INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIV, Issue X, October 2025
www.ijltemas.in Page 527
Optimized Python Approach for Vehicle License Plate Detection
and Recognition
1 Ravi Tomar, 1 Manoj Kumar, 1 Sharad Kumar, 1 Sachin Kumar, 1 Jagdeep Singh, 2 Vikas Sharma
School of Engineering & Technology, Shri Venkateshwara University, Gajraula, U.P. India
2 Department of Computer Applications, SRM Institute of Science and Technology, Delhi NCR Campus, Ghaziabad, U.P.
India
DOI: https://doi.org/10.51583/IJLTEMAS.2025.1410000067
Abstract—Automatic license plate detection and recognition (ALPR) is a critical component of intelligent transportation
systems, enabling applications such as traffic monitoring, parking management, and law enforcement. This paper presents an
efficient Python-based approach that integrates image processing, computer vision, and optical character recognition (OCR)
techniques for real-time license plate detection and recognition. The proposed framework employs pre-processing methods to
enhance image quality, followed by contour detection and segmentation to accurately localize license plates. Subsequently,
character recognition is performed using OCR to extract alphanumeric information from the detected plates. The system is
evaluated on diverse datasets containing vehicles under varying lighting and environmental conditions. Experimental results
demonstrate high detection accuracy, precision, recall, and F1-score while maintaining low computational overhead, making it
suitable for deployment in real-world scenarios. The Python-based implementation ensures flexibility, scalability, and ease of
integration with existing IoT and smart city infrastructures.
Keywords—ALPR, Python, Computer Vision, Image Processing, Optical Character Recognition, Vehicle Identification,
Intelligent Transportation Systems.
I. Introduction
Automatic License Plate Recognition (ALPR) has emerged as one of the most widely adopted technologies in modern intelligent
transportation systems (ITS). With the exponential increase in the number of vehicles worldwide, efficient traffic management,
surveillance, and law enforcement have become increasingly challenging. Conventional methods of vehicle identification, such as
manual monitoring and human intervention, are inefficient, error-prone, and unable to cope with the rapid pace of urbanization
and population growth. As a result, automated solutions that can seamlessly integrate with existing infrastructure are necessary to
address issues such as congestion, road safety, toll collection, and crime prevention. Valova et al. [1] proposed a conceptual
model for a parking system integrated with ALPR, highlighting the potential for seamless vehicle entry and exit management.
ALPR systems serve as a cornerstone in this regard, offering automated identification of vehicles by detecting and recognizing
license plates from captured images or videos. The significance of ALPR technology lies in its versatility and applicability across
various domains. In urban areas, ALPR facilitates traffic flow monitoring, congestion control, and automated traffic violation
detection. For parking management systems, it ensures seamless entry and exit by authenticating vehicles without the need for
manual intervention. Similarly, law enforcement agencies rely heavily on ALPR for detecting stolen vehicles, tracking suspects,
and enforcing legal compliance. With the rise of smart cities and Internet of Things (IoT) applications, ALPR has become even
more critical, enabling real-time vehicle monitoring and data sharing across interconnected platforms. Thus, developing efficient,
accurate, and scalable ALPR systems has become a pressing need for researchers and practitioners alike is shown in Fig.1. The
process of license plate recognition is inherently challenging due to the diversity of real-world conditions. Vehicles are captured
under different environmental circumstances, such as varying lighting conditions, weather disturbances, occlusions, or camera
angles. Additionally, license plates differ in size, color, font style, and orientation depending on the country or region, further
complicating recognition tasks. Noise in images caused by motion blur or low-resolution cameras adds to the complexity.
