RICE INSECTS CLASSIFICATION USIING TRANSFER LEARNING AND CNN
0 likes21 views
This document summarizes a study that used transfer learning and convolutional neural networks (CNNs) to classify different rice insect pests from images. The researchers used pre-trained CNN models like AlexNet and VGG16 and fine-tuned them on a dataset of rice insect images. AlexNet achieved the highest classification accuracy of 98%. Transfer learning helped address the classification problem with minimal training requirements compared to training CNNs from scratch. The study aims to help with early detection of insect pests to prevent crop damage.
![© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 295
RICE INSECTS CLASSIFICATION USIING TRANSFER LEARNING AND CNN
Anindita Rath, Rasmita Routray, and Dr.Prof Sribasta Behera
Odisha University of Technology and Research, Bhubaneswar
----------------------------------------------------------------------***------------------------------------------------------------------------
International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 08 | Aug 2022 www.irjet.net p-ISSN: 2395-0072
Index Terms-- CNN, Deep learning, Transfer learning,
VGG16, Alexnet Classification.
I. INTRODUCTION
griculture field is one of the central points that are identified
with social steadiness and monetary improvement.
A
Abstract-- One of the biggest challenges to food security
worldwide is insect pest attacks. Entomology has had
many applications in many biological domains (i.e. insect
counting as a biodiversity index). To meet a growing
biological demand and to compensate a decreasing
workforce amount, automated entomology has been
around for decades. This challenge has been tackled by
computer scientists as well as by biologists themselves.
This thesis investigates the ways to classify different
insect pests using various techniques. Generally these
approaches undergo feature extraction, classification
methods on the tested datasets. Although various
techniques were proposed, transfer learning based
methods are limited in literature which addresses the
aforesaid problem. Presently two transfers learning
based on CNN architectures were performed. The pre
trained CNN models such as Alexnet and VGG16 were
selected for our experiments. From the experimental
results, it is observed that transfer learning can address
this classification with minimal training requirements
and the Alexnet is more effective in comparison to the
VGG16 CNN model in terms of accuracy.
Nonetheless, a few hundred distinct types of insects are
discoveredconnectedwithputawaygrains andtheiritems, and
insects that assault our stores of oat sustenance constitute a
standout amongst the most genuine dangers to our
development. The manual grouping of such creepy crawly
bothers in paddy fields can be tedious and requires generous
specialized ability. The undertaking turns out to be all the
more difficult when bug irritations are to be perceived from
still pictures utilizing a mechanized framework. Pictures of
one bug vermin might be taken from various perspectives,
messed foundation, or may endure change, for example,
revolution, commotion, and soforth.
A. CNN
In deep learning, a convolutional neural network (CNN or
ConvNet) is a class of deep, feed-forward artificial neural
networks, most commonly applied to analyzing visual imagery.
CNNs have emerged from the study of brain’s visual cortex.
These type of deep neural nets have been used in image
recognition since 1980s[9].
A CNN architecture is formed by a stack of distinct layers that
transform the input volume into an output volume (e.g.
holding the class scores) through a differentiable function. A
few distinct types of layers are commonly used and they are
convolutional layer, pooling layer, ReLU activation layer and
fullyconnected layer [2].
Feature learning
It is the first part of the architecture which receives the
image input and extracts important features. These important
features are extracted using convolutional layers. The pooling
layers are used for reducing the spatial dimensionality of the
representation, saving a lot of computational power. and also
reducing the risk of over fitting.
Classification
As the name suggests, in this part the input is the extracted
features from the feature learning part which are used for
training the fully connected layers for classification. The final
fully connected usually outputs the prediction of the image.
