OpenCV Library Research Articles

ЭКСПЕРИМЕНТАЛЬНАЯ ОЦЕНКА ТОЧНОСТИ ПРОСТРАНСТВЕННОЙ СТАБИЛИЗАЦИИ КВАДРОКОПТЕРА DJI AIR 2S*

The paper proposes a method for assessing the actual accuracy of spatial stabilization of a DJI Air 2S quadcopter using its video camera and processing the resulting video sequences with a developed software package in the mode of hovering over a reference object. The software package was implemented using the OpenCV library in Python. To register deviations of stabilization systems from a given position, it is proposed to use a reference field with known geometric parameters, information about which is entered into the processing algorithm. The main task of the software package is to identify the boundaries of the reference field and measure the geometric sides in pixels which allows you to determine the center of the field and measure its deviation relative to the center of the image as well as determine the height of the unmanned aerial vehicle flight. In the work the height measurement algorithm was tested. This technique is used to evaluate the accuracy of quadcopter stabilization with various combinations of navigation sensors used for this purpose. A statistical analysis of the obtained results was carried out and conclusions were drawn about compliance with the characteristics of positioning accuracy declared by the manufacturer using optical obstacle avoidance systems and satellite navigation systems.

Comparative Analysis of Algorithms for Laser Line Identification for 3D Scanning Devices

A comparative analysis of algorithms for laser line center determination and recognition has been conducted. The issues in this area are described, outlining the principles, pros, and cons of each method. Additionally, the possibilities of their application using programming code are demonstrated. Algorithm testing was performed using Python language tools and the OpenCV library. It is shown that the quality of the result in the extremum method significantly depends on the quality of the selected parameters for the Butterworth filter. In contrast, the gray gravity method substantially relies on the accuracy of determining the laser incidence angle. The computational efficiency and accuracy for each algorithm are also analyzed.

Intelligent Face Recognition Based Students’ Attendance System

This paper aims at developing an intelligent face recognition students’ attendance system to enhance school attendance tracking. The system is made up of four phases- database of students’ details, face detection, face recognition and attendance report. The database stores all the students’ details and images of the students’ face captured while face detection and recognition is carried out with convolutional neural network algorithms from the face recognition and opencv library. As faces are detected and recognized from live streaming video of the classroom, attendance are recorded into an excel file and then sent to a real time database. The system also contains a mobile interface through which the course instructors can access information at all time. The methodology adopted for this work is object-oriented analysis and design methodology (OOADM) while programming language used is Python. This work has helped immensely to eliminate the issue of proxy attendance as well as reduce the time wasted in manual attendance system. It is very beneficial to schools and other institutions where attendance is required.

Использование внешней системы компьютерного зрения для управления группой мобильных роботов

The paper considers a centralized approach to the joint management of a group of robots using an external computer vision system. Developed algorithms and path planning software was experimentally tested on simulated work sites in the V-REP robotic modeling software environment for a group of three mobile robots, the control was performed in the Jupyter-notebook program using the OpenCV library.