Traditional ALPR approaches primarily relied on rule-based methods, including edge detection, morphological operations, and
geometric analysis for license plate localization. While these methods performed adequately under controlled conditions, they
often struggled in dynamic environments where lighting variations or plate distortions occur. With the advancements in artificial
intelligence, particularly deep learning, many modern ALPR systems leverage convolutional neural networks (CNNs) and object
detection frameworks for robust plate detection and recognition. However, these deep learning-based solutions often require large
datasets for training, significant computational resources, and complex hardware setups, which may limit their deployment in
cost-sensitive or resource-constrained environments. In this context, Python has emerged as a powerful programming language
for developing efficient and flexible ALPR systems. Python’s extensive libraries for image processing (OpenCV), machine
learning (scikit-learn, TensorFlow, PyTorch), and optical character recognition (Tesseract OCR) make it an ideal choice for
building end-to-end recognition pipelines. Its simplicity, scalability, and community-driven ecosystem allow researchers and
developers to rapidly prototype, test, and deploy ALPR solutions in real-world scenarios. Moreover, Python-based
implementations are highly adaptable, making them compatible with diverse IoT and smart city platforms, thereby supporting
seamless integration and interoperability. The approach presented in this paper emphasizes efficiency, accuracy, and real-time
performance while minimizing computational overhead. The proposed framework begins with pre-processing steps to enhance
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIV, Issue X, October 2025
www.ijltemas.in Page 528
input image quality through techniques such as grayscale conversion, noise removal, and contrast adjustment. These steps ensure
that license plate regions are more distinguishable from the background. Next, contour detection and segmentation are applied to
localize license plates within vehicle images. This method leverages geometric properties and edge detection algorithms to
identify rectangular regions corresponding to license plates. Once the license plate is isolated, segmentation techniques are used
to separate individual characters. Optical Character Recognition (OCR) is then employed to extract alphanumeric text, enabling
the final step of vehicle identification. The efficiency of this Python-based approach lies in its balance between traditional image
processing techniques and lightweight OCR methods, avoiding the heavy computational demands of deep learning models while
still ensuring robustness in challenging environments. The system is designed to operate effectively under diverse conditions,
including variations in illumination, plate design, and background noise. This makes the framework highly suitable for
deployment in practical applications such as traffic surveillance, toll collection booths, parking management systems, and real-
time law enforcement monitoring. Another notable aspect of the proposed approach is its scalability and flexibility. The Python-
based pipeline can be easily extended or customized to accommodate region-specific license plate formats, varying camera
hardware, or integration with cloud-based services for large-scale deployments. Additionally, the system supports real-time
operation, enabling timely decision-making in scenarios where rapid vehicle identification is critical. By ensuring high detection
accuracy along with low computational requirements, the proposed framework provides a cost-effective solution that bridges the
gap between resource-intensive deep learning models and less reliable traditional approaches.
Fig. 1. Automatic License Plate Recognition Architecutre
II. Literature Review
Automatic License Plate Recognition (ALPR) has emerged as a pivotal technology in intelligent transportation systems, enabling
efficient traffic management, parking solutions, and law enforcement. Dayal et al. [2] explored machine learning techniques for
automatic number plate recognition, demonstrating that supervised learning models can achieve high recognition accuracy with
optimized preprocessing and segmentation. Similarly, Du et al. [3] leveraged embedded technology combined with the YOLOv11
model for license plate detection, emphasizing real-time applicability and computational efficiency in dynamic traffic
environments. Recent studies have focused on integrating deep learning frameworks to enhance detection and recognition
performance. DC et al. [4] utilized a Single Shot Detector (SSD) with MobileNet for ALPR, achieving robust performance with
low computational overhead, while Agarwal et al. [6] combined EasyOCR and TensorFlow for intelligent video-based license plate
detection, underscoring the effectiveness of end-to-end deep learning pipelines in processing streaming video data. Rafek et al. [7]
further demonstrated the applicability of deep learning models in video stream recognition, achieving high accuracy in real-time car
plate recognition scenarios. OCR-based approaches continue to play a critical role in character recognition. Pallempati et al. [9]
implemented OCR methods for handwritten character recognition and vehicle number recognition, highlighting the versatility of
OCR for various text-based recognition tasks. Additionally, Ismael and Saeed [13] explored mobile robot navigation using
computer vision and OCR techniques, illustrating the broader applicability of OCR in autonomous systems. Kharatmol et al. [15]
emphasized the use of OCR and image processing in extracting meaningful information from educational credentials, indicating
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIV, Issue X, October 2025
www.ijltemas.in Page 529
cross-domain utility of these methods. Beyond ALPR, the integration of image processing and AI techniques has been applied in
broader security and monitoring contexts. Vashishth et al. [5] investigated image processing for counterfeit currency detection, and
Sharma et al. [14] explored advanced facial feature predictions for child search systems, indicating the growing role of AI in
surveillance and automated monitoring. Thamilarasi et al. [10] developed an AI-based smart traffic guard system for real-time
detection and penalty enforcement, reflecting the increasing adoption of intelligent systems for traffic compliance and safety.
