ITERATIVE ADAPTATION ALGORITHMS IN MULTICRITERIA TASKS

Based on a survey released by the TomTom Traffic Index in 2018, Indonesia was ranked seventh in the category of the most congested country in the world. One of the factors affecting traffic congestion in Indonesia is an inflexible and conventional traffic management system. In this regard, it is necessary to have a better traffic management system such as a Smart Traffic Light. One way to implement a smart traffic light system is to make a vehicle detection and counting system on the traffic CCTV video automatically. The methods used in this research are Haar Cascade Classifiers and Convolutional Neural Network. Haar Cascade Classifiers have fast computation processes and CNN is applied to validate the detection results of the Haar Cascade method for better accuracy. The average level of accuracy achieved by the system on quiet test data is 82%, normal test data is 69%, and busy test data is 60%. Meanwhile, the average computation time needed by the system for the quiet test data is 0.63 seconds, the normal test data is 0.52 seconds, and the busy test data is 1.05 seconds.

Emergency Medical Services (EMS) require rapid response and efficient routing to ensure timely patient care. However, urban traffic congestion and static routing methods often delay ambulance arrivals. To addressthis, this paper proposes an intelligent ambulance routing and traffic-signal preemption framework, termed SLnO-CC, integrating Sea Lion Optimization (SLnO) for optimal route planning and a Haar Cascade Classifier (CC) for real-time emergency vehicle detection and signal control. The proposed model wasevaluated across eight real-world traffic scenarios within a 15 km urban area, benchmarking against A*, Advanced A* with Dispersion Index, Ant Colony Optimization (ACO), and standalone SLnO. Experimental results demonstrate that SLnO-CC achieved the lowest average response time (9.06 min) and travel time(5.36 min), outperforming A* (9.70 min, 12.20 min) and ACO (9.44 min, 11.47 min) by 6.6% and 13.2%, respectively. In terms of total routing efficiency, SLnO-CC reduced the overall distance and time by 17.8% and 19.6%, respectively, compared with existing baselines. The Haar Cascade–based preemption module achieved 96.8% detection accuracy under varying illumination and occlusion. Overall, the SLnO-CC framework enhances routing adaptability, congestion awareness, and emergency responsiveness—ensuring total response time remains within 10 minutes over a 15 km operational range with high detection reliability.

The aim of our study is finding an efficient artificial intelligence (AI) model for automatic segmentation of carotid artery from ultrasound images which may save time and lower the operator dependency during diagnosis process. A secondary objective is to provide a method also for lowering the operator dependency in carotid artery ultrasound imaging by qualitative mask evaluation and anatomy better understanding using 3D ultrasound reconstructions. A 3D ultrasound prototype based on a high frequency standard ultrasound machine and a pose reading sensor was used to acquire carotid artery and thyroid gland 2D ultrasound images. In-house-developed software applications were used for 2D image manual and automatic segmentation, 3D ultrasound reconstructions and for the 3D visualization of the scanned area. Using the 3D ultrasound reconstructions, for spatial orientation, an experienced operator created the masks remotely from the 2D ultrasound original images for three types of anatomic elements: the carotid artery wall, the lumen of the blood vessel and the thyroid gland. U-Net, U-Net++ and MultiResU-Net models were tested for automatic segmentation on a 2931 images data set. The best segmentation results in terms of IoU (Intersection over Union) and Dice metrics were obtained for the MultiResU-Net architecture: 0.9146 respectively 0.8578. The results from our preliminary study confirmed the future potential of automatic segmentation for the chosen AI models and the fact that 3D reconstructions based on 2D ultrasound images and pose reading sensors information helped the operator to obtain good quality remotely annotated ground truth (GT) masks of the carotid artery and thyroid gland.

This paper addresses the problem of detecting partially occluded objects from 2D images. The detection of partially occluded objects is performed and compared using feature-based training and color-based object segmentation. The occluded objects are very difficult to be detected based only on their features since, all the essential features may not be visible to the learned model due to occlusion. Haar cascade classifier has been utilised for feature-based training and the k-means clustering is utilized for color-based tracking. Various input images are provided Haar classifier as well as the K-means clustering to detect the objects in the 2D images and the subsequent results are compared and analysed. For segmenting the 2D objects using k-means clustering, the average recall and average precision varies from 0.70 to 0.98. The variation is based upon the veracity of the occluded objects. The average precision rate for detecting the occluded 2D objects through the developed method is between 0.24 and 0.60. And it is noted that the average recall for the respective detection lies between 0.25 and 0.70.

