Open Source Computer Vision Research Articles

Defect monitoring in additive manufacturing through OpenCV-based remote video access

Errors in additive manufacturing are a significant obstacle to its widespread adoption in various application. The developments of 3D printed parts are prone to various types of defects, which are difficult to monitor in real-time through human vision. The use of 3D printers has increased in diverse fields, leading to a significant rise in the need of continuous monitoring during the manufacturing process. Prompt action should be