B. Alex Deep Neural Network
Alex Net’s design has 60,000 total parameters spread over
eight layers. Three fully linked layers and five convolutional
layers make up these eight layers. Other significant
advancements include the use of multiple GPUs for training
and the use of enhanced versions such as flipping, scaling,
and noising of the pictures for training. Furthermore, the
network employed ReLU (Rectified Linear Unit) activation
functions instead of tanh (hyperbolic tangent), whichever
helped minimize the network’s training time and was a
present solution to the ”vanishing gradient” difficulty at the
time. When constructing the feature map, the pooling layers](https://image.slidesharecdn.com/irjet-v9i849-221028101709-62a15b6f/85/RICE-INSECTS-CLASSIFICATION-USIING-TRANSFER-LEARNING-AND-CNN-1-320.jpg)
![© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 296
Alexnet is observed, and then based on performance metrics,
the Alexnetmodel is found to be more accurate. The model’s
testing is done using performance evaluation metrics such as
overall accuracy, average accuracy, Kappa and loss. The
Alexnet model achieved the highest accuracy of 98%. One of
the main problems faced in a larger neural network is the
vanishing gradient problem.
International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 08 | Aug 2022 www.irjet.net p-ISSN: 2395-0072
IV. REFERENCES
[1] Kamal Sarkar “Rice Leaf Diseases Classification Using
CNN With Transfer Learning”, Jadavpur University 2020
IEEE Calcutta Conference (CALCON).
[2] T. Gupta,”Plant leaf disease analysis using image
processing technique with modified SVM-CS classifier,”
Int. J. Eng. Manag. Technol, no. 5, pp. 11-17, 2017.W.-K.
Chen, Linear Networks and Systems. Belmont, CA:
Wadsworth, 1993, pp. 123–135.
[3] Y. Es-saady,T. El Massi,M. El Yassa,D. Mammass, and A.
Benazoun, ”Automatic recognition of plant leaves
diseases based on serial combination of two SVM
classifiers,” International Conference on Electrical and
Information Technologies (ICEIT) pp. 561-566, 2016.
[4] Bengio, Yoshua. ”Learning deep architectures for AI.”
Foundations and trends® in Machine Learning 2.1 (2009):
1-127
[5] Mohammed, Mohssen, Muhammad Badruddin Khan, and
Eihab Bashier Mohammed Bashier. Machine learning:
algorithms and applications. Crc Press, 2016.
[6] Brownlee, Jason. Deep learning for computer vision:
image classification, object detection, and face
recognition in python. Machine Learning Mastery, 2019.
[7] Carroll, Matthew W., et al. ”Use of spectral vegetation
indices derived from airborne hyperspectral imagery for
detection of European corn borer infes- tation in Iowa
corn plots.” Journal of Economic Entomology 101.5 (2008):
1614-1623.
[8] Dimililer, Kamil, and Salah Zarrouk. ”ICSPI: intelligent
classification sys- tem of pest insects based on image
processing and neural arbitration.” Ap- plied Engineering
in Agriculture 33.4 (2017): 453
[9] Estruch, Juan J., et al. ”Transgenic plants: an emerging
approach to pest control.” Nature biotechnology 15.2
(1997): 137-141.Motorola Semiconductor Data Manual,
Motorola Semiconductor Products Inc., Phoenix, AZ,
1989.
[10] Devaraj, Abirami, et al. ”Identification of plant disease
using image pro- cessing technique.” 2019 International
Conference on Communication and Signal Processing
(ICCSP). IEEE, 2019.](https://image.slidesharecdn.com/irjet-v9i849-221028101709-62a15b6f/85/RICE-INSECTS-CLASSIFICATION-USIING-TRANSFER-LEARNING-AND-CNN-3-320.jpg)
RICE INSECTS CLASSIFICATION USIING TRANSFER LEARNING AND CNN
- 1. © 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 295 RICE INSECTS CLASSIFICATION USIING TRANSFER LEARNING AND CNN Anindita Rath, Rasmita Routray, and Dr.Prof Sribasta Behera Odisha University of Technology and Research, Bhubaneswar ----------------------------------------------------------------------***------------------------------------------------------------------------ International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 09 Issue: 08 | Aug 2022 www.irjet.net p-ISSN: 2395-0072 Index Terms-- CNN, Deep learning, Transfer learning, VGG16, Alexnet Classification. I. INTRODUCTION griculture field is one of the central points that are identified with social steadiness and monetary improvement. A Abstract-- One of the biggest challenges to food security worldwide is insect pest attacks. Entomology has had many applications in many biological domains (i.e. insect counting as a biodiversity index). To meet a growing biological demand and to compensate a decreasing workforce amount, automated entomology has been around for decades. This challenge has been tackled by computer scientists as well as by