ABSTRACT
Pests are a significant challenge in paddy cultivation, resulting in a global loss of approximately 20 % of rice yield. Early detection of paddy insects can help to save these potential losses. Several ways have been suggested for identifying and categorizing insects in paddy fields, employing a range of advanced, noninvasive, and portable technologies. However, none of these systems have successfully incorporated feature optimization techniques with Deep Learning and Machine Learning. Hence, the current research provided a framework utilizing these techniques to detect and categorize images of paddy insects promptly. Initially, the suggested research will gather the image dataset and categorize it into two groups: one without paddy insects and the other with paddy insects. Furthermore, various pre-processing techniques, such as augmentation and image filtering, will be applied to enhance the quality of the dataset and eliminate any unwanted noise. To determine and analyze the deep characteristics of an image, the suggested architecture will incorporate 5 pre-trained Convolutional Neural Network models. Following that, feature selection techniques, including Principal Component Analysis (PCA), Recursive Feature Elimination (RFE), Linear Discriminant Analysis (LDA), and an optimization algorithm called Lion Optimization, were utilized in order to further reduce the redundant number of features that were collected for the study. Subsequently, the process of identifying the paddy insects will be carried out by employing 7 ML algorithms. Finally, a set of experimental data analysis has been conducted to achieve the objectives, and the proposed approach demonstrates that the extracted feature vectors of ResNet50 with Logistic Regression and PCA have achieved the highest accuracy, precisely 99.28 %. However, the present idea will significantly impact how paddy insects are diagnosed in the field.
ABSTRACT
The "BDWaste" dataset contains two significant categories of waste, namely digestible and indigestible, in Bangladesh. Each category represents 10 distinct species of waste. The digestible categories are sugarcane husk, fish ash, potato peel, paper, mango peel, rice, shell of malta, lemon peel, banana peel, and egg shell. On the other hand, the indigestible species are polythene, cans, plastic, glass, wire, gloves, empty medicine packets, chip packets, bottles, and masks. The research uploaded the primarily collected dataset on Mendeley, and the dataset contains a total of 2497 raw images, of which 1234 were digestible and 1263 belonged to indigestible species. Each species is stored in a fixed file based on its name and categories. Also, each species contains an indoor (with a visible surface) and an outdoor (with a surface that can be seen generally) image. The dataset is free from any blurry, dark, noisy, or invisible images. The research also performed waste classification with pre-trained convolutional neural network models such as MobileNetV2 and InceptionV3. The research found the highest accuracy of 96.70% in the indigestible waste classification and 99.70% in the digestible waste classification. The researchers presume that this data can be used in the future in different types of research, such as sustainable development, sustainable environments, agricultural development, recycling processes, and other computer vision-based applications.
ABSTRACT
COVID-19, caused by SARS-CoV-2, has been declared as a global pandemic by WHO. Early diagnosis of COVID-19 patients may reduce the impact of coronavirus using modern computational methods like deep learning. Various deep learning models based on CT and chest X-ray images are studied and compared in this study as an alternative solution to reverse transcription-polymerase chain reactions. This study consists of three stages: planning, conduction, and analysis/reporting. In the conduction stage, inclusion and exclusion criteria are applied to the literature searching and identification. Then, we have implemented quality assessment rules, where over 75 scored articles in the literature were included. Finally, in the analysis/reporting stage, all the papers are reviewed and analysed. After the quality assessment of the individual papers, this study adopted 57 articles for the systematic literature review. From these reviews, the critical analysis of each paper, including the represented matrix for the model evaluation, existing contributions, and motivation, has been tracked with suitable illustrations. We have also interpreted several insights of each paper with appropriate annotation. Further, a set of comparisons has been enumerated with suitable discussion. Convolutional neural networks are the most commonly used deep learning architecture for COVID-19 disease classification and identification from X-ray and CT images. Various prior studies did not include data from a hospital setting nor did they consider data preprocessing before training a deep learning model.
