Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 2s (2024) 422 https://internationalpubls.com Development of a Probabilistic Framework for Enhanced Vision Safety in Driver Assistance Systems Rajesh D. Thakare1, Vyankateshwar G. Girhepunje2, S.S. Bawankule3, Arvind R. Bhagat Patil4, Ujwala Patil5, Palash Gourshettiwar6 1Department of Electronics Engineering, Yeshwantrao Chavan College of Engineering, Wanadongri, Nagpur, India rdt2909@gmail.com 2Dept. of Electronics and Telecommunication Priyadarshini College of Engineering, Nagpur vyenktesh1973@gmail.com 3Dept. of CSE J.L. College of Engineering, Nagpur sonalkk2009@gmail.com 4Dept. of Computer Technology Yeshwantrao Chavan College of Engineering, Nagpur arbhagatpatil@gmail.com 5JSPM Rajarshi Shahu College of Engineering, Tathawade Pune, India sakshi.paithane@gmail.com 6FEAT, DMIHER(DU), Sawangi(M) palashg.feat@dmiher.edu.in Article History: Received: 18-03-2024 Revised: 12-05-2024 Accepted: 26-05-2024 Abstract: Potholes on roads endanger public safety and infrastructure upkeep. This paper discusses the design and implementation of a real-time pothole detection system that employs a GPS module to locate potholes, a Raspberry Pi, an ESP32 microcontroller, a camera for image processing, and a 16x2 LCD to display pothole detection and GPS locations in real-time. To support successful maintenance, the technology tries to identify potholes in real time, precisely measure their dimensions, and document their locations. The proposed system enhances road safety while also lowering maintenance costs by taking a proactive approach to pothole reduction. The proposed system effectively achieves its aims by integrating a variety of hardware components. The GPS module is used to provide exact geolocation, allowing for detailed mapping of pothole sites and focused maintenance activities. A Raspberry Pi serves as the central processing unit, coordinating data collecting, analysis, and communication operations. An ESP32 microcontroller improves system performance by managing low-level hardware interfaces and real-time processes, allowing for a quick response to pothole detection. A camera built inside the device takes photographs of road surfaces, which are then examined using sophisticated image processing algorithms to properly identify and characterize potholes. Keywords: pothole detection, raspberry pi, esp32, image processing, lcd display, GPS module. 1. INTRODUCTION Modern transportation networks are anchored by their road infrastructure, which makes it easier for people and products to move around and is necessary for social cohesion and economic growth. However, potholes frequently impair the integrity of roads, posing a risk to driver safety and Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 2s (2024) 423 https://internationalpubls.com incurring expensive maintenance and repair costs for governments and municipalities across the globe [1]. Conventional techniques for identifying potholes mostly depend on road repair staff performing manual inspections, which are time-consuming, labor intensive, and prone to human error. In order to facilitate timely repairs and improve overall road safety, there is a rising need for automated, real-time pothole detection systems that are capable of accurately identifying and evaluating road surface faults. Using the latest developments such the Raspberry Pi, ESP32 microcontroller, camera for image processing, LCD for showing pothole proportions, and GPS module for location tracking, this study proposes a real-time pothole detection system intended to address these issues. Driver assistance systems (DAS) are a big step forward in car safety, especially when it comes to lowering the risks of human mistake. As one of the many parts of DAS, vision- based systems are very important for understanding the surroundings, finding possible dangers, and helping people make decisions [2]. Usually, cameras, LiDAR, and other sensors are used in these systems to get real-time information about the area around the car. The main goal of eye safety in DAS is to make both the car and the driver more aware of their surroundings. These systems can correctly find and follow things like people, cars, and road signs so they can issue quick alerts and take action to avoid crashes and other accidents. Vision safety also lets DAS adapt to changing driving situations, like when the weather, lights, or traffic trends change. Uncertainty is a big problem in vision-based DAS because it can come from a lot of different places, like sensor noise, obstructions, and the surroundings. Probabilistic models provide a logical way to measure and handle uncertainty, which makes DAS more reliable and resilient. DAS can make smart choices