НЕЙРОСЕТЕВАЯ СИСТЕМА РАСЧЕТА УРОЖАЙНОСТИ ПОДСОЛНЕЧНИКА

The article is devoted to the solution of the problem of developing a neural network system for calculating the yield of sunflower. The developed system is designed to simplify, automate and speed up the inspector's work on calculating the biological yield of sunflower. When assessing the biological yield, most of the time is taken by counting the seeds of a sunflower head. It is proposed to solve this problem using a trained convolutional neural network. The developed system allows you to add the name of a new calculation area, use phone cameras to obtain images of sunflowers (for mobile applications), add images to the system, search for and select the required subspecies of sunflower from the directory, obtain the values of the average weight of seeds of the selected subtype of sunflower, set the area of the determined sowing field, calculate the yield based on the added images, receive the result of calculating the biological yield of sunflower, and also receive previously calculated yield values by name from the device's memory and by the year of the assessment. The Python programming language was chosen to solve the problem of counting seeds in a sunflower head and developing a neural network system. The Roboflow system was used to create a sample with segmented images of sunflower heads. TenserFlow was used as a machine learning library. Keras and OpenCV libraries were used for segmentation and object recognition in the image. Google Colab was used as the development environment for the neural network system. The mobile application was implemented in Java using the Android Studio development environment. The desktop application was implemented in C# using the Visual Studio development environment on Windows Forms technology. The developed neural network system, based on the image transferred to the input of the neural network, detects a sunflower head, segments part of the image with the head, cutting out this area, then counts the seeds in the head and calculates the yield. The YOLOv8 convolutional neural network was used to segment the sunflower head. Using the Roboflow service, a data sample of 10,000 sunflower images was formed. Part of the data (70%) was used for training, and the rest of the data was used for testing the model. Testing of the neural network system showed its efficiency and the possibility of further practical use. Thus, the developed system allows accelerating the assessment of sunflower yield by automating this process.

СВЕРТОЧНАЯ НЕЙРОСЕТЕВАЯ МОДЕЛЬ РАСПОЗНАВАНИЯ РУКОПИСНЫХ МАТЕМАТИЧЕСКИХ ВЫРАЖЕНИЙ

This article is devoted to solving the problem of developing a convolutional neural network model for recognizing handwritten mathematical expressions consisting of Arabic numerals and basic symbols of mathematical operations. Its solution required the formation of training and test samples, development of the structure of the neural network and its training, selection of a method for segmenting the image into individual characters, as well as testing and evaluating the results of the convolutional neural network model. In the work, image processing and classification was carried out using a multilayer convolutional neural network. To build a convolutional neural network model, the deep learning library Keras was used in this work. To train the model, a set of handwritten mathematical symbols was used, obtained using the publicly available machine learning library FastAI and consisting of 10,379 images of handwritten mathematical expressions. To test the equation classifier, a random expression generator was created based on the initial data of the handwritten training sample. Examples of generated mathematical expressions are shown. To simplify the learning process, all images were converted to binary format, saving it in black and white. Thus, image processing required the steps of generating data, receiving input images, scaling them, building an architecture and training a convolutional neural network, as well as its practical use. The result of image processing is to obtain a table of probabilities that the image belongs to a particular class of numbers and expressions of mathematical operations. To divide the images into different areas, a segmentation process was carried out using contour analysis methods. The OpenCV library was used to detect the outlines of numbers and mathematical symbols. Image segmentation required the steps of converting the gray image to binary format, finding contours using the findContours function, and checking for contour intersections. After detecting objects in the image, elements are extracted from the mathematical expression, analyzed using a trained neural network, and the string is digitized. The constructed neural network model showed high accuracy of image recognition of mathematical expressions on training and test data samples. The recognition accuracy was no less than 98%.

AI-Based Self-Learning Robotic Arm Using Microcontroller

Robotic arms are essential components of day-to-day life likely manufacturing and logistics operations to provide more speed, more flexible decision-making and reduce the possibility of errors. The objective of this project is to make a robotic arm with machine learning features to automate the assembly by object and colour detection using openCV, and TensorFlow library. The basic arm is made of light material to provide maximum mobility which is controlled by an ATmega328P microcontroller with the aid of an Arduino development environment for hassle-free interfacing sensors. A webcam is used for sensing the surroundings. Feed from waveform is processed by openCV and TensorFlow library Using a Python environment the real-time coordinates data and colour information will be fed to the microcontroller unit that controls the arm servos accordingly to grip the appropriate object and place them in the predetermined location. The arm is also capable of learning trajectory which is taught to it using human hands. The use of easily available components reduces the cost but the same time leaves much room for future improvement. There are two approaches to machine learning used in robotic systems, the first one is a conventional approach to machine learning which is a complicated, expansive asset that is not suitable for several types of automation industries. A self-learning approach, called a behaviour-based robotic arm with a machine learning feature to automate the assembly by the object and colour detection using openCV, a python environment to be built to address the limitations of conventional robotics methodologies. This approach aims to enhance the capability of robotic platforms to offer more flexible decision-making, cost-effectiveness, reduced human error, and higher precision in processes.