Krishna et al. [11] and Sofiykov et al. [12] emphasized real-time vehicular traffic management using deep learning and computer
vision frameworks, demonstrating practical deployments in urban traffic scenarios. Reddy et al. [8] extended deep learning
applications to network traffic analysis, showing the potential of AI in detecting anomalies beyond the visual domain. In parallel,
security-focused applications in networking have also gained prominence. Sharma and Teotia [16] proposed an optimized Graph
Neural Network (GNN)-based intrusion detection system for dynamic Mobile Ad Hoc Networks (MANETs), highlighting
lightweight architecture, online adaptation, and dynamic graph representation for real-time intrusion detection. Complementing
this, Sharma et al. [17] conducted a comparative study of deep learning models, including convolutional and recurrent neural
networks, for fake news classification, demonstrating significant variations in model performance and underscoring the potential of
hybrid approaches. Furthermore, Sharma and Teotia [18] provided a comprehensive analysis of security mechanisms in MANETs,
characterizing threats such as routing attacks, denial-of-service (DoS), and blackhole vulnerabilities. Collectively, these studies
underscore the evolution of ALPR systems from traditional image processing to sophisticated deep learning frameworks, while also
showing the broader application of AI and security mechanisms across domains.
III. Proposed Methodology
The proposed Python-based framework for Automatic License Plate Recognition (ALPR) is designed to achieve efficient and
accurate vehicle identification by integrating traditional image processing techniques with Optical Character Recognition (OCR).
Unlike resource-intensive deep learning-based solutions, the methodology emphasizes lightweight operations that maintain high
accuracy while ensuring real-time performance and minimal computational overhead. The framework consists of five major
stages: image acquisition, preprocessing, license plate localization, character segmentation, and character recognition and shown
below in Fig. 2.
Fig. 2. Python Based ALPR Architecture
A. Image Acquisition: The first step involves capturing images of vehicles through surveillance cameras, traffic monitoring
systems, or pre-collected datasets. The quality of input images varies significantly due to environmental factors such as lighting
conditions, weather, and motion blur. Hence, the system is designed to handle heterogeneous image inputs with varying
resolutions and noise levels. To ensure adaptability, the acquisition process is compatible with both static images and video
frames, enabling real-time operation in dynamic traffic environments.
B. Preprocessing: Preprocessing is a crucial step for enhancing the quality of raw input images and making license plates more
distinguishable from the background. The preprocessing pipeline includes:
1. Grayscale Conversion – Input images are converted to grayscale to reduce computational complexity and eliminate
unnecessary color information.
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIV, Issue X, October 2025
www.ijltemas.in Page 530
2. Noise Reduction – Filters such as Gaussian blur or bilateral filtering are applied to suppress image noise caused by poor
lighting or low-quality sensors.
3. Contrast Enhancement – Histogram equalization or adaptive contrast techniques are used to improve the visibility of
edges and textual information on the license plate.
4. Edge Detection – Methods such as the Canny edge detector highlight prominent boundaries in the image, facilitating the
localization of rectangular plate regions.
These steps collectively improve the robustness of subsequent operations and ensure accurate plate detection even under
challenging conditions.
C. License Plate Localization: The next stage involves identifying the region of interest (ROI) corresponding to the license
plate. The methodology leverages contour-based techniques for localization:
1. Contour Detection – Candidate regions are extracted using edge information. Contours are analysed to identify
rectangular shapes, which are likely to correspond to license plates.
