International Journal of Innovations in Science & Technology
Not a member yet
813 research outputs found
Sort by
Intelligent license Plate Recognition System
Get PDFSince the 19th century, the number of vehicles has been increasing rapidly with the growth of the human population. To supervise vehicles, license plates are used all over the world. The license plate is the unique identity for vehicles; that’s why it is always used to monitor and keep records of vehicles by law enforcement, border monitoring, parking control, and many other applications. Monitoring a huge number of vehicles is a difficult task using traditional (manual) methods. The Intelligent License Plate Recognition (ILPR) system overcomes these problems by recognizing plate identities without human involvement through artificial intelligence and machine learning processes. This system extracts the identity number allocated to each vehicle from the license plate and can provide information about a specific vehicle. It can be further applied in regulated zones such as military areas, parking control, toll collection, and for identifying non-tax-paid vehicles. For developing the ILPR system, text extraction and deep learning techniques must be combined.
The ILPR system, developed by integrating Deep Learning (DL), Image Processing (IP), and image-to-text extraction approaches, is used to detect plate identity. YOLOv8 is used for object detection and the OCR engine for text extraction. The system will be capable of detecting live license plates with high accuracy, which will help in regulated zones and traffic system applications
Numerical Simulation of Flow Past a Square Object Detached with Controlling Object at Various Reynolds Number
Get PDFA two-dimensional (2-D) numerical study has been conducted for flow past of two different configurations of square objects by using the numerical technique Lattice Boltzmann Method (LBM). In these configurations, one object plays the role of the main object, while the second object acts as a controlling object positioned in two different ways, such as firstly placed at the top right corner of the main object (first configuration) and in the second configuration the control object is put at the bottom right corner of the main object at L = 20d (where d is the size of the object). The primary goal of this study was to investigate the impact of the control object on the main object to reduce fluid force and suppress the vortex shedding. Initially, the code\u27s validity was checked, and the effect of the computational domain was studied to determine accurate upstream (Lup), and downstream (Ldown) distances and height of channel (H). Subsequently, all the numerical computations were performed by considering the range of Reynolds numbers (Re = U∞d/ʮ) Re = 80 to 200. The results are presented in terms of vorticity contour, drag (Cd) and lift coefficients (Cl), and physical parameters, including Cdmean, Cdrms, Clrms, and St. In the vorticity contour, three distinct modes of flow structures were observed for the first configuration (where the control object is placed at the bottom corner of the main object), such as i) Von Karman vortex street (VKVS) flow mode, ii) Two rows vortex street (TRVS) flow mode and iii) Critical flow mode (CF). For the second configuration, two different types of flow modes are identified, dominating the critical flow behavior, those are i) Irregular vortex shedding (IVS) flow mode and ii) Critical flow (CF) mode. The values of Cdmean, Cdrms, Clrms, and St are calculated against the Reynolds number. For the main object in both configurations, the value of Cdmean decreases at the lower range of Reynolds numbers and then continuously increases at larger values of Reynolds numbers. However, for the control object, the mean drag coefficient consistently increased with an increment in the range of Reynolds numbers. The maximum value of Cdmean is attained at Re = 200, reaching the value of 2.0708 for the configuration where the control object is placed at the top right corner. Similarly, the highest value of the Strouhal number is obtained for the control object; placed at the bottom right corner for C2, with a value of 0.1321 occurring at either Re = 100 or Re = 120
Spatial Investigation of Soil Erosion Risk in the High Rainfall Zone of Pakistan by Using Rusle Model
Get PDFThe degradation of soil quality and agricultural sustainability is threatened by soil erosion, which poses a serious threat to livelihoods and food security. Maintaining soil fertility and reducing the danger of erosion require efficient evaluation and management strategies. This research presents an innovative approach to assessing soil erosion in Nowshera District, leveraging remote sensing technology coupled with Geographic Information System (GIS) tools. The study intends to offer a more thorough and accurate understanding of erosion patterns and drivers in the area by incorporating these cutting-edge approaches. Cloud-free LANDSAT 8 multispectral images, characterized by minimal vegetation cover, serve as the primary dataset for this analysis. The integration of the RUSLE model with GIS and remote sensing techniques enables the calculation of soil erosion rates throughout the research region. The study demonstrates variation in soil erosion parameters across different locations, as indicated by the range of R factor values, which range from 603.43 to 696.43 MJ mm/ha/h/year. The southeastern portion demonstrates significantly lower erosion rates than the northwestern part, which can be linked to variations in topography and land use patterns. This study highlights the significance of using remote sensing techniques to evaluate soil erosion changes over time and provide valuable information for land management plans in Nowshera District, Pakistan. This information may be used to make better decisions regarding conservation planning and agricultural sustainability
