Browse Items (11810 total)
Sort by:
-
Advancements in Deep Learning Techniques for Potato Leaf Disease Identification Using SAM-CNNet Classification
Potato leaf diseases like Late Blight and Early Blight significantly challenge potato cultivation, impacting crop yield and quality worldwide. Potatoes are a staple for over a billion people and crucial for food security, especially in developing countries. The economic impact is substantial, with Late Blight alone causing annual damages over $6 billion globally. Effective detection and management are essential to mitigate these effects on agricultural productivity and economic stability. This paper presents a novel approach to potato leaf disease detection using advanced deep learning and optimization techniques. Key components include data normalization to eliminate noise, feature extraction using GoogLeNet, and hyperparameter tuning through the Elk Herd Optimizer (EHO). Additionally, a Spatial Attention Mechanism and Convolutional Neural Network (SAM-CNNet) are employed for robust classification. The method is validated using the Plant Village dataset, yielding an accuracy of 98.58%, with precision of 97.68%, recall of 98.42%, and F1-Score of 98.21%, demonstrating exceptional performance and reliability. This study highlights the proposed approach's efficacy in accurately identifying and classifying potato leaf diseases, offering a promising solution for precision agriculture and crop management. Copyright: 2024 The authors. This article is published by IIETA and is licensed under the CC BY 4.0 license. -
Dandelion Algorithm for Optimal Location and Sizing of Battery Energy Storage Systemsin Electrical Distribution Networks
This paper describes a new way to improve the performance of an EDN by integrating distributed battery energy storage systems (BESs) in the best way possible. This method is based on the Dandelion Algorithm (DA). The search space for BES locations is first predetermined using loss sensitivity factors (LSFs), and then DA is used to determine the optimal locations and sizes. The reduction of real power distribution loss is regarded as the primary objective function, and the impact of BESs is extended to examine the network voltage profile, voltage stability, and GHG emissions. IEEE 33-busEDN is used to calculate the computational efficiency of LSF-DA. Results show that DA is more efficient than Archimedes optimization (AOA), future search algorithm(FSA), pathfinder algorithm(PFA), and butterfly optimization algorithm(BOA) algorithms. Furthermore, the results show that the proposed DA enhances all technological and environmental factors and RDN performance. The Author(s), under exclusive license to Springer Nature Switzerland AG 2024. -
Liberalisation and cashew industry: evidence from India (1965 to 2018)
We examine the impact of liberalisation on production, import, export and area under cultivation of cashew industry in India during 1965 to 2018 period using regression method. We divide data into two sub-periods. The liberalisation and pre-liberalisation period is from 1965 to 1991 and the post-liberalisation period covers the period from 1992 to 2018. We find that cashew production is not influenced post trade liberalisation. This study also finds trade liberalisation has a significant and positive impact on export. Further, we reveal an insignificant impact of liberalisation on import. This study show that the area under cultivation is not changed after the trade liberalisation. 2024 Inderscience Publishers. All rights reserved. -
A colorimetric chemosensor for distinct color change with (E)-2-(1-(3-aminophenyl)ethylideneamino)benzenethiol to detect Cu2+ in real water samples
The study reports the synthesis of chemosensor (E)-2-(1-(3-aminophenyl)ethylideneamino)benzenethiol (C1), a highly sensitive, colorimetric metal probe that shows distinct selectivity for the detection of Cu2+ ion in various real water samples. Upon complexation with Cu2+ in CH3OH/H2O (60:40 v/v) (aqueous methanol), the C1 demonstrate significant enhancement in the absorption at 250nm and 300nm with a color change from light yellow to brown which was visualized using naked-eye. Therefore, these properties make C1 as an effective candidate for on-site Cu2+ ions detection. The emission spectrum of C1 illustrated TURN-ON recognition of Cu2+ with a limit of detection (LOD) of 46nM. Furthermore, Density Functional Theory (DFT) calculations were performed to better understand the interactions between C1 and Cu2+. The obtained results suggested that the electron clouds present around the NH2 innitrogen and sulfur in SH play a pivotal role in the formation of a stable complex. The computational results were in good agreement with the experimental UVvisible spectrometry results. Graphical abstract: [Figure not available: see fulltext.] 2023, The Author(s), under exclusive licence to The Japan Society for Analytical Chemistry. -
Smart Attendance Management System using IoT