2. Geometric Filtering – Features such as aspect ratio, area, and rectangularity are evaluated to eliminate false positives.
Only regions matching the expected dimensions of license plates are retained.
3. Segmentation of ROI – The selected region is cropped from the input image, isolating the license plate from the
background and preparing it for character-level processing.
This contour-based approach offers high precision while avoiding the computational burden of deep learning detectors, making it
suitable for real-time systems.
D. Character Segmentation: Once the license plate region is localized, segmentation techniques are applied to separate
individual alphanumeric characters. This step is vital, as incorrect segmentation directly impacts recognition accuracy. The
segmentation process includes:
1. Thresholding – Adaptive thresholding converts the license plate image into a binary representation, distinguishing
characters from the plate background.
2. Morphological Operations – Techniques such as dilation and erosion are used to refine character boundaries and
eliminate small artifacts.
3. Character Isolation – Projection profiles and connected component analysis are applied to detect individual characters,
which are then extracted as separate sub-images for recognition.
By combining thresholding with morphological operations, the framework ensures reliable segmentation even in cases where
characters are closely spaced or partially degraded.
E. Character Recognition: The final stage involves recognizing segmented characters using an Optical Character Recognition
(OCR) engine. In this work, the Tesseract OCR library is integrated due to its compatibility with Python and proven efficiency in
recognizing alphanumeric text. The process consists of:
1. Normalization – Each segmented character is resized and normalized to maintain consistency across inputs.
2. OCR Processing – The normalized characters are fed into Tesseract OCR for recognition.
3. Post-processing – Detected text is validated against region-specific license plate formats using regular expressions or
predefined templates, thereby improving recognition accuracy, and reducing false detections.
This stage outputs the final alphanumeric license plate string, completing the recognition pipeline.
F. System Integration and Real-Time Operation: The entire methodology is implemented in Python using libraries such as
OpenCV for image processing and contour detection, NumPy for matrix operations, and Tesseract OCR for character recognition.
The modular design of the system allows seamless integration with IoT platforms, smart city infrastructures, and cloud-based data
management systems. Furthermore, the lightweight nature of the approach ensures deployment on low-cost hardware with limited
computational resources, such as Raspberry Pi or embedded devices, without compromising accuracy.
IV. Result & Analysis
The proposed Python-based framework for Automatic License Plate Recognition (ALPR) was evaluated on multiple datasets
containing vehicle images captured under diverse environmental conditions, including varying lighting, occlusions, and plate
designs. The evaluation focused on assessing the system’s accuracy, precision, recall, and F1-score for both license plate
detection and character recognition. Additionally, computational performance was measured to verify the framework’s suitability
for real-time deployment. The proposed Python-based Automatic License Plate Recognition (ALPR) framework was evaluated in
a controlled experimental setup to assess its effectiveness under diverse real-world conditions. The experiments were conducted
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIV, Issue X, October 2025
www.ijltemas.in Page 531
using a standard computing environment comprising an Intel Core i7 processor, 16 GB RAM, and Python 3.11, ensuring that the
framework could operate efficiently without requiring specialized hardware such as GPUs. OpenCV was utilized for all image
processing operations, including preprocessing, contour detection, and character segmentation, while Tesseract OCR was
integrated for alphanumeric character recognition. The evaluation was performed on a combination of publicly available license
plate datasets and custom-collected images from urban traffic environments to ensure diversity in terms of vehicle types, plate
sizes, fonts, lighting conditions, and angles. The datasets included high-resolution as well as low-resolution images, simulating
realistic scenarios encountered in surveillance and traffic monitoring. Ground-truth annotations for license plate locations and
corresponding text were used to calculate key performance metrics such as accuracy, precision, recall, and F1-score. To validate
the system’s suitability for real-time deployment, the average processing time per image was recorded, demonstrating the
efficiency of the lightweight Python-based pipeline. The experimental setup was designed to rigorously test the robustness,
scalability, and accuracy of the proposed ALPR system while ensuring reproducibility and practical applicability in real-world
scenarios. To evaluate the effectiveness of the proposed license plate detection and recognition framework, several standard
performance metrics were employed, providing a comprehensive assessment of both detection and recognition
capabilities. Accuracy measures the overall correctness of the system by quantifying the proportion of correctly detected and
recognized license plates relative to the total number of vehicles processed. Precision evaluates the reliability of the system by
calculating the fraction of correctly detected license plates among all regions identified as plates, thereby indicating the rate of
false positives. Recall, on the other hand, reflects the system’s ability to identify all true license plates in the dataset, measuring
the proportion of correctly detected plates against the total number of actual plates and highlighting the occurrence of false
negatives. To balance the trade-off between precision and recall, the F1-score is computed as their harmonic mean, offering a
single metric that represents the overall effectiveness of the detection and recognition pipeline. These metrics are particularly
important for real-world deployment, as high precision ensures fewer incorrect detections, high recall ensures minimal missed
plates, and a high F1-score confirms the robustness and reliability of the system across diverse environmental conditions.