Sherpa: Implementing a Hybrid Recommendation System for Next-Gen Tourist Experience
Get PDFIn the digital era, Sherpa revolutionizes personalized tourism with an AI-driven recommendation system, fostering meaningful connections between travelers and local guides. This study explores Sherpa\u27s integration of collaborative and content-based filtering—specifically, singular value decomposition (SVD) and cosine similarity—to tailor travel experiences uniquely. Our methodology includes a detailed examination of Sherpa\u27s algorithm and its implementation within a cross-platform, MERN Stack-powered backend. We assess the system\u27s efficacy in aligning recommendations with individual user preferences, based on quantitative user feedback and engagement metrics. Initial results demonstrate a significant improvement in personalized experience satisfaction. The paper concludes that Sherpa\u27s innovative approach not only enhances the quality of travel recommendations but also sets a new standard for interactive and adaptive tourism platforms. Through continuous algorithmic refinement, Sherpa is positioned to lead a transformative shift in how travelers explore new destinations, offering not just journeys, but transformative experiences
Beyond CNNs: Encoded Context for Image Inpainting with LSTMs and Pixel CNNs
Get PDF
ur paper presents some creative advancements in the image in-painting techniques for small, simple images for example from the CIFAR10 dataset. This study primarily targeted on improving the performance of the context encoders through the utilization of several major training methods on Generative Adversarial Networks (GANs). To achieve this, we upscaled the network Wasserstein GAN (WGAN) and compared the discriminators and encoders with the current state-of-the-art models, alongside standard Convolutional Neural Network (CNN) architectures. Side by side to this, we also explored methods of Latent Variable Models and developed several different models, namely Pixel CNN, Row Long Short Term Memory (LSTM), and Diagonal Bidirectional Long Short-Term Memory (BiLSTM). Moreover, we proposed a model based on the Pixel CNN architectures and developed a faster yet easy approach called Row-wise Flat Pixel LSTM. Our experiments demonstrate that the proposed models generate high-quality images on CIFAR10 while conforming the L2 loss and visual quality measurement
Low-Cost Smart Metering Using Deep Learning
Get PDFUtility services like electricity, water, and gas are essential for modern living, and their demand has been rising worldwide. However, traditional manual meter reading is a standard procedure for billing purposes. This is not only labor and time-intensive but also prone to mistakes, which results in incorrect billing and revenue losses. In the era of advanced AI, leveraging cutting-edge technology to automate meter readings has become increasingly viable. However, Existing AI-based meter reading systems have limitations in detecting and recognizing meters from a distance. This research addresses these problems by presenting a novel system that utilizes the YOLOv8 model to detect meter screens from a distance. In addition, the system uses a fine-tuned Paddle OCR to recognize meter readings. A Novel dataset curated for the meter screen detection, recognition, and end-to-end OCR tasks related to electricity, gas, and water utility meters has been presented, containing up to 8,044 images. The proposed system was trained and extensively tested on the proposed dataset to gauge its performance. The system achieved an exceptional mean Average Precision (mAP) of 0.995 for both analog and digital meters on the detection task; furthermore, the system achieved an accuracy of 96.92% in the recognition task, which is 70% better than the accuracy of Pre-trained Paddle OCR. Moreover, an all-encompassing evaluation that combines detection and recognition using Paddle OCR and YOLOv8, i.e., the end-to-end OCR task, achieved an accuracy of 97.8%. Lastly, the system achieved an inference speed of up to 6 frames per second, guaranteeing real-time effectiveness
Visually: Assisting the Visually Impaired People Through AI-Assisted Mobility
Get PDFThis research introduces “Visually”, a revolutionary mobile application that aims to address the complications that visually impaired people come across in their daily lives. By deploying advanced deep learning models for real-time object detection, facial recognition, and currency identification with voice outputs for each feature, the “Visually” application strives to enhance the autonomy, independence, and mobility of visually impaired people. The system undergoes thorough training on a diverse dataset, incorporating augmentation techniques to enhance the robustness of the models. The project\u27s multifaceted objectives include a user-friendly interface, real-time object detection, multi-modal recognition, Text-to-Speech audio output, and an overarching aim of enriching the lives of visually impaired individuals. Driven by the global prevalence of visual impairment and the demand for cost-effective solutions, “Visually” is aligned with international efforts for accessibility and inclusivity. For cross-platform compatibility, the machine learning models have been integrated whilst being deployed with TensorFlow Lite. With Offline availability, the application ensures accessibility even in rural areas with limited network connectivity. To make a substantial societal impact "Visually" aims to contribute to a more inclusive and equitable society, by transforming the way visually impaired individuals navigate around the environment. Positioned at the intersection of technology, accessibility, and empowerment, the “Visually” project is poised to bring about positive change for a community that frequently encounters unique challenges in their daily lives