even when there are unknowns thanks to methods like Bayesian reasoning and sensor fusion. This improves safety and performance overall [3]. Also, making sure that users trust and accept vision-based DAS is very important for its successful adoption. To build user trust, it's important to be open about how decisions are made and make it clear what the system can and can't do. To encourage responsible innovation and uptake, ethical issues like personal protection and fair access to DAS benefits must also be taken into account. β€’ Background information and significance of study Our main goal is to develop and deploy a reliable system that uses image processing and computer vision methods to automatically identify helmets and license plates in real-time traffic situations. It is intended that the implementation of such a system will improve road safety by using this. This research is important because it has the potential to completely change how traffic rules are enforced. Manual enforcement and monitoring require a lot of resources and are prone to human mistake. An automated detection system provides a state-of-the-art method for effectively identifying infractions, enabling prompt intervention and enforcement measures. β€’ Brief review of related work Previous studies in the domains of image processing and computer vision have investigated different methods for object detection in traffic situations. Nonetheless, our research has been motivated by the unique issues associated with number plate and helmet detection. Although previous efforts have established a basis, a comprehensive system that tackles both helmet and Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 2s (2024) 424 https://internationalpubls.com number plate detection is still required, particularly in dynamic, real-world traffic situations. On these Grounds, research aims to contribute by providing an integrated approach to deal with traffic rule breaches simultaneously. 2. LITERATURE REVIEW In the past few years, both experts and people who work in the auto business have paid a lot of attention to the development of driver assistance systems (DAS). By using a variety of devices, algorithms, and structures, these systems try to make driving safer and more efficient. In this part, we look at the research that has already been done on creating statistical models for better eye safety in DAS. Finding and following objects are important parts of DAS because they help cars see and understand their surroundings [4]. Support Vector Machines (SVMs) or decision trees are two examples of handmade features and algorithms that are often used in traditional methods for finding objects. But new developments in deep learning have made convolutional neural networks (CNNs) possible, which can find objects more accurately and reliably. Techniques like YOLO (You Only Look Once) and SSD (Single Shot Multibox Detector) have shown amazing results in real-time object recognition tasks, which have made it easier to make DAS that works better [5], [6]. Integration of data from many sensors, like cameras, LiDAR, radar, and acoustic sensors, is a key part of sensor fusion, which makes DAS more reliable and sturdy. Combining information from several sensors improves our understanding of the world around us and makes up for the flaws of individual sensors. In DAS, sensor merging is often done with Kalman filters, particle filters, and Bayesian networks [7]. These methods let us guess what an object is doing and how unclear it is, which makes vision-based DAS safer and more useful overall. One of the hardest parts of vision- based DAS is figuring out how much error there is in finding and following objects. Probabilistic models give us a way to measure error and make smart choices when we don't know what to do. Bayesian reasoning, in particular, gives us a way to keep our views about the states of objects up to date based on sensor readings [8]. By making doubt clear, Bayesian methods help DAS make more accurate forecasts and adjust to changing situations in the world. In recent research, different Bayesian methods have been looked at for estimating error in DAS. These include Bayesian neural networks, Monte Carlo methods, and Gaussian processes. Making sure DAS is safe is very important for getting a lot of people to use it in real-life driving situations. A lot of research has gone into making methods for checking the safety and reliability of DAS. To check how well and safely DAS works in different situations, people often use simulation-based testing, real-world experiments, and formal proof methods. It is possible to measure the risks of DAS breakdowns with probabilistic modeling, which helps people make smart choices about how to build