НЕЙРОСЕТЕВАЯ МОДЕЛЬ РАСПОЗНАВАНИЯ ЛЕСНЫХ ПОЖАРОВ

This article is devoted to the development of a convolutional neural network model for recognizing forest fires from images. A review of approaches for recognizing forest fires is made: based on convolutional neural networks, recurrent neural networks, attention neural networks, graph deep neural networks, as well as object detectors. To solve this problem, the feasibility of using convolutional neural networks is justified due to a number of their advantages. The formal formulation of the problem and the stages of implementing a convolutional neural network model are considered: preparing data for analysis, choosing software tools, as well as building, training and testing the model. The stages of collecting and preparing data for analysis are considered. To obtain initial data manually using search queries, images of forest fires belonging to the classes “Fire” and “No fire” were found, presented from different angles, at different scales and times of day. Data preprocessing has been carried out: changing the image size, changing the contrast, rotating. Examples of images characterizing forest fires are presented. Training, validation and test data samples were generated. The amount of data obtained for each image class (“Fire” and “No Fire”) is shown. To build the neural network model, the Python 3.10 programming language version was used. The libraries chosen for writing program code were TensorFlow, Keras, NumPy, OpenCV, Matplotlib and Python Imaging Library. The architecture of the constructed model, consisting of 10 layers, is presented. A convolutional neural network model was trained and tested in the Kaggle Notebook cloud computing environment. Model training graphs are presented. An example of how the model works when recognizing images of forest fires is shown. The accuracy of forest fire recognition during training on the validation set was 97%. The accuracy of the model on test data was 94%. The results of the research allowed us to draw a conclusion about the adequacy of the constructed model. It is concluded that the constructed model can be effectively used as a tool for solving the problem of recognizing forest fires as part of an intelligent computer vision system.

Novel Artificial Intelligence Tool for Real-time Patient Identification to Prevent Misidentification in Health Care.

Errors in the identification of true patients in a health-care facility may result in the wrong dose or dosage being given to the wrong patient at the wrong site during radiotherapy sessions, radiopharmaceutical administration, radiological scans, etc. The aim of this article is to reduce the error in the identification of correct patients by implementation of the Python deep learning-based real-time patient identification program. The authors utilized and installed Anaconda Prompt (miniconda 3), Python (version 3.9.12), and Visual Studio Code (version 1.71.0) for the design of the patient identification program. In the field of view, the area of interest is merely face detection. The overall performance of the developed program is accomplished over three steps, namely image data collection, data transfer, and data analysis, respectively. The patient identification tool was developed using the OpenCV library for face recognition. This program provides real-time patient identification information, together with the other preset parameters such as disease site, with a precision of 0.92%, recall rate of 0.80%, and specificity of 0.90%. Furthermore, the accuracy of the program was found to be 0.84%. The output of the in-house developed program as "Unknown" is provided if a patient's relative or an unknown person is found in restricted region. This Python-based program is beneficial for confirming the patient's identity, without manual interventions, just before therapy, administering medications, and starting other medical procedures, among other things, to prevent unintended medical and health-related complications that may arise as a result of misidentification.

Smart Trash Bin to Prevent Animal Disturbance Using Raspberry Pi and Deep Learning

A trash can is an indispensable component of human daily life. It is ubiquitous and extremely useful for keeping the city clean. However, littering is a frequent and widespread problem. Not only do humans litter, but so do animals, particularly monkeys. Animals are constantly searching for food, and when we discard our food waste, they are attracted and leave their natural habitats. Following that, the user may forget to lock the trash can, which may end up in the hands of the animal. With the aid of a deep learning system that detects and identifies human subjects, a smart trash bin with an autonomous locking mechanism will be used in this project to reduce animal disturbance. In this paper, Raspberry-Pi 4 module is being used. Moreover, this article describes a human subject identification method and its real-time displacement using the OpenCV library of programming roles, which is primarily aimed at Raspberry Pi with camera module. The smart trash bin system helps to reduce animal disturbance and minimise littering.

