

Section: Introductionmentioning

confidence: 99%

“… Fuentes et al (2017) proposed a deep-learning-based approach using three architectures, namely, Faster Region-based Convolutional Neural Network (Faster R-CNN), Region-based Fully Convolutional Network (R-FCN), and Single Shot Multibox Detector (SSD) that can effectively recognize nine different types of diseases and pests in tomato plants. In a recent study, Rahman et al (2020) developed a CNN approach for detecting diseases and pests (five classes of diseases, three classes of pests and one class of healthy plant and others) from rice plant images. A total number of 1,426 images were collected that were captured using four different types of cameras and the system achieved a mean validation accuracy of 94.33 %.…”

Section: Introductionmentioning

confidence: 99%

A real-time approach of diagnosing rice leaf disease using deep learning-based faster R-CNN framework

Bari

Islam

Rashid

et al. 2021

PeerJ Computer Science

176

View full text Add to dashboard Cite

The rice leaves related diseases often pose threats to the sustainable production of rice affecting many farmers around the world. Early diagnosis and appropriate remedy of the rice leaf infection is crucial in facilitating healthy growth of the rice plants to ensure adequate supply and food security to the rapidly increasing population. Therefore, machine-driven disease diagnosis systems could mitigate the limitations of the conventional methods for leaf disease diagnosis techniques that is often time-consuming, inaccurate, and expensive. Nowadays, computer-assisted rice leaf disease diagnosis systems are becoming very popular. However, several limitations ranging from strong image backgrounds, vague symptoms’ edge, dissimilarity in the image capturing weather, lack of real field rice leaf image data, variation in symptoms from the same infection, multiple infections producing similar symptoms, and lack of efficient real-time system mar the efficacy of the system and its usage. To mitigate the aforesaid problems, a faster region-based convolutional neural network (Faster R-CNN) was employed for the real-time detection of rice leaf diseases in the present research. The Faster R-CNN algorithm introduces advanced RPN architecture that addresses the object location very precisely to generate candidate regions. The robustness of the Faster R-CNN model is enhanced by training the model with publicly available online and own real-field rice leaf datasets. The proposed deep-learning-based approach was observed to be effective in the automatic diagnosis of three discriminative rice leaf diseases including rice blast, brown spot, and hispa with an accuracy of 98.09%, 98.85%, and 99.17% respectively. Moreover, the model was able to identify a healthy rice leaf with an accuracy of 99.25%. The results obtained herein demonstrated that the Faster R-CNN model offers a high-performing rice leaf infection identification system that could diagnose the most common rice diseases more precisely in real-time.

show abstract

Identification and recognition of rice diseases and pests using convolutional neural networks

Cited by 374 publications

References 13 publications

Определение Заболеваний Маниока Методами Компьютерного Зрения

Определение Заболеваний Маниока Методами Компьютерного Зрения

Rapid Detection of Rice Disease Based on FCM-KM and Faster R-CNN Fusion

A real-time approach of diagnosing rice leaf disease using deep learning-based faster R-CNN framework

Contact Info

Product

Resources

About