A. License Plate Detection: License plate localization was evaluated on 1,500 vehicle images. The proposed contour-based
approach successfully detected 1,430 plates, resulting in a detection accuracy of 95.3%. The precision and recall for detection
were 96.1% and 94.7%, respectively, indicating that the framework effectively balances false positives and false negatives. The
high F1-score of 95.4% demonstrates the robustness of contour detection combined with geometric filtering in accurately
localizing license plates across diverse conditions. TABLE I. summarizing the performance of the proposed contour-based license
plate detection system on 1,500 vehicle images. Fig. 3. illustrating the performance of the license plate detection system,
showing Accuracy (95.3%), Precision (96.1%), Recall (94.7%), and F1-score (95.4%) for evaluating system effectiveness.
License Plate Detection Analysis
Performance Metric Value
Accuracy 95.30%
Precision 96.10%
Recall 94.70%
F1-score 95.40%
Fig. 3. License Plate Detection on Urban Dataset
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIV, Issue X, October 2025
www.ijltemas.in Page 532
B. Character Recognition: Following plate localization and character segmentation, the OCR module recognized alphanumeric
characters on the plates. Character recognition was evaluated across all segmented characters (approx. 9,000 characters). The
system achieved an overall recognition accuracy of 92.8%, with precision of 93.5%, recall of 92.2%, and an F1-score of 92.8%.
Misrecognitions occurred primarily in cases of low-resolution images, blurred characters, or unusual plate fonts, which slightly
affected the overall performance. Nevertheless, post-processing using template-based validation improved the recognition
reliability and reduced errors. TABLE II. summarizing the performance of the OCR-based character recognition module on
approximately 9,000 segmented license plate characters. Fig. 4. illustrates the performance metrics of character recognition on
the urban dataset, including Accuracy (92.8%), Precision (93.5%), Recall (92.2%), and F1-score (92.8%).
Character Recognition on Urban Dataset
Character Recognition Performance Value
Accuracy 92.80%
Precision 93.50%
Recall 92.20%
F1-score 92.80%
Fig. 4. Character Recognition Performance Analysis on Urban Dataset
The proposed Python-based ALPR framework demonstrates both high computational efficiency and robust performance, making
it suitable for real-time applications in traffic surveillance, parking management, and IoT-enabled smart city systems. The average
processing time per image was measured at approximately 0.42 seconds per frame, highlighting the low computational overhead
of the contour-based license plate localization combined with OCR-based character recognition. This lightweight design allows
the system to operate effectively on standard computing devices without requiring high-end GPUs or complex deep learning
architectures. Experimentally, the framework achieves high accuracy, precision, recall, and F1-score in both license plate
detection and character recognition, maintaining robust performance under diverse environmental conditions, including variable
lighting, different plate orientations, and partial occlusions. The methodology’s efficiency ensures low latency suitable for real-
time deployment, while the Python-based implementation provides flexibility and scalability, allowing seamless integration with
IoT platforms and smart city infrastructure. Overall, the results indicate that the framework strikes a strong balance between
accuracy, speed, and practical deployability. While some limitations remain—such as handling severely blurred or damaged
plates—the proposed approach offers a cost-effective, reliable, and easily deployable solution for real-world ALPR applications
without the computational burden associated with deep learning models.