Performance Evaluation of Face Recognition System (FRS) Developed with Haar Cascade and MongoDB Integration in Recognition of Covered Faces

Although traditional face recognition systems (FRS) can detect with a certain success rate whether a mask is worn, they may fail due to the fact that most of the faces of the people who wear masks are covered. The difficulties arising from the fact that a significant part of the faces of individuals wearing masks are covered limits the performance of existing FRSs. In this research, it is aimed to integrate the Haar Cascade method with the MongoDB database in real time using the OpenCV library for mask-wearing face recognition and to demonstrate its performance with extensive experiments. In the experiments, the data set created within the scope of this study from realistic face images, in which most of the masked faces are covered, was used. Our research has shown that the accuracy of face recognition is 85% for masked faces, 61% for unmasked faces, and 41% when half of the face is covered by a different object. It is considered that this study will contribute to the literature in terms of providing a more effective and applicable mask detection solution by combining the Haar Cascade method with real-time database management integration.

Enhancing User Authentication with Facial Recognition and Feature-Based Credentials

This research proposes a novel and trustworthy user authentication method that creates individualized and trusted credentials based on distinctive facial traits using facial recognition technology. The ability to easily validate user identification across various login methods is provided by this feature. The fundamental elements of this system are face recognition, feature extraction, and the hashing of characteristics to produce usernames and passwords. This method makes use of the OpenCV library, which is free software for computer vision. Additionally, it employs Hashlib for secure hashing and Image-based Deep Learning for Identification (IDLI) technology to extract facial tags. For increased security and dependability, the system mandates a maximum of ten characters for users and passwords. By imposing this restriction, the system increases its resilience by reducing any possible weaknesses in its defense. The policy also generates certificates that are neatly arranged in an Excel file for easy access and management. To improve user data and provide reliable biometric authentication, this study intends to create and implement a recognition system that incorporates cutting-edge approaches such as face feature extraction, feature hashing, and password creation. Additionally, the system has robust security features using face recognition.

Enhancing healthcare data security: a two-step authentication scheme with cloud technology and blockchain

In the modern world, medical data leakage has many external and internal threats. Information systems of medical organizations are constantly subject to various types of cyber-attacks and unauthorized penetration attempts, which leads to the publication of patient medical data online. Existing authentication schemes using blockchain technologies in medical organization systems ensure the integrity of medical data and secure access to patient data. However, one of the serious reasons for unauthorized access to the healthcare system is the human factor, which manifests itself in a negligent attitude towards account security, non-compliance with the rules and policies of information security, and transferring to third parties personal login details to the information system of a medical organization. This paper proposes a solution to this problem through an improved two-step authentication scheme using cloud technology and blockchain. The combined use of cloud technologies and blockchain is a distinctive feature of the proposed authentication scheme since it provides two levels of protection: 1) two-step authentication, the second stage of which includes biometrics through a mobile application. It prevents unauthorized access to the system by third parties; 2) cloud encryption keys for decrypting medical data, which are also accessed through the user's biometrics. The practical part of the paper includes the implementation of biometric login in Python using the OpenCV library. As a result of the practical part, unique fingerprint samples were obtained. The biometric user verification algorithm is designed for a mobile application, which we plan to implement in the future

Gesture-Based Control of Multimedia Player Using Python and OpenCV

The increasing significance of computers in our daily lives, coupled with the rise of ubiquitous computing, has necessitated effective human-computer interaction. Hand gesture recognition systems have emerged as a real-time video-based solution for detecting and interpreting hand gestures, offering intelligent and natural human-computer interaction (HCI) methods. This project focuses on leveraging human hands as input devices for computer operation. Developed using Python and the OpenCV library, the program utilizes a computer webcam to capture and analyze hand shapes and patterns. The program provides real-time feedback by displaying recognized hand gestures on the live video stream. The ultimate outcome of this project is an application that enhances user experiences in contactless systems. The project recognizes and detects human hand motions using the Python computer language through a process flow that includes background subtraction, hand ROI segmentation, contour detection, and finger recognition. Techniques for processing images are used, including hand gesture detection, pattern recognition, thresholding, and contour detection. The processing of incoming photos and the creation of related keystrokes are made possible by OpenCV, a rich set of image processing tools