Our world is going through a lot of misery due to the outbreak of coronavirus disease (COVID-19) from the end of 2019 affecting many people’s livelihood and causing deaths. Wearing a face mask has become very normal today as the whole world is trying to recover back to normalcy and continue their activities. Therefore, it has become crucial to implement this policy especially in public areas, offices, etc. Constant monitoring cannot be done on every individual. Hence, technology plays a prominent role for a steady implementation. The method consists of two stages of model creation using convolutional neural network (CNN) and model implementation using cascaded classifier (Haar cascade classifier). This model consists of an in-depth reading program. A rigid model is built using TensorFlow, Keras, and OpenCV. By varying the datasets and epoch numbers, we tend to maximize the accuracy and minimize the loss score. This model will be helpful for the authorities to track safety violations, promote the mask usage, and secure the work environment when integrated in various devices.

The steady increase in population and overcrowding has become an unavoidable factor in any public gathering or on the street during any festive occasions. The intelligent monitoring technology has been developing in recent years and human tracking has made a lot of progress. In this paper, we propose a method to manage the crowd by keeping in track the count of the people in the scene. In our study, we develop a system using Raspberry Pi 3 board that consists of ARMv8 CPU that detects the human heads and provide a count of humans in the region using OpenCV-Python. A Haar cascade classifier is trained for human head detection. Human tracking is achieved by indicating the direction of movement of the person. The results of the analysis will be helpful in managing the crowd in any area with high density of crowds.

The paper proposes a self-driving car model also called autonomous, robotic or driver-less car is one that operates and navigates using its intelligence. The basic idea behind the paper is to develop a 1/10 scale RC car to portray an automated car. The model consists of the following software and hardware components such as CNN (Convolutional neural network), Monocular vision algorithm, Haar cascade classifier, Raspberry Pi Board model B+, Pi camera, Arduino, and an Ultrasonic sensor. The Pi camera and ultrasonic sensor are attached to the raspberry pi board to collect input images along with sensor data to stream these data to the server which in our case is the laptop. The (CNN) convolutional neural network running on the server will be used to enable lane detection to provide steering predictions that are left, right, forward based on the input image. The haar cascade classifier will be used to detect signals and stop sign and monocular vision algorithms to calculate distance from them. The ultrasonic sensor will be used for front collision avoidance by stopping the car at a certain distance before the obstacle ahead. The navigation commands such as right, left, forward, stop will be sent to the car through Arduino which is connected to the RC car's remote, this will make the car drive autonomously based on neural network predictions and some hard coded rules. Thus this model will enable autonomous driving by self-navigation via lane detection, stopping at detection of stop signs, red lights and moving on green signal and front collision avoidance in a cost-effective manner. Hardcoded rules, obstacle avoidance mechanisms, machine learning models, and smart object discrimination will help the system follow traffic rules and navigate the car using artificial intelligence.

The purpose of the paper is to recognize an abducted child from the photos of children available, with the help of face recognition and face matching technique and to adopt an abandoned child from the data available, with the help of string comparing technique. The work in this paper is extended by providing the concept of an adoption system for the abandoned child. Haar cascade classifier, an OpenCV classifier can be used for face recognition, the mean method can be used for face matching, and DB can be used for string comparisons. The dataset consists of 500 images for child recognition and child matching, and 50 images for child adoption. The accuracy obtained from the face recognition algorithm is 80.7%.KeywordsHaar cascade classifierJSON modelOpenCV and DB

The Face -Based Attendance System is an innovative solution designed to automate and streamline the process of recording attendance in academic institutions and workplaces. Traditional methods, such as manual sign-ins and RFID-based systems, are often time-consuming and susceptible to proxy attendance. This project leverages advanced computer vision and machine learning techniques, particularly using Open CV and the Haar Cascade Classifier, to detect and recognize faces in real time. The system captures live video input through a webcam, processes the facial features, and matches them with stored profiles to mark attendance automatically. This approach not only improves accuracy and efficiency but also enhances security and reduces administrative overhead. The system is implemented using Python and integrates a graphical user interface (GUI) for ease of use. Overall, this project demonstrates the potential of facial recognition technology in creating smart, secure, and contact-less attendance systems. Keywords: Face recognition, attendance system, computer vision, machine learning, OpenCV, Haar Cascade Classifier, real-time detection, Python, automation, contactless attendance.

This paper proposes an eye state detection system using Haar Cascade Classifier and Circular Hough Transform. Our proposed system first detects the face and then the eyes using Haar Cascade Classifiers, which differentiate between opened and closed eyes. Circular Hough Transform (CHT) is used to detect the circular shape of the eye and make sure that the eye is detected correctly by the classifiers. The accuracy of the eye detection is 98.56% on our database which contains 2856 images for opened eye and 2384 images for closed eye. The system works on several stages and is fully automatic. The eye state detection system was tested by several people, and the accuracy of the proposed system is 96.96%.