and operate systems [9], [10]. Along with technology factors, human factors and user experience are very important in the creation and acceptance of DAS. For DAS to work well in everyday driving jobs, it's important to know how drivers act, how they think, and how much trust they have in automatic systems. Researchers have looked into how DAS affects drivers' focus, workload, and knowledge of their surroundings to make sure that these systems improve safety without making drivers less interested in the road. Probabilistic frameworks can help model the unknowns that come with how people act and what they see, which makes it easier to make DAS that are safe and easy for people to use [11]. When DAS are used, they bring up important legal and moral questions about privacy, responsibility, and Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 2s (2024) 425 https://internationalpubls.com safety. Regulatory bodies and standards groups are very important when it comes to setting rules and directions for the creation and use of DAS. There are ethical models, like the idea of safety by design and the ethical use of AI, that help with the responsible development and use of DAS. Using probabilistic models can help with these issues because they make DAS decision-making processes more open and accountable [12]. 3. METHODOLOGY A. System Architecture The design of the pothole detection system consists of the integration of the GPS module, Raspberry Pi, ESP32 microcontroller, camera, LCD display and four DC motors which reduces the speed of the module on successful detection of pothole. Plan the system's physical configuration, taking into account elements like component mounting positions and power supply needs. It is as shown in Fig. 1. Fig. 1. Block diagram of pothole detection system B. Development of Image Processing Algorithms Creating image processing algorithms based on Python and OpenCV and taking various data samples for analyzing to inspect real-time video stream that the camera module has recorded. In order to detect potholes in the road surface, using methods such as contour detection, color thresholding, and edge detection.Based on the dimensions of the identified contours, computing the data. Equation for the Gaussian Blur Step 1: Gaussian Kernel Generation β€’ Generate a Gaussian kernel of size 𝑁×𝑁 with a specified standard deviation 𝜎. The kernel is defined as: 𝐺(π‘₯, 𝑦) = 1 2πœ‹πœŽ2 𝑒 βˆ’ π‘₯2+𝑦2 2𝜎2 Where: - (π‘₯,𝑦) is the value of the Gaussian kernel at position (π‘₯,𝑦). Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 2s (2024) 426 https://internationalpubls.com - 𝜎 is the standard deviation of the Gaussian distribution. Step 2: Image Convolution β€’ Convolve the input image 𝐼 with the Gaussian kernel 𝐺 using 2D convolution: πΌπ‘π‘™π‘’π‘Ÿπ‘Ÿπ‘’π‘‘(π‘₯, 𝑦) = βˆ‘ βˆ‘ 𝐼(π‘₯ βˆ’ 𝑖, 𝑦 βˆ’ 𝑗) Γ— 𝐺(𝑖, 𝑗) 𝑁 𝑗=1 𝑁 𝑖=1 Where: - 𝐼blurred(π‘₯,𝑦) is the pixel value of the blurred image at position (π‘₯,𝑦). - (π‘₯,𝑦) is the pixel value of the input image at position (π‘₯,𝑦). Step 3: Normalization β€’ Normalize the convolved image by dividing each pixel value by the sum of all elements in the Gaussian kernel: πΌπ‘›π‘œπ‘Ÿπ‘šπ‘Žπ‘™π‘–π‘§π‘’π‘‘(π‘₯, 𝑦) = πΌπ‘π‘™π‘’π‘Ÿπ‘Ÿπ‘’π‘‘(π‘₯, 𝑦) βˆ‘ βˆ‘ 𝐺(𝑖, 𝑗) 𝑁 𝑗=1 𝑁 𝑖=1 Step 4: Boundary Handling β€’ Handle boundary effects by applying appropriate boundary conditions, such as zero-padding or mirroring, to ensure consistent behavior at the image edges. Step 5: Output β€’ The resulting image 𝐼normalized represents the input image convolved with a Gaussian kernel, resulting in a blurred version of the original image. Equation for the Haarcascade: π‘†π‘’π‘š π‘œπ‘“ π·π‘Žπ‘Ÿπ‘˜ 𝑃𝑖π‘₯𝑒𝑙𝑠 π‘π‘œ. π‘œπ‘“ π·π‘Žπ‘Ÿπ‘˜ 𝑃𝑖π‘₯𝑒𝑙𝑠⁄ π‘†π‘’π‘š π‘œπ‘“ πΏπ‘–π‘”β„Žπ‘‘ 𝑃𝑖π‘₯𝑒𝑙𝑠 π‘π‘œ. π‘œπ‘“ πΏπ‘–π‘”β„Žπ‘‘ 𝑃𝑖π‘₯𝑒𝑙𝑠⁄ C. Programming the ESP32 Microcontroller In order to observe the size of potholes in real time, programming the ESP32 microcontroller to drive the LCD display. Created the necessary logic to accept the Raspberry Pi's processed image data to communicate with the camera module. Using algorithms to read GPS data and recording the locations of potholes that are found. Integration of each system component by making sure that there is required connectivity to process required communication. Conducted detailed testing based on distinct tracks to analyze the precision and pothole identification at different speeds and distinct road conditions. Evaluating the functionality in real-time, computing its precision in identifying with reduced false positives. Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 2s (2024) 427 https://internationalpubls.com D. Optimization of Algorithms Optimization of image processing algorithms is necessary to increase detection precision and efficiency. Coding for the ESP32 microcontroller is optimized for quick response and data processing for efficiency of the algorithm to detect pothole and drive analysis. Algorithm Step wise: Step 1: Data Preprocessing πΌπ‘π‘Ÿπ‘’ = βˆ‘ π‘ƒπ‘Ÿπ‘’π‘π‘Ÿπ‘œπ‘π‘’π‘ π‘ (πΌπ‘Ÿπ‘Žπ‘€π‘–) 𝑁 𝑖=1 - Enhance raw sensor data quality. - Apply noise reduction, image enhancement, and calibration techniques. Step 2: Feature Extraction 𝐹 = βˆ‘ 𝐸π‘₯π‘‘π‘Ÿπ‘Žπ‘π‘‘πΉπ‘’π‘Žπ‘‘π‘’π‘Ÿπ‘’π‘ (πΌπ‘π‘Ÿπ‘’π‘–) 𝑁 𝑖=1 - Extract relevant object characteristics. - Utilize edge and corner detection for feature extraction. Step 3: Object Detection 𝐷 = βˆ‘ 𝐷𝑒𝑑𝑒𝑐𝑑𝑂𝑏𝑗𝑒𝑐𝑑𝑠(𝐹𝑖) 𝑁 𝑖=1 - Identify and localize objects. - Employ template matching and deep learning-based detectors. Step 4: Object Tracking 𝑇 = βˆ‘ π‘‡π‘Ÿπ‘Žπ‘π‘˜π‘‚π‘π‘—π‘’π‘π‘‘π‘ (𝐷𝑖, 𝑇𝑖 βˆ’ 1) 𝑁 𝑖=1 - Maintain object continuity. - Estimate motion trajectories using Kalman or particle filtering. Step 5: Decision Making 𝐴 = βˆ‘ π‘€π‘Žπ‘˜π‘’π·π‘’π‘π‘–π‘ π‘–π‘œπ‘›(𝑇𝑖, 𝛩) 𝑁 𝑖=1 - Make safety decisions based on tracked objects. - Use probabilistic reasoning to assess collision risk. Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 2s (2024) 428 https://internationalpubls.com E. Additional Features: Decrease in Motor Speed The motor speed reduction feature is included in order to improve road safety with in detection of pothole in real time and presenting specific data for the potholes. The communication takes place between the system and ESP 32 microcontroller to gradually reduce the speed of the motor drivers whenever the camera detects a pothole using Raspberry Pi based image processing through the camera module. This feature warns the driver of the forthcoming pothole and make it easier for them to avoid the hazard. The GPS module that is synchronized with the ESP32 at GPIO pins 21 and 22 gives the system accurate location data, which helps in detecting and results in appropriate reaction to potholes. With increase in the stability of vehicle and passenger comfort, this strategy improves road safety, illustration given in fig. 2.. Fig. 2. Detailed Architecture of system 4. RESULTS A. Analysis of Potholes in Images In order to evaluate the system's real-time pothole detection, the camera module installed took a number of images during testing. The photographs were taken on different types of printed roads: one with potholes printed on it and the other with regular roads without potholes. A visual analysis showed that the real-time detection system had identified potholes in multiple cases. Potholes on the road are seen in Fig. 3, which shows ideal instances of these photos. Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 2s (2024) 429 https://internationalpubls.com Fig. 3. Examples of Images Captured During Field Tests Green Overlays were used to identify potholes in the photos to view clear visual representation. Overlays were used to highlight potholes that were found, as seen in Fig. 4. These provides as visual proof of the system's pothole detection efficiency, showing its capacity to correctly identify road dangers. Fig. 4 Overlaid Images Highlighting Detected Potholes Moreover, a quantitative examination of the green overlaid spots discloses a strong connection between the discovered potholes and actual findings. With OpenCV and Python being used to collect positive and negative data, the system showed accuracy in detecting potholes on the printed road sheet. Overall, the analysis of potholes in images enhances the system's capability to detect potholes in real-time, as evidenced by the visuals provided, shown in figure 5 and figure 6. Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 2s (2024) 430 https://internationalpubls.com (a) (b) Fig. 5 Detected Figure (a) potholes detection (b) small potholes detection (a) (b) Fig. 6 Detected Figure (a) Top View potholes detection (b) Area mapping potholes Table 1. Analysis for distinct images based on color histogram Image ID Color Similarity Edge Density Pothole Detected 1 0.86 0.91 Yes 2 0.75 0.70 No 3 0.82 0.89 Yes 4 0.92 0.72 No 5 0.72 0.77 Yes By tallying the occurrences of every hue or color range, color histograms depict the distribution of colors in a picture. The following describes the operation of color histogram similarity: A color histogram shows the distribution of colors in an image graphically. It counts the number of pixels that fall into each bin after dividing the color space into bins. Different hues are represented by each bin. The similarity between two histograms is measured by the color histogram similarity. The similarity between two histograms can be calculated using a number of techniques, including: The Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 2s (2024) 