6 research outputs found
A Comprehensive Review on Lane Marking Detection Using Deep Neural Networks
No full textLane marking recognition is one of the most crucial features for automotive vehicles as it is one of the most fundamental requirements of all the autonomy features of Advanced Driver Assistance Systems (ADAS). Researchers have recently made promising improvements in the application of Lane Marking Detection (LMD). This research article has taken the initiative to review lane marking detection, mainly using deep learning techniques. This paper initially discusses the introduction of lane marking detection approaches using deep neural networks and conventional techniques. Lane marking detection frameworks can be categorized into single-stage and two-stage architectures. This paper elaborates on the network’s architecture and the loss function for improving the performance based on the categories. The network’s architecture is divided into object detection, classification, and segmentation, and each is discussed, including their contributions and limitations. There is also a brief indication of the simplification and optimization of the network for simplifying the architecture. Additionally, comparative performance results with a visualization of the final output of five existing techniques is elaborated. Finally, this review is concluded by pointing to particular challenges in lane marking detection, such as generalization problems and computational complexity. There is also a brief future direction for solving the issues, for instance, efficient neural network, Meta, and unsupervised learning
Bangladeshi crops leaf disease detection using YOLOv8
No full textThe agricultural sector in Bangladesh is a cornerstone of the nation’s economy, with key crops
such as rice, corn, wheat, potato, and tomato playing vital roles. However, these crops are highly
vulnerable to various leaf diseases, which pose significant threats to crop yields and food security
if not promptly addressed. Consequently, there is an urgent need for an automated system that
can accurately identify and categorize leaf diseases, enabling early intervention and management.
This study explores the efficacy of the latest state-of-the-art object detection model, YOLOv8 (You
Only Look Once), in surpassing previous models for the automated detection and categorization
of leaf diseases in these five major crops. By leveraging modern computer vision techniques, the
goal is to enhance the efficiency of disease detection and management. A dataset comprising 19
classes, each with 150 images, totaling 2850 images, was meticulously curated and annotated
for training and evaluation. The YOLOv8 framework, known for its capability to detect multiple
objects simultaneously, was employed to train a deep neural network. The system’s performance
was evaluated using standard metrics such as mean Average Precision (mAP) and F1 score. The
findings demonstrate that the YOLOv8 framework successfully identifies leaf diseases, achieving a
high mAP of 98% and an F1 score of 97%. These results underscore the significant potential of this
approach to enhance crop disease management, thereby improving food security and promoting
agricultural sustainability in Bangladesh
A deep learning approach for lane marking detection applying encode-decode instant segmentation network
No full textA lot of people suffer from disability and death due to unintentional road accidents, which also result in the loss of a significant amount of financial assets. Several essential features of Advanced Driver Assistance Systems (ADAS) are being incorporated into vehicles by researchers to prevent road accidents. Lane marking detection (LMD) is a fundamental ADAS technology that helps the vehicle to keep its position in the lane. The current study employs Deep Learning (DL) methodologies and has several research constraints due to various problems. Researchers sometimes encounter difficulties in LMD due to environmental factors such as the variation of lights, obstacles, shadows, and curve lanes. To address these limitations, this study presents the Encode-Decode Instant Segmentation Network (EDIS-Net) as a DL methodology for detecting lane marking under various environmental situations with reliable accuracy. The framework is based on the E-Net architecture and incorporates combined cross-entropy and discriminative losses. The encoding segment was split into binary and instant segmentation to extract information about the lane pixels and the pixel position. DenselyBased Spatial Clustering of Application with Noise (DBSCAN) is employed to connect the predicted lane pixels and to get the final output. The system was trained with augmented data from the Tusimple dataset and then tested on three datasets: Tusimple, CalTech, and a local dataset. On the Tusimple dataset, the model achieved 97.39% accuracy. Furthermore, it has an average accuracy of 97.07% and 96.23% on the CalTech and local datasets, respectively. On the testing dataset, the EDIS-Net exhibited promising results compared to existing LMD approaches. Since the proposed framework performs better on the testing datasets, it can be argued that the model can recognize lane marking confidently in various scenarios. This study presents a novel EDIS-Net technique for efficient lane marking detection. It also includes the model's performance verification by testing in three different public datasets
An Approach to Detect Suicidal Bengali Posts from Social Media Using Machine Learning Algorithms
No full textIn this modern era, suicide is one of the critical issues. According to the WHO, more than seven million people die
due to suicide every year. Suicide is also the second cause of unnatural death for persons between the ages of 15 and 29.