V. Conclusion
The proposed Python-based framework for vehicle number plate detection and recognition demonstrated robust performance
across multiple evaluation metrics, achieving high accuracy, precision, recall, and F1-score, which underscores its effectiveness
for real-time applications. The integration of advanced image processing techniques and machine learning algorithms enabled
reliable character recognition even under challenging urban conditions, highlighting the framework's practical applicability in
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIV, Issue X, October 2025
www.ijltemas.in Page 533
smart transportation systems and automated monitoring. Furthermore, the system exhibited efficient computational performance,
processing images in near real-time, making it suitable for deployment in dynamic traffic environments. Looking forward, future
work can focus on enhancing system robustness under extreme weather and low-light conditions, integrating deep learning-based
object detection models for improved recognition accuracy, and expanding the framework to handle multiple languages and
regional variations in license plates. Additionally, exploring edge-computing implementations and IoT-based deployment can
further optimize latency, scalability, and real-time decision-making, paving the way for smarter, more autonomous traffic and
surveillance solutions.
References
1. I. Valova, T. Kaneva and N. Valov, "Conceptual Model of a Parking System with Automatic License Plate
Recognition," 2024 IEEE 30th International Symposium for Design and Technology in Electronic Packaging (SIITME),
Sibiu, Romania, 2024, pp. 168-173, doi: 10.1109/SIITME63973.2024.10814907.
2. A. Dayal, A. Porwal, A. Gupta, A. K. Srivastava and M. Mahendru, "Automatic Number Plate Recognition System
using Machine Learning Techniques," 2024 International Conference on Electrical Electronics and Computing
Technologies (ICEECT), Greater Noida, India, 2024, pp. 1-4, doi: 10.1109/ICEECT61758.2024.10739064.
3. X. Du, H. Chen, J. Huan and Y. Du, "License Plate Recognition System Based on Embedded Technology and
YOLOv11," 2025 5th International Symposium on Computer Technology and Information Science (ISCTIS), Xi'an,
China, 2025, pp. 868-871, doi: 10.1109/ISCTIS65944.2025.11065377.
4. K. DC, M. P. Blessy, M. Premalatha and S. Pooja, "Single Shot Detector Based MobileNet for Automatic Detection and
Recognition of License Plates," 2024 International Conference on Advances in Modern Age Technologies for Health
and Engineering Science (AMATHE), Shivamogga, India, 2024, pp. 1-6, doi: 10.1109/AMATHE61652.2024.10582122.
5. T. K. Vashishth, V. Sharma, B. Kumar, S. Chaudhary, R. Panwar and K. K. Sharma, "Guarding Finances: The Role of
Image Processing in Unmasking Counterfeit Currency," 2024 2nd International Conference on Advancement in
Computation & Computer Technologies (InCACCT), Gharuan, India, 2024, pp. 668-671, doi:
10.1109/InCACCT61598.2024.10551141.
6. A. Agarwal, D. Samal, L. Samal, R. K. Nayak and S. N. Mehta, "An Intelligent Video Based Automatic License Plate
Detection and Analysis Using Easy OCR and Tensorflow," 2025 International Conference on Innovations in Intelligent
Systems: Advancements in Computing, Communication, and Cybersecurity (ISAC3), Bhubaneswar, India, 2025, pp. 1-
6, doi: 10.1109/ISAC364032.2025.11156342.
7. S. N. Rafek, S. N. K. Kamarudin and Y. Mahmud, "Deep Learning-Based Car Plate Number Recognition (CPR) in
Videos Stream," 2024 5th International Conference on Artificial Intelligence and Data Sciences (AiDAS), Bangkok,
Thailand, 2024, pp. 538-543, doi: 10.1109/AiDAS63860.2024.10730590.