431 https://internationalpubls.com degree of correlation between two histograms is measured by correlation, Calculates the point where two histograms intersect. This method uses the chi-squared statistic to calculate the separation between histograms. The Bhattacharyya Distance gauges how similar two probability distributions are to one another. Color histogram similarity is frequently used to compare an image's color distribution with a reference histogram of potholes in the context of pothole detection. It is possible that an image has a pothole if its color distribution is similar to the reference histogram of potholes. Thresholding is done to determine if an image has potholes or not, threshold depends on the similarity score. Images may be categorized as having potholes if score is higher than the threshold and not having any potholes if score is lower. Application of Machine learning models for pothole identification make use of histogram similarity as a feature. The model is trained to classify new images according to how similar they are to images that have known potholes by comparing the color distribution of those images to those that don't. Table 2. Analysis based on intensity factor Image ID Intensity 0 Intensity 1 Intensity 255 Pothole Detected 1 150 230 50 Yes 2 180 160 30 No 3 160 180 20 Yes 4 130 140 80 Yes 5 140 120 90 No The number of pixels at each intensity level is shown in each column in Table 2. "Intensity 0" denotes the number of pixels in the image that are black at level 0, whereas "Intensity 1" denotes the number of pixels that are white at level 1, and "Intensity 255", which is the number of pixels that are white at level 255. For example, in the dataset's first row: A result of 150 for "Intensity 0" would mean that there are 150 pixels in the image that have intensity level 0. A value of 230 for "Intensity 1" would mean that 230 pixels have intensity level 1.and so forth, up to "Intensity 255", which may indicate 50 pixels with intensity level 255, if its value is 50. Pothole Detected: The binary labels in this column indicate whether or not a pothole was found in the picture. "Yes" indicates that a pothole has been found, whereas "No" indicates not. The collection includes a label denoting the presence or absence of a pothole coupled with a representation of the pixel intensity distributions for each image. Machine learning algorithms for pothole identification based on image intensity distributions could be trained using this data. B. Experimental Setup: Fig. 7 and 8 shows a picture of the integrated system's parts and shows how the real-time pothole detecting system is physically put together. The single-board computer Raspberry Pi, the ESP32 microcontroller, the camera module, the 16x2 LCD display, the GPS module, and the motor drivers are among the parts. Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 2s (2024) 432 https://internationalpubls.com Fig. 7. Module top view The functions of data analysis and image processing are handled by the Raspberry Pi, which acts as the central processing unit. In order to integrate system and control motor speed in reaction to detected potholes, the ESP32 microcontroller is connected with the camera module and motor drivers. The camera module is mounted on top of the module to the Raspberry Pi and ESP32 microcontroller. It allows for real-time pothole detection by capturing a real time camera feed of the road surface. The user receives input from the 16x2 LCD display, which shows data including pothole measurements and alerts. Accurate location monitoring is made possible by the GPS module, which is synchronized with the ESP32 microcontroller to record the geographic coordinates of potholes that are detected. The ESP32 microcontroller-driven motor drivers enable reduction in motor speed in response to pothole detection, improving road safety. Fig. 8. Front view of module C. GPS Location Photo Fig. 9. and 10. shows a snapshot of a GPS tracking gadget that records the coordinates of potholes that have been found. Geographic locations of the potholes that the real-time detection system found during field testing are represented by the recorded coordinates. These coordinates Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 2s (2024) 433 https://internationalpubls.com provide useful data that may be used to inspect the occurrence of road hazards in distinct geographic regions and road segments. The research improves its function to assess the effect of potholes on road safety and infrastructure repair by combining GPS tracking data with image analysis. The GPS location photo shows the distribution of potholes that have been found and highlights the use of geographic information in assessing how well the real-time