biologists themselves. This thesis investigates the ways to classify different insect pests using various techniques. Generally these approaches undergo feature extraction, classification methods on the tested datasets. Although various techniques were proposed, transfer learning based methods are limited in literature which addresses the aforesaid problem. Presently two transfers learning based on CNN architectures were performed. The pre trained CNN models such as Alexnet and VGG16 were selected for our experiments. From the experimental results, it is observed that transfer learning can address this classification with minimal training requirements and the Alexnet is more effective in comparison to the VGG16 CNN model in terms of accuracy. Nonetheless, a few hundred distinct types of insects are discoveredconnectedwithputawaygrains andtheiritems, and insects that assault our stores of oat sustenance constitute a standout amongst the most genuine dangers to our development. The manual grouping of such creepy crawly bothers in paddy fields can be tedious and requires generous specialized ability. The undertaking turns out to be all the more difficult when bug irritations are to be perceived from still pictures utilizing a mechanized framework. Pictures of one bug vermin might be taken from various perspectives, messed foundation, or may endure change, for example, revolution, commotion, and soforth. A. CNN In deep learning, a convolutional neural network (CNN or ConvNet) is a class of deep, feed-forward artificial neural networks, most commonly applied to analyzing visual imagery. CNNs have emerged from the study of brain’s visual cortex. These type of deep neural nets have been used in image recognition since 1980s[9]. A CNN architecture is formed by a stack of distinct layers that transform the input volume into an output volume (e.g. holding the class scores) through a differentiable function. A few distinct types of layers are commonly used and they are convolutional layer, pooling layer, ReLU activation layer and fullyconnected layer [2]. Feature learning It is the first part of the architecture which receives the image input and extracts important features. These important features are extracted using convolutional layers. The pooling layers are used for reducing the spatial dimensionality of the representation, saving a lot of computational power. and also reducing the risk of over fitting. Classification As the name suggests, in this part the input is the extracted features from the feature learning part which are used for training the fully connected layers for classification. The final fully connected usually outputs the prediction of the image. B. Alex Deep Neural Network Alex Net’s design has 60,000 total parameters spread over eight layers. Three fully linked layers and five convolutional layers make up these eight layers. Other significant advancements include the use of multiple GPUs for training and the use of enhanced versions such as flipping, scaling, and noising of the pictures for training. Furthermore, the network employed ReLU (Rectified Linear Unit) activation functions instead of tanh (hyperbolic tangent), whichever helped minimize the network’s training time and was a present solution to the ”vanishing gradient” difficulty at the time. When constructing the feature map, the pooling layers
- 2. © 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 295 considerably. C. VGG16 Deep Neural Network The VGG-16 network model is a convolution-based network model that is widely used in computer vision methods, and it performed well in the Image Net 2014 competition. For classification, we deleted the top layers of this model and re- placed them with Flatten, Dense, Drop, and dense-Soft Max layers. To avoid over fitting, the drop layer is utilized to drop certain values at random. The Soft Max layer is used for emotion classifications into many categories. Excepting the last layer, the ReLu activation function has been used in all of the layers. D. Figures and Tables E. Proposed Transfer Learning Transfer learning is a machine learning research issue that concentrated on storing and transferring proficiency learned whileaddressing one issuestoa different but related issues. TABLE I Confusion matrix for binary classification Fig.1 Class specific accuracy for VGG16 and Alexnet transfer learning classifiers International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 09 Issue: 08 | Aug 2022 www.irjet.net p-ISSN: 2395-0072 additionally introduce a stride of 4 pixels, implying that each of the local receptive fields overlapped, lowering their model’s error F. Dataset