Youth in nations like Bangladesh struggle with schoolwork, employment, relationships, drug use, and family issues, all of
which are significant or minor contributors on the road to depression. In Bangladesh, people are uncomfortable discussing
this ailment openly and frequently mistake this problem as madness. Many at-risk persons use social platforms to talk about
their issues or get knowledge on related topics. This study aims to prevent suicide by identifying suicidal posts on social
media. We collected suicidal-related data from Kaggle. (We use nine algorithms for three features). The prediction model
achieved good performance. Stochastic Gradient Descent is the best model with the highest accuracy for unigram features,
87.23%. For bigram features, the Multinomial Naive Bayes is the best model with the highest accuracy, 88.69%. The best
model with the highest accuracy for trigram features, 86.13%, is Stochastic Gradient Descent. This research demonstrates the
chance that a machine-learning strategy can reduce the risk of suicide. Hopefully, this model will serve as a guide for
lowering potential suicide risk in the future. The study concludes with a summary of several practical concerns that may be
considered to improve model performanc
Light-Weight Deep Learning Model for Accelerating the Classification of Mango-Leaf Disease
No full textMango leaf diseases represent a serious threat to world agriculture, necessitating prompt and accurate detection to avert catastrophic effects. In response, this study suggests a light-weight, deep learning-based method for automatically classifying mango leaf diseases. The model is based on the original DenseNet architecture, which is well known for its effectiveness in image classification tasks. Custom layers have been added over the existing layer of the original DenseNet model. The proposed model has been compared with other existing pre-trained models. Based on comparisons, the proposed model, DenseNet78, proved to be efficient even on a relatively small dataset, where the conventional model failed. The proposed model ensured generalization across regions, disease variants, and diverse datasets of mango leaves. The results demonstrate that the fine-tuned DenseNet architecture (DenseNet78), along with an ideal growth rate, modifying block size, and a number of layers, provides optimum accuracy, with 99.47% accuracy in identifying healthy mango leaves and 99.44% accuracy in detecting various mango leaf diseases. The results also demonstrate that the model is effective in accelerating the training process because of careful comparative analysis of all the available alternatives, including the most effective combination of optimizers, learning rate schedulers, and loss functions. The study's conclusion is an automated approach for diagnosing mango leaf disease using an improved and optimized DenseNet architecture (DenseNet78)
A remote-controlled global navigation satellite system based rover for accurate video-assisted cadastral surveys
No full textOne of the main tasks of a cadastral surveyor is to accurately determine property boundaries by measuring control points and calculating their coordinates. This paper proposes the development of a remotely-controlled tracking system to perform cadastral measurements. A Bluetooth-controlled rover was developed, including a Raspberry Pi Zero W module that acquires position data from a VBOX 3iSR global navigation satellite system (GNSS) receiver, equipped with a specific modem to download real-time kinematic (RTK) corrections from the internet. Besides, the Raspberry board measures the rover speed with a hall sensor mounted on a track, adjusting the acquisition rate to collect data at a fixed distance. Position and inertial data are shared with a cloud platform, enabling their remote monitoring and storing. Besides, the power supply section was designed to power the different components included in the acquisition section, ensuring 2 hours of energy autonomy. Finally, a mobile application was developed to drive the rover and real-time monitor the travelled path. The tests indicated a good agreement between rover measurements and those obtained by a Trimble R10 GNSS receiver (+0.25% mean error) and proved the superiority of the presented system over a traditional metric wheel