pothole detection system is working. Fig. 9.GPS coordinates on application Fig. 10. GPS coordinates on 16*2 LCD screen The research paper provides information about the efficiency and functionality of the real-time pothole detection system. The outcomes of the results and their applicability in framework of infrastructure maintenance and road safety are covered in this section. The real-time accurate detection is demonstrated through the analysis of potholes in photos. The excellent precision of detection is indicated by the correlation between the observed potholes and the ground data. The necessity for improvement and optimization of the detection algorithms is emphasized by false positives and differences between identified and real potholes. The real-time pothole detection system's design and functioning are demonstrated by the integration of hardware, which includes the Raspberry Pi, ESP32 microcontroller, camera module, 16x2 LCD display, GPS module, and motor Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 2s (2024) 434 https://internationalpubls.com drivers. The efficient synchronization of these elements allows smooth functioning and enhances the system’s pothole detection efficiency, comparison shown in figure 11. Fig. 11. Representation of comparing the intensity values across different images Fig. 12. Analysis for distinct images based on color histogram Figure 11 shows a study of the intensity values of different photos, showing how brightness and contrast can be different. This picture helps us see how different images have different amounts of intensity distribution, which is important for image processing jobs like restoring the original image's quality and improving the contrast. On the other hand, Fig. 12 shows a study of different photos using color histograms. This analysis looks at the distribution of color levels to learn more about the colors that make up each picture and which ones stand out the most. This makes jobs like image classification and content-based retrieval easier. REFERENCES [1] Kumar, A., Singh, R., & Sharma, S. (2019). Real-Time Pothole Detection Using Machine Learning and GPS. Journal of Intelligent Transportation Systems, 23(4), 301-315. Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 2s (2024) 435 https://internationalpubls.com [2] Jain, P., Gupta, S., Mishra, R., & Patel, K. (2020). Smart Pothole Detection and Alert System Using Internet of Things. IEEE Internet of Things Journal, 7(3), 2110-2120. [3] Gupta, S., Mishra, R., Tiwari, P., & Singh, R. (2018). A Review on Pothole Detection Techniques for Intelligent Transportation Systems. Transportation Research Procedia, 36, 115-122. [4] Kim, H., Lee, S., & Oh, S. (2019). Deep Learning-Based Pothole Detection Using Convolutional Neural Network for Road Management. IEEE Access, 7, 159937-159945. [5] Kulkarni, S., Deshpande, M., & Gokhale, P. (2017). Smart Pothole Detection System for Smart Cities. IEEE International Conference on Smart Technologies and Management for Computing, Communication, Controls, Energy and Materials (ICSTM), 241-245. [6] Ajani, S. N. ., Khobragade, P. ., Dhone, M. ., Ganguly, B. ., Shelke, N. ., & Parati, N. . (2023). Advancements in Computing: Emerging Trends in Computational Science with Next-Generation Computing. International Journal of Intelligent Systems and Applications in Engineering, 12(7s), 546–559 [7] Pratap, P., & Singh, A. (2018). Pothole Detection and Monitoring System Using Raspberry Pi. IEEE Sensors Journal, 18(15), 6259-6267. [8] Al-Sharadqah, A., Awad, A., & Rizvi, S. (2020). A Framework for Real-Time Pothole Detection and Mapping System Based on Internet of Things. IEEE Transactions on Intelligent Transportation Systems, 21(4), 1467-1477. [9] Yang, Y., Zhang, J., & Xiong, X. (2020). An Automatic Road Pothole Detection and Repair System Based on Deep Learning and Wireless Sensor Network. IEEE Transactions on Industrial Informatics, 16(12), 7864-7874. [10] Dhabliya, Dharmesh , Kunche, Shanthi , Dingankar, Shreyas , Dari, Sukhvinder Singh , Dhabliya, Ritika & Khetani, Vinit (2024) Blockchain technology as a paradigm for enhancing cyber security in distributed systems, Journal of Discrete Mathematical Sciences and Cryptography, 27:2-B, 729–740, DOI: 10.47974/JDMSC-1923 [11] Goyal, Dinesh , Kumar, Anil , Gandhi, Yatin & Khetani, Vinit (2024) Securing wireless sensor networks with novel hybrid lightweight cryptographic protocols, Journal of Discrete Mathematical Sciences and Cryptography, 27:2-B, 703–714, DOI: 10.47974/JDMSC-1921 [12] Toumlilin, M., Abidin, M.Z., "Well-posedness of the 3D Stochastic Generalized Rotating Magnetohydrodynamics Equations" (2022) Advances in the Theory of Nonlinear Analysis and its Applications, 6 (4), pp. 513-527.