Accidents related with vulnerable pedestrians around crosswalks are continued so that proactive safety support system is required. Pedestrian detection from frames captured by a camera is a significant and yet challenging task. An autonomous road crossing surveillance system would be ideal for tracking pedestrians who want to cross the road and assist them. A practical solution for aiding pedestrians regularly, a road crossing surveillance with real-time Pedestrian Detection. Since the background subtraction from videos and images is still a persistent problem and difficult to accomplish. A Haar Cascade Classifier with the full-body detection algorithm is used to detect people in real-time captured by a camera. Keywords: Pedestrian crossing, surveillance, road crossing surveillance, surveillance system, autonomous, Haar Cascade Classifier.

Traditional attendance systems often encounter challenges in efficiently and accurately recording attendance. This research aims to introduce an innovative solution through the development of an intelligent anti-spoofing attendance system based on facial recognition using the Haar Cascade Classifier method. Designed to overcome the inefficiencies in attendance recording, this system ensures the accuracy of educational staff attendance records. Its development method relies on the Haar Cascade Classifier, employing image processing to detect learned object features, particularly focusing on facial recognition. Research findings indicate that the implementation of this system achieves an average accuracy rate of 98.90% in attendance recording. The facial recognition technology ensures reliable attendance recording with confidence levels exceeding 80%, signifying precise facial identification that addresses various challenges and ensures attendance data integrity. Not only does the system identify educational staff with high accuracy, but it also provides prompt responses for efficient attendance logging and verification. Beyond its technical benefits, this study significantly contributes to the development of smarter and more efficient attendance technology. The system plays a crucial role in enhancing the discipline of educational staff at STMIK IKMI Cirebon and streamlining attendance management and evaluation across educational institutions.

Traffic violations are common in the area zebra crossing at the location of traffic lights, violations are generally caused by the negligence of motorists who do not comply with existing regulations. As a result, there were many traffic accidents that could have been avoided. Starting from this problem, a simulation model tool is needed that is able to provide direct appeals to drivers when a violation occurs. Based on this, the purpose of this research is to create a simulation model to prevent accidents caused by motorists who are negligent of the rules by identifying vehicles that stop crossing the line boundaries. zebracross or drivers waiting for an out of place traffic light using imagery HaarCascade Classifiers which is programmed using Python and OpenCV to help process digital images. The results of this study that the Haar Cascade Classifier algorithm can characterize motorbikes well, with the best average accuracy value of 91.5% at a resolution of 720p. Based on the results obtained shows that the algorithm haar cascade classifier can detect violations that exist at traffic lights. Hence this indicates the rate of traffic violations can reduce by utilizing the algorithm haar cascade classifier.

Intima-media thickness (IMT) is now being considered as an indicator of atherosclerosis. Our group has developed several feature-based IMT measurement algorithms such as the Completely Automated Layer EXtraction (CALEX) (which is a class of patented AtheroEdge Systems from Global Biomedical Technologies, Inc., CA, USA). These methods are based on the hypothesis that the highest pixel intensities are in the far wall of the common carotid artery (CCA) or the internal carotid artery (ICA). In this paper, we verify that this hypothesis holds true for B-mode longitudinal ultrasound (US) images of the carotid wall. This patented methodology consists of generating the composite image (the arithmetic sum of images) from the database by first registering the carotid image frames with respect to a nearly straight carotid-artery frame from the same database using: (1) B-spline-based nonrigid registration and (2) affine registration. Prior to registration, we segment the carotid-artery lumen using a level-set-based algorithm followed by morphological image processing. The binary lumen images are registered, and the transformations are applied to the original grayscale CCA images. We evaluated our technique using a database of 200 common carotid images of normal and pathologic carotids. The composite image presented the highest intensity distribution in the far wall of the CCA/ICA, validating our hypothesis. We have also demonstrated the accuracy and improvement in the IMT segmentation result with our CALEX 3.0 system. The CALEX system, when run on newly acquired US images, shows the IMT error of about 30 μm. Thus, we have shown that the CALEX algorithm is able to exploit the far-wall brightness for accurate IMT measurements.

This paper deals with the passive location problem for coherent near-field sources with the non-uniform sparse linear array (non-USLA). Specifically, we divide the non-USLA into three subarrays and exploit the fact that the coherent sources are located in near field for the whole array, whereas they are regarded as far-field targets with respect to the subarray 1/2 among the non-USLA. We first roughly derive azimuth angles with the subarray 1/2 using iterative adaptive algorithm (IAA). Second, based on the obtained angles, the possible locations of coherent sources are derived according to the law of Sines. Finally, the accurate location estimations of the sources are achieved with the whole non-USLA through searching around the possible locations using IAA. At the analysis stage, we assess the effectiveness of the proposed method through numerical simulations showing its capability to obtain the locations of near-field coherent sources accurately and effectively.