Harnessing Deep Learning in Computer Vision for Effective Face Mask Detection in the Global COVID-19 Crisis

Objective: This study aims to explore the use of deep learning algorithms in computer vision to address the effective detection of face mask use in the global COVID-19 crisis.
 Results: Deep learning techniques enable the execution of tasks by computers and smart devices without human intervention, including image identification and predictions. Its applications have demonstrated significant advances in several areas, particularly in engineering and health research. The study highlights that computer vision scientists can contribute to the prevention, control and management of the fight against COVID-19 and other airborne viruses. Computer vision employs algorithmic tools to process images, perform associations and transmit relevant information.
 Implications for Research, Practice and Social: Construction of a deep learning model using a dataset of people with and without face masks. The developed model, implemented in Python with the help of OpenCV, Keras and TensorFlow libraries, presented highly promising results, reaching an accuracy of 99% in predictions in test images.
 Originality/Value: This study highlights the originality and value of deep learning techniques in computer vision as an effective means of tackling virus-borne pandemics such as COVID-19, and contributing to a preventive, efficient and cost-effective approach to use of face masks.
 Keywords: Computer vision; Convolutional Neural Network; Face mask.

Interactive Hand Gesture Control System for Augmented Reality-based Games

This research paper presents the development of an innovative 2D augmented reality (AR) car game utilizing Python, OpenCV, and Pygame. The primary objective of the game is to provide an interactive platform for individuals with motor skill challenges, such as Parkinson’s disease, to enhance their hand and gesture control abilities. The game involves controlling a virtual car superimposed onto the real-world environment using an AR marker. Hand gestures captured by a webcam control the car's movements, creating a novel and accessible gaming experience. The implementation incorporates computer vision techniques from the OpenCV library to detect an AR marker, interpret hand gestures, and map them to specific actions within the game. The Pygame library facilitates the creation of an engaging gaming environment, where users can employ intuitive hand movements to navigate the virtual car. The system's performance is evaluated in varying lighting conditions, and the results demonstrate a promising accuracy rate in gesture detection. The limitations and potential improvements are discussed, emphasizing the system's application in assisting individuals with motor impairments and its broader implications for human-computer interaction. This research provides insights into the intersection of augmented reality, computer vision, and assistive technology, offering a foundation for further advancements in accessible gaming experiences. Keywords: Augmented Reality, Hand Gesture Control, Computer Vision, Game Interaction, Assistive Technology

Overcoming optical image challenges in automatic grain size measurement using a novel computer vision algorithm applied to hot deformation of Al-Zn-Mg Powder metallurgy alloy

The present study introduces a novel approach that utilizes the Python programming language along with the OpenCV, SciPy, and NumPy libraries. This approach addresses challenges related to grain size measurement in optical images, including issues such as fused grains caused by fragmented boundaries, noise resulting from lens aberrations, and grains located at the image edges. These challenges have been successfully overcome, enabling the segmentation of microstructures and the automated determination of grain characteristics, including grain counting, ASTM grain size, and grain size in µm. The developed algorithm was validated using optical images obtained after hot deformation, and a comparative analysis was performed with manual measurements. The results showed an average absolute relative error (AARE) of 3.1% for the number of grains, 0.19% for ASTM grain size, and 1.59% for grain size in µm, with an R-value of 0.999.