Taking student attendance is mandatory in an educational organization, and maintaining those attendance plays a vital role. The conventional way of taking student attendance in any institution is time-consuming and challenging, because in the conventional procedure taking attendance/Roll call is performed manually by calling student names as per their roll numbers and marking 'absent(A)' or 'present(P)' on the attendance/logbook accordingly in every class per day. To improve teaching efficiency/teaching time in classrooms by reducing the time required for Roll call's, we have proposed a biometric student attendance system based on IoT. The proposed system records students' attendance using the facial-based biometric system and stores the attendance details on the server through the internet. In this system, the Raspberry pi camera captures the student face images and compares them with the stored images in the database. If the captured image is comparable with the stored image, then the student's attendance is recorded on the remote server as a present(P) in class; otherwise, attendance is recorded as absent (A). The developed system has been tested for sample classes, and the results proved that the system is simple, cost-effective, and portable for managing students' attendance. 2022 IEEE. -
Sensitive crop leaf disease prediction based on computer vision techniques with handcrafted features
Agricultural production is considered the primary source of the economy of many countries. Tomato and Potatoes are the most sensitive and consumable vegetables worldwide. However, during the growth of these crops, they suffer from many leaf diseases, which lead to loss of productivity and economy of the farmers. Many farmers detect and find plant diseases that are more time-consuming, expensive, and require expert decisions following the naked eye method. Therefore, early and accurate diagnosis of Tomato and Potato crops leaf diseases plays a vital role in sustainable agriculture. So, this research paper proposes an efficient leaf disease classification model based on computer vision techniques. The proposed Adaptive Deep Neural Network (ADNN) leaf disease classification method is a hybrid model which combines an optimized long short-term memory (OLSTM) and convolution neural network (CNN). The weight values supplied in the LSTM classifier are optimally selected using the Adaptive Raindrop Optimization algorithm. The handcrafted features are extracted from the segmented image and fused with the hybrid deep neural network to improve the classifier performance. The ADNN method consists of five steps: preprocessing, feature extraction, segmentation, handcrafted feature extraction, and classification. At first, the images are given to the preprocessing stage to remove the noise from leaf images. Then, the image-affected portion is segmented using an enhanced radial basis function neural network. After the segmentation process, the segmented image is given as an input to the adaptive deep neural network (ADNN) that classifies various types of diseases in the Potato and Tomato leaves. The efficiency of the ADNN model based on the OLSTM-CNN approach is determined concerning multiple metrics, namely Accuracy, Precision, Recall, F-measure, Specificity, and Sensitivity. The ADNN model achieved the best Accuracy of 98.02% for Tomatoes and 98% for Potatoes. The ADNN is compared with existing state-of-the-art CNN, LSTM, ResNet50, and MobileNet techniques. The performance analysis proved that the ADNN model improved efficiency in terms of all metrics and methods. 2023, The Author(s) under exclusive licence to The Society for Reliability Engineering, Quality and Operations Management (SREQOM), India and The Division of Operation and Maintenance, Lulea University of Technology, Sweden. -
Hybrid Deep Learning-Based Potato andTomato Leaf Disease Classification
Predicting potato and tomato leaf disease is vital to global food security and economic stability. Potatoes and tomatoes are among the most important staple crops, providing essential nutrition to millions worldwide. However, many tomato and potato leaf diseases can seriously reduce food productivity and yields. We are proposing a hybrid deep learning model that combines optimized CNN (OCNN) and optimized LSTM (OLSTM). The weight values of LSTM and CNN models are optimized using the modified raindrop optimization (MRDO) algorithm and the modified shark smell optimization (MSSO) algorithm, respectively. The OCNN model is used to extract potato leaf image features and then fed into the OLSTM model, which handles data sequences and captures temporal dependencies from the extracted features. Precision, recall, F1-score, and accuracy metrics are used to analyze the output of the proposed OCNN-OLSTM model. The experimental performance is compared without optimizing the CNN-LSTM model, individual CNN and LSTM models, and existing MobileNet and ResNet50 models. The presented model results are compared with existing available work. We have received an accuracy of 99.25% potato and 99.31% for tomato. The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024. -
A Comprehensive Review onCrop Disease Prediction Based onMachine Learning andDeep Learning Techniques
