Rice grain disease identification using dual phase convolutional neural network based system aimed at small dataset.

Abstract

Although Convolutional neural networks (CNNs) are widely used for plant disease detection, they require a large number of training samples while dealing with wide variety of heterogeneous background. In this paper, a CNN based dual phase method has been proposed which can work effectively on small rice grain disease dataset with heterogeneity. At the first phase, Faster RCNN method is applied for cropping out the significant portion (rice grain) from an image. This initial phase results in a secondary dataset of rice grains devoid of heterogeneous background. Disease classification is performed on such derived and simplified samples using CNN architecture. Comparison of the dual phase approach with straight forward application of CNN on the small grain dataset shows the effectiveness of the proposed method which provides a 5 fold cross validation accuracy of 88.92%.

Design of Accurate Multi-class Optimized Lightweight Convolution Neural Network for Rice Varieties Classification

“…It is 5.6 times greater than the proposed OpLW-CNN model. The overall classi cation metrics of the proposed OpLW-CNN model achieve more accuracy of 2.6-13.4% [10,11,13,14] and reduces the computation resources of 58-82% compared to benchmark CNN models [10][11][12][13][14].…”

“…Therefore, the proposed model is outperforming the classi cation metrics by 6% and 13% in terms of accuracy. [10] adopted the three CNN models such as VGG-16, VGG-19, and ResNet50. From these models, [10] claims that VGG-19 is considered to be better with an accuracy of 88%.…”

“…[10] adopted the three CNN models such as VGG-16, VGG-19, and ResNet50. From these models, [10] claims that VGG-19 is considered to be better with an accuracy of 88%. The computation of the VGG-19 is estimated and attains to be 144M.…”

“…In recent years, Machine learning (ML) and Deep Learning (DL) approaches have been effective in the accurate and quick evaluation of a massive volume of data. Using these techniques an automated, cost-effective, e cient, and non-destructive classi cation of rice varieties is being proposed [10][11][12][13][14]. This increases product quality and adheres to food safety standards [6].…”

Convolutional neural network in rice disease recognition: accuracy, speed and lightweight

“…Rice disease recognition based on convolutional neural networks and images requires a large amount of data to train the model and validate its performance. Different types of rice diseases may behave similarly on images, and a large number of images must be available to ensure the effectiveness of recognizing different diseases (Ahmed et al, 2023). The convolutional neural network, as a deep learning model, has a complex structure and contains many parameters, so the demand for data during training is relatively large (Aggarwal et al, 2023;Dogra et al, 2023).…”