Open Access. Powered by Scholars. Published by Universities.®
Articles 1 - 2 of 2
Full-Text Articles in Engineering
A Study On Image Processing Techniques And Deep Learning Techniques For Insect Identification, Vinita Abhishek Gupta, M.V. Padmavati, Ravi R. Saxena, Pawan Kumar Patnaik, Raunak Kumar Tamrakar
A Study On Image Processing Techniques And Deep Learning Techniques For Insect Identification, Vinita Abhishek Gupta, M.V. Padmavati, Ravi R. Saxena, Pawan Kumar Patnaik, Raunak Kumar Tamrakar
Karbala International Journal of Modern Science
Automatic identification of insects and diseases has attracted researchers for the last few years. Researchers have suggested several algorithms to get around the problems of manually identifying insects and pests. Image processing techniques and deep convolution neural networks can overcome the challenges of manual insect identification and classification. This work focused on optimizing and assessing deep convolutional neural networks for insect identification. AlexNet, MobileNetv2, ResNet-50, ResNet-101, GoogleNet, InceptionV3, SqueezeNet, ShuffleNet, DenseNet201, VGG-16 and VGG-19 are the architectures evaluated on three different datasets. In our experiments, DenseNet 201 performed well with the highest test accuracy. Regarding training time, AlexNet performed well, …
A Novel Insect And Pest Identification Model Based On A Weighted Multipath Convolutional Neural Network And Generative Adversarial Network, Vinita Abhishek Gupta, M.V. Padmavati, Ravi R. Saxena, Raunak Kumar Tamrakar
A Novel Insect And Pest Identification Model Based On A Weighted Multipath Convolutional Neural Network And Generative Adversarial Network, Vinita Abhishek Gupta, M.V. Padmavati, Ravi R. Saxena, Raunak Kumar Tamrakar
Karbala International Journal of Modern Science
Timely identification of insects and their management play a significant role in sustainable agriculture development. The proposed hybrid model integrates a weighted multipath convolutional neural network and generative adversarial network to identify insects efficiently. To address the shortcomings of single-path networks, this novel model takes input from numerous iterations of the same image to learn more specific features. To avoid redundancy produced due to multipath, weights have been assigned to each path. For Xie2 dataset, the model shows 3.75%, 2.74%, 1.54%, 1.76%, 1.76%, 2.74 %, and 2.14% performance improvement from AlexNet, ResNet50, ResNet101, GoogleNet, VGG-16, VGG-19, and simple CNN respectively. …