Leaf diseases cause direct crop losses in agriculture, and farmers cannot detect the disease early. If the diseases are not detected early and correctly, the farmer must undergo huge losses. It may lead to the wrong pesticide or over pesticide, directly affect crop productivity and economy, and indirectly affect human health. Sensitive crops have various leaf diseases, and early prediction of these diseases remains challenging. This paper reviews several machine learning (ML) and deep learning (DL) methods used for different crop disease segmentation and classification. In the last few years, computer vision and DL techniques have made tremendous progress in object detection and image classification. The study summaries the available research on different diseases on various crops based on machine learning (ML) and deep learning (DL) techniques. It also discusses the data sets used for research and the accuracy and performance of existing methods. It does mean that the methods and available data sets presented in this paper are not projected to replace published solutions for crop disease identification, perhaps to enhance them by finding the possible gaps. Seventy-five articles are analysed and reviewed to find essential issues that involve additional study for future research in this domain to promote continuous progress for data sets, methods, and techniques. It mainly focuses on image segmentation and classification techniques used to solve agricultural problems. Finally, this paper provides future research scope and challenges, limitations, and research gaps. 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. -
Potato Leaf Disease Identification using Hybrid Deep Learning Model
The potato is one of the most significant crops in the world. However, it is prone to several leaf diseases that can result in significant productivity losses, leading to economic challenges. Early and precise disease identification is essential for sensitive crops like potatoes. Deep learning approaches have demonstrated excellent potential in image-based disease classification tasks in recent years. This paper presents a hybrid strategy for classifying potato leaf image diseases by integrating Optimised Convolutional Neural Network (OCNN) and Long Short-Term Memory (LSTM) networks. The Adaptive Shark Smell Optimisation (ASSO) technique is used to optimize the weights of CNN models. The CNN component is initially used to extract pertinent characteristics from Potato leaves, capturing significant visual patterns related to various diseases. These extracted features are then fed into the LSTM model, which utilizes its sequential learning capability to model the temporal dependencies among the extracted features. The model performance is analyzed based on Accuracy, Precision, Recall, and F1-score criteria. Experimental results showed that the hybrid OCNN-LSTM model outperforms the individual CNN, LSTM, and MobileNet models. The proposed model results are compared with existing state-of-the-art work, and it was found that the OCNN-LSTM model performed better and received 99.02% accuracy. 2023 IEEE. -
Plant Leaf Disease Classification Using Optimal Tuned Hybrid LSTM-CNN Model
Tomatoes are widely cultivated and consumed worldwide and are susceptible to various leaf diseases during their growth. Therefore, early detection and prediction of leaf diseases in tomato crops are crucial. Farmers can take proactive measures to prevent the spread and minimize the impact on crop yield and quality by identifying leaf diseases in their early stages. Several Machine Learning (ML) and Deep Learning (DL) frameworks have been developed recently to identify leaf diseases. This research presents an efficient deep-learning approach based on a hybrid classifier by optimizing the CNN and LSTM models, which helps to enhance classification accuracy. Initially, Median Filtering (MF) is used for leaf image pre-processing. Then, an improved watershed approach is used for segmenting the leaf images. Subsequently, enhanced Local Gabor Pattern (LGP) and statistical and color features are extracted. An optimized CNN and LSTM are used for classification, and the weights are tuned using the SISS-OB (Self Improved Shark Smell With Opposition Behavior) algorithm. Finally, we have analyzed the performance using various measures. Since we have done segmentation, feature extraction, and optimization improvisations, our proposed methodology results are higher than other available methods and existing works. The results obtained at Learning Percentage (LP) is 90% which is far superior to those obtained at other LPs. The FNR (False Negative Rate) is much lower (0.05) at the 90th LP. The proposed model achieved better classification performance in terms of Accuracy of 97.13%, Sensitivity of 95.09%, Specificity of 95.24%, Precision of 94.31%, F measure of 96.71% and MCC 87.34%. 2023, The Author(s), under exclusive licence to Springer Nature Singapore Pte Ltd. -
Enhanced radial basis function neural network for tomato plant disease leaf image segmentation
Primary crop losses in agriculture are due to leaf diseases, which farmers cannot identify early. If the diseases are not detected early and correctly, then the farmer will have to undergo huge losses. Therefore, in the field of agriculture, the detection of leaf diseases in tomato crops plays a vital role. Recent advances in computer vision and deep learning techniques have made disease prediction easy in agriculture. Tomato crop front side leaf images are considered for research due to their high exposure to diseases. The image segmentation process assumes a significant role in identifying disease affected areas on tomato leaf images. Therefore, this paper develops an efficient tomato crop leaf disease segmentation model using an enhanced radial basis function neural network (ERBFNN). The proposed ERBFNN is enhanced using the modified sunflower optimization (MSFO) algorithm. Initially, the noise present in the images is removed by a