SSOCANET SSOCANET - Empowering VANETs with Salp Swarm Optimization-Enhanced Clustering Algorithm
No full textVehicular Ad hoc networks (VANETs) present significant challenges due to the dynamic nature of vehicle movements, leading to a constantly changing vehicular network topology. This instability results in packet loss, network fragmentation, message reliability, and scalability issues. To address these challenges, clustering has emerged as a promising solution to escalate vehicle communication efficiency. However, determining the optimal number of clusters remains a crucial problem. The proposed solution, the Salp Swarm Optimization-Enhanced Clustering Algorithm for VANET (SSOCANET), leverages the foraging behavior of salps to optimize cluster formation based on multiple objectives. SSOCANET achieves an optimal number of clusters by employing carefully designed objective functions, minimizing communication overhead and end-to-end communication latency in a network. The simulation results demonstrate the superior performance of SSOCANET compared to other clustering approaches, offering a robust solution for VANETs
Exploring the Dynamics of Urban Sprawl Using GIS & RS Techniques and By Modeling Using Ca-Markov Model in District Peshawar
Get PDFUrban Sprawl is described by the unplanned and uneven growth pattern in the built-up areas, determined by several processes and leading to ineffective resource utilization. Pakistan, a developing country, is struggling with extreme population growth and currently ranks fifth globally in terms of population size. Peshawar, the provincial capital of Khyber Pakhtunkhwa, has undergone significant urbanization in recent decades for various reasons, necessitating a comprehensive analysis to inform urban planning. In the present work the urban sprawl of the Peshawar district has been studied from 2010-2020, and future predictions for the year 2030 are evaluated. This research uniquely utilizes the CA-Markov model to predict urban sprawl for the year 2030, a method not previously applied in the earlier studies in the study area. This research is carried out to examine the land use pattern, to find out the urban sprawl from 2010 to 2020 using remotely sensed satellite data for three periods (2010, 2015 and 2020). The object-Based Image Analysis (OBIA) approach was used to examine the land use patterns. The LULC prediction till 2030 is done by using the CA-Markov Model in a GIS environment. The pattern of development of urban sprawl in Peshawar is typical of most Pakistani major cities, where ribbon sprawl is common along major roads, while leapfrog sprawl is dominant in the city’s outer edge. The LULC changes derived from the OBIA method show that urban area expanded from 23% to 39% of the whole area, while agriculture decreased from 44% to 35% over ten years. To grip land use changes better, the paper proposes a method for the simulation of spatial patterns. The simulating method can be divided into two parts: one is a quantitative forecast by using the Markov model and the other is simulating the spatial pattern changes by using the CA model. The above two models construct the simulative model of the spatial pattern of land use. CA–Markov is used to simulate the spatial pattern of land use in Peshawar for 2030, which indicates that the urban land will reach a total of 44% consuming areas from Barren Land and Vegetation land
Impact Assessment of Monsoon Precipitation on Groundwater Level in Lahore District GEE Script
Get PDFIntroduction/Importance of Study: Precipitation is a crucial component of the global water cycle and a primary source of freshwater, with groundwater being vital for drinking water worldwide, especially in regions like Lahore, Pakistan, where it supports 60-70% of the population.
Novelty Statement: This study uniquely addresses the impact of monsoon precipitation on groundwater levels in Lahore, providing a comprehensive analysis that has not been previously undertaken.
Material and Method: Data from the Pakistan Meteorological Department and WASA’s hydrology branch (2018-2022) were analyzed using GIS-based Inverse Distance Weighted Interpolation and statistical methods to assess precipitation patterns and groundwater levels.
Result and Discussion: The findings indicate that monsoon rainfall significantly raises groundwater levels by 2-3 meters due to seepage and infiltration. Spatial and temporal analyses revealed that the monsoon period, especially July and August, contributes the most to groundwater recharge. Despite this, the overuse of groundwater during non-monsoon months and extensive urban infrastructure limit overall groundwater recharge. The study found a positive correlation between monsoon precipitation and groundwater levels, emphasizing the critical role of sustainable aquifer management to maintain groundwater resources.
Concluding Remarks: Sustainable management of aquifer recharge is essential to ensure the long-term availability of groundwater resources in Lahore