Description G. Performance Evaluators For validation of the approach, 40% testing and 60% training samples are considered for each class randomly. The observation was replicated over 30 assessments and their mean calculationsareshown. For contrasting the performance of various methods, the commonly used performance measures for multi class data sets like class specific Accuracy, Overall Accuracy (OA), κ, and Average Accuracy (AA) are listed in tables. Descriptions of the measures are presented below. For ease of understanding the binary classification, the confusionmatrix(CM) is presented in Figure 7.2. In the figure, columns represent the original class labels (supplied with the data) as True and False, similarly each row represents the outcome of the classifier. True Positive (TP) and True Negative (TN) are defined as both the original (ground truth) and the obtained (classified) class labels are true and false respectively. While the contradictions are presented as False Positive (FP) and False Negative (FN) which are off-diagonal in the CM. Let, total N samples are tested which is equal to the The IP102 dataset contains more than 75,000 images belongs to 102 categories. A natural long- tailed distribution presents on it. In addition, we annotate 19,000 images with bounding boxes for object detection. The IP102 has a hierarchical taxonomy and the insect pests which mainly affect one specific agricultural product are grouped into the same upper-level category. N = Σ (TP +TN +FP +FN) . So, higher the TP and TN values lead to a better accuracy; on the contrary, higher the FP and FN values reject the classifier. II. ACKNOWLEDGMENT I would also like to extend my gratitude to Dr.(Prof) Sribatsa Behera and Mrs Rasmita Routray who has helped me in gathering the images of rice insects. They constantly comes up with new ideas with her expertise and experience in the field. III.CONCLUSION This study facilitates the early diagnosis of plant diseases to prevent crop loss and the spread of diseases. The CNN model is used to predict different insect pests correctly. The performance of various pre-trained CNN models such as VGG,
- 3. © 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 296 Alexnet is observed, and then based on performance metrics, the Alexnetmodel is found to be more accurate. The model’s testing is done using performance evaluation metrics such as overall accuracy, average accuracy, Kappa and loss. The Alexnet model achieved the highest accuracy of 98%. One of the main problems faced in a larger neural network is the vanishing gradient problem. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 09 Issue: 08 | Aug 2022 www.irjet.net p-ISSN: 2395-0072 IV. REFERENCES [1] Kamal Sarkar “Rice Leaf Diseases Classification Using CNN With Transfer Learning”, Jadavpur University 2020 IEEE Calcutta Conference (CALCON). [2] T. Gupta,”Plant leaf disease analysis using image processing technique with modified SVM-CS classifier,” Int. J. Eng. Manag. Technol, no. 5, pp. 11-17, 2017.W.-K. Chen, Linear Networks and Systems. Belmont, CA: Wadsworth, 1993, pp. 123–135. [3] Y. Es-saady,T. El Massi,M. El Yassa,D. Mammass, and A. Benazoun, ”Automatic recognition of plant leaves diseases based on serial combination of two SVM classifiers,” International Conference on Electrical and Information Technologies (ICEIT) pp. 561-566, 2016. [4] Bengio, Yoshua. ”Learning deep architectures for AI.” Foundations and trends® in Machine Learning 2.1 (2009): 1-127 [5] Mohammed, Mohssen, Muhammad Badruddin Khan, and Eihab Bashier Mohammed Bashier. Machine learning: algorithms and applications. Crc Press, 2016. [6] Brownlee, Jason. Deep learning for computer vision: image classification, object detection, and face recognition in python. Machine Learning Mastery, 2019. [7] Carroll, Matthew W., et al. ”Use of spectral vegetation indices derived from airborne hyperspectral imagery for detection of European corn borer infes- tation in Iowa corn plots.” Journal of Economic Entomology 101.5 (2008): 1614-1623. [8] Dimililer, Kamil, and Salah Zarrouk. ”ICSPI: intelligent classification sys- tem of pest insects based on image processing and neural arbitration.” Ap- plied Engineering in Agriculture 33.4 (2017): 453 [9] Estruch, Juan J., et al. ”Transgenic plants: an emerging approach to pest control.” Nature biotechnology 15.2 (1997): 137-141.Motorola Semiconductor Data Manual, Motorola Semiconductor Products Inc., Phoenix, AZ, 1989. [10] Devaraj, Abirami, et al. ”Identification of plant disease using image pro- cessing technique.” 2019 International Conference on Communication and Signal Processing (ICCSP). IEEE, 2019.