Gaussian filter followed by CLAHE (contrast-limited adaptive histogram equalization) based on contrast enhancement and un-sharp masking. Then, color features are extracted from each leaf image and given to the segmentation stage to segment the disease portion of the input image. The performance of the proposed ERBFNN approach is estimated using different metrics such as accuracy, Jaccard coefficient (JC), Dice's coefficient (DC), precision, recall, F-Measure, sensitivity, specificity, and mean intersection over union (MIoU) and are compared with existing state-of-the-art methods of radial basis function (RBF), fuzzy c-means (FCM), and region growing (RG). The experimental results show that the proposed ERBFNN segmentation model outperformed with an accuracy of 98.92% compared to existing state-of-the-art methods like RBFNN, FCM, and RG, as well as previous research work. 2022 Elsevier B.V. -
Concept Drift Detection for Social Media: A Survey
The research over information retrieval from social media data has progressed for streaming data since the last decade. Recently, academic researchers have witnessed users' changing topics, trends, and intent on social media. This change of information with time takes into account the temporal attribute for real-time data, and thus, advances in this domain are exponentially growing. Although concept drift is still not explored due to a shortage of available datasets, concept drift for social media is minimally explored. This manuscript makes attempts to identify the types of concept drift for social media data, discuss the historical perspective of concept drift on social media, and enlist the possible research directions. 2021 IEEE. -
Investigation on AI-Based Techniques in Applications for Detecting Fatal Traffic Accidents
The difficulties with road accident rates today rank among the top concerns for health and social policy in nations across the continents. In this essay, we've spoken about the fatalities and injuries brought on by traffic accidents in several Indian states. We have also shed light on the various factors that contribute to traffic accidents. Many researchers have reported various methods for identifying automobile crashes or accidents that are discussed in this work. Additionally, we covered collision avoidance systems and their various kinds. An examination of the analysis techniques used to comprehend the numerous causes causing accidents is also included in the study. Traditional models are frequently used to identify problems such driver weariness, drowsiness, driving while intoxicated, and distractions. 2023 IEEE. -
A Comparison of Similarity Measures in an Online Book Recommendation System
To assist users in identifying the right book, recommendation systems are crucial to e-commerce websites. Methodologies that recommend data can lead to the collection of irrelevant data, thus losing the ability to attract users and complete their work in a swift and consistent manner. Using the proposed method, information can be used to offer useful information to the user to help enable him or her to make informed decisions. Training, feedback, management, reporting, and configuration are all included. Our research evaluated user-based collaborative filtering (UBCF) and estimated the performance of similarity measures (distance) in recommending books, music, and goods. Several years have passed since recommendation systems were first developed. Many people struggle with figuring out what book to read next. When students do not have a solid understanding of a topic, it can be difficult determining which textbook or reference they should read. 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. -
Quantitative assessment of blockchain applications for Industry 4.0 in manufacturing sector
Blockchain is one of the emerging digital technologies that will play a role in the breakthroughs of the fourth industrial revolution. The use of blockchain technology has the potential to greatly benefit businesses of all sizes by increasing their data's integrity, privacy, and openness. The term Industry 4.0? refers to the amalgamation of recent advances in manufacturing technology that have helped businesses cut production times significantly. The industrial and supply chain industries can benefit from these technological advancements in a number of ways. Increased efficiency in production and a more stable supply chain are just two of the many benefits that blockchain promises to bring to the manufacturing industry. The study focuses on Blockchain's huge potential in the context of Industry 4.0. Understanding the role of Blockchain technology in Industry 4.0 is examined, along with its various drivers, enablers, and associated capabilities. The several sub-domains of Industry 4.0 that can benefit from the implementation of Blockchain technology are also covered. The present research is primary and exploratory in nature. The sample size of the study is 256. The responses obtained from workers working in manufacturing sector in Delhi/NCR. The responses from workers obtained through structured questionnaire. The several sub-domains of Industry 4.0 found that can benefit from the implementation of Blockchain technology. At last, the existing study found the most important uses of Blockchain technology in the fourth industrial revolution. 2023 -
A Unified Approach to Predict and Understand Acute Myeloid Leukemia Diagnosis