8. P. C. S. Reddy, P. Shirley Muller, S. N Koka, V. Sharma, N. Sharma and S. Mukherjee, "Detection of Encrypted and
Malicious Network Traffic using Deep Learning," 2023 International Conference on Ambient Intelligence, Knowledge
Informatics and Industrial Electronics (AIKIIE), Ballari, India, 2023, pp. 1-6, doi:
10.1109/AIKIIE60097.2023.10390386.
9. I. Pallempati, D. V. Lakshmi and M. S. Das, "Handwritten Character Recognition and Vehicle Number Recognition
using OCR Method," 2024 Third International Conference on Trends in Electrical, Electronics, and Computer
Engineering (TEECCON), Bangalore, India, 2024, pp. 41-45, doi: 10.1109/TEECCON64024.2024.10939189.
10. N. Thamilarasi, B. Shanmathi, S. Kannagi, G. M, S. K. M and S. P, "Development of an AI-Based Smart Traffic Guard
System for Real-Time Detection and Automated Penalty Enforcement of Traffic Violations," 2025 6th International
Conference on Inventive Research in Computing Applications (ICIRCA), Coimbatore, India, 2025, pp. 804-809, doi:
10.1109/ICIRCA65293.2025.11089648.
11. K. V. S. S. R. Krishna, P. Rajyalakshmi, K. L. Asritha, M. Divyasri and K. S. Chaitragna, "Vehicle Number Plate
Detection using Deep Learning," 2024 International Conference on Integrated Circuits and Communication Systems
(ICICACS), Raichur, India, 2024, pp. 1-5, doi: 10.1109/ICICACS60521.2024.10498324.
12. D. P. Sofiykov and P. V. Georgieva, "A Computer Vision System for Real-Time Vehicular Traffic Management," 2025
MIPRO 48th ICT and Electronics Convention, Opatija, Croatia, 2025, pp. 63-68, doi:
10.1109/MIPRO65660.2025.11131789.
13. K. D. Ismael and E. A. Saeed, "Controlling Mobile Robot Navigation Equipped with Computer Vision Perception for
Text using OCR Technique," 2024 21st International Multi-Conference on Systems, Signals & Devices (SSD), Erbil,
Iraq, 2024, pp. 1-7, doi: 10.1109/SSD61670.2024.10549002.
14. V. Sharma, T. K. Vashishth, S. Kumar, R. Kumar, S. Chaudhary and S. Vidyant, "Next-Gen Child Search: Advanced
Predictions with Facial Features and Skin Color," 2025 Seventh International Conference on Computational Intelligence
andCommunication Technologies (CCICT), Sonepat, India, 2025, pp. 551-556, doi: 10.1109/CCICT65753.2025.00089.
15. J. Kharatmol, N. Visawadia, T. Gurav, S. Daulatkar and K. Dhargalkar, "Extracting Meaningful Data from Education
Credentials Using OCR and Image Processing," 2025 12th International Conference on Computing for Sustainable
Global Development (INDIACom), Delhi, India, 2025, pp. 01-06, doi: 10.23919/INDIACom66777.2025.11115713.
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIV, Issue X, October 2025
www.ijltemas.in Page 534
16. V. Sharma and S. Teotia, “Optimization of Graph Neural Networks for Real-Time Intrusion Detection in Dynamic
Mobile Ad-Hoc Networks,” International Journal of Environmental Sciences, vol. 11, no. 11s, pp. 740–748, Jun. 2025,
doi: 10.64252/79452g17.
17. V. Sharma, A. Yadav, M. Kumar, S. Kumar, S. Kumar, and J. Singh, “A Comparative Study of Deep Learning Models
for Fake News Classification,” International Journal of Latest Technology in Engineering, Management & Applied
Science (IJLTEMAS), vol. 14, no. 9, pp. 188–195, 2025, doi: 10.51583/IJLTEMAS.2025.1409000026.
18. V. Sharma and S. Teotia, “A Comprehensive Analysis of Security Mechanisms and Threat Characterization in Mobile
Ad Hoc Networks,” International Journal of Latest Technology in Engineering, Management & Applied Science
(IJLTEMAS), vol. 14, no. 5, pp. 732–737, May 2025.