ALGORITMO PARA A DETECÇÃO DE ÁRVORES URBANAS A PARTIR DE IMAGENS 360

Trees are indispensable to human life, they absorb carbon dioxide and release oxygen, protect ecosystems, reduce erosion and help to reduce the environment temperature. The manual identification of trees on public roads requires expense and time for recording and managing the data collected, since urban regions can be very large. We developed in this paper a method for the trees detection in urban areas from a 360 video. A YOLO neural network was trained to detect the trees from frames of the equirectangular video (360 images). We used Computer Vision techniques with the OpenCV library to develop algorithms to segment the regions that fit the detected trees in the rectilinear field of view (gnomonic projection), in order to verify if the trees are on the sidewalks. The results obtained showed around 80% success in detecting trees using YOLO, and an accuracy of 71% in the algorithm that checks if the trees are on the sidewalk.

Biometric Authentication based on Liveness Detection Using Face Landmarks and Deep Learning Model

This paper describes the approach to active liveness detection of the face using facial features and movements. The project aims to create a better method for detecting liveness in real-time on an application programming interface (API) server. The project is built using Python programming with the computer vision libraries OpenCV, dlib and MediaPipe and the deep learning library Tensorflow. There are five modules in active liveness detection progress related to different parts or movements on the face: headshakes, nodding, eye blinks, smiles, and mouths. The functionality of modules runs through face landmarking through dlib and MediaPipe and detection of face features through Tensorflow Convolutional Neural Network (CNN) trained in two different approaches: smile detection and eye-blink detection. The result of implementing face landmarking shows an accurate result through the pre-trained model of MediaPipe and the pre-trained parameter of the dlib 68 landmarking model. And more than 90% classification model accuracy in precision, recall, and f1-score for both trained CNNs in detecting smiles and eyes blinking through the Scikit-Learn classification report. In addition, the prototype API is also implemented using the Python RESTful API library, FastAPI, to test the detection functionality in the prototype Android application. The prototype result is outstanding, as the model excellently requests and retrieves from the API server. The possible research path gives the success of real-time detection on API servers for easy implementation of liveness detection on low-spec client devices.

Detection of White Blood Cells Using Algorithms that Analyze the Binary Properties of Red, Blue, and Hue Components

In this paper, a unique method for the identification of white blood cells (WBCs) is proposed. The system is based on morphological analysis and makes use of the red, blue, and hue components of digital pictures. The purpose is to design a technology that is both accurate and efficient for the detection of white blood cells, which is essential for a variety of medical diagnosis. Preprocessing of the pictures, segmentation of white blood cells (WBCs), feature extraction based on morphological features, and classification using machine learning approaches are all components of the procedure. Method: Increasing contrast and removing noise are the first steps in the proposed technique, which starts with picture preparation. In order to separate white blood cells (WBCs) from the background, the next step is to conduct segmentation, which involves a mix of thresholding and morphological processes. After the white blood cells have been segmented, characteristics such as area, perimeter, circularity, and intensity are retrieved from them. These attributes are given into a machine learning classifier, trained on a dataset of labeled WBC pictures, to discriminate between WBCs and other cells or artifacts. Python as well as OpenCV libraries are used in the implementation of the technique. Result: In the process of identifying white blood cells (WBCs) from microscopic pictures, the suggested algorithm yields encouraging results. In the identification of white blood cells (WBC), the evaluation on a dataset consisting of a variety of blood samples reveals good accuracy, sensitivity, and specificity. With regard to both accuracy and computing efficiency, the algorithm's performance is superior to that of other approaches that are currently in use. Moreover, the system demonstrates a high level of stability when confronted with differences in picture quality and staining processes. Conclusion: In conclusion, the algorithm that was created offers a dependable and automated method for the detection of white blood cells (WBC) based on the identification of morphological characteristics in digital pictures. Through the use of the red, blue, and hue components, it does an excellent job of distinguishing white blood cells from other cellular components. The precision and efficiency of the approach make it acceptable for incorporation into clinical workflows, which will help in the identification of a variety of blood illnesses in a timely and accurate manner using the method. It is possible that further improvements and validation on bigger datasets might make its use in clinical settings easier, which would thus contribute to improvements in patient care and diagnosis.