Acute myeloid leukemia (AML) is a rapidly progressing disease that affects myeloid cells in blood and bone marrow. These abnormal cancerous cells called blast cells are non-functional cells that increase rapidly in bone marrow and are released into blood stream which crowd out the healthy functional cells leading to weak immune system. This life-threatening disease needs to be diagnosed at early stage and hence requires fully automated system for early detection of leukemia to aid pathologists and doctors. Most of the automated machine learning and AI models are not transparent and require techniques to explain model prediction. This paper presents methods to classify blood microscopic images into healthy or acute myeloid leukemia. Among all the methods implemented, Gradient Boosting outperforms with an accuracy of 96.67%. This paper also focuses on explainable AI to interpret model prediction and feature importance which further helps in understanding decision-making process of classification model and optimize it. 2024, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. -
A Study of Preprocessing Techniques on Digital Microscopic Blood Smear Images to Detect Leukemia
Digital microscopic blood smear images can get distorted due to the noise as a result of excessive staining during slide preparation or external factors during the acquisition of images. Noise in the image can affect the output of further steps in image processing and can have an impact on the accuracy of results. Hence, it is always better to denoise the image before feeding it to the automatic diagnostic system. There are many noise reduction filters available; the selection of the best filter is also very important. This paper presents a comparative study of some common spatial filters like wiener filter, bilateral filter, Gaussian filter, median filter and mean filter which are efficient in noise reduction, along with their summary and experimental results. Performing comparative analysis of result based on PSNR, SNR and MSE values, it can be determined that median filter is most suitable method for denoising digital blood smear images. 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. -
Machine Learning Model to Detect Chronic Leukemia in Microscopic Blood Smear Images
Chronic leukemia is a slow-progressing form of disease, If not diagnosed on time can progress and increase the risk of life-threatening complications. It is essential to develop a fully automated system to recognize and categorize type of leukemia for proper evaluation and treatment. This paper aims to provide a machine learning model to identify and classify chronic lymphocytic leukemia, chronic myeloid Leukemia and healthy cells. Digital microscopic blood smear images were automatically cropped into single nucleus and segmented using watershed algorithm. Grey level co-occurrence matrix (GLCM) and geometrical features were extracted from the segmented nucleus images and random forest algorithm is used to classify chronic leukemia and healthy cells. This prognosis aids pathologists and physicians in identifying leukemic patients early and selecting the most effective course of action. 2023 IEEE. -
Segment Anything Model (SAM) to Segment lymphocyte from Blood Smear Images
Automated lymphocyte segmentation from smear images plays an important role in disease diagnosis and monitoring, aiding in the assessment of immune system function and pathology detection. This study proposes an approach for lymphocyte segmentation utilizing Segment Anything Model (SAM) which is a deep learning model. Our method leverages a pre trained SAM architecture and fine-tunes it on a custom dataset comprising smear images containing lymphocytes. The pretrained model's ability of versatile segmentation combined with fine-tuning on the specific dataset enhances its performance in accurately identifying lymphocyte boundaries. We evaluate the proposed approach on a diverse set of smear images, demonstrating its effectiveness in segmenting lymphocytes with impressive IOU score and Dice Score. SAM deep learning model, fine-tuned on custom datasets, holds promise for robust and efficient lymphocyte segmentation from blood smear images. 2024 IEEE. -
Acute Leukemia Subtype Recognition in Blood Smear Images with Machine Learning
Acute leukemia is a swiftly progressing blood cancer affecting white blood cells which poses a significant threat to the immune system and often leads to fatal outcomes if not detected and treated promptly. The current manual diagnostic method, being time-consuming and prone to errors, necessitates an urgent shift toward a comprehensive automated system. This paper presents an innovative approach to automatically identify acute leukemia cells and their subtypes by analyzing microscopic blood smear images. The proposed methodology involves the segmentation of clustered lymphocytes, isolation of nuclei, and extraction of diverse features from each nucleus. A random forest classifier is then trained to categorize nuclei into healthy or cancerous, with further precision in classifying cancerous nuclei into specific subtypes. The method achieves an impressive 97% accuracy across all evaluations, holding profound implications for pathologists and medical practitioners in their decision-making processes. 2024, J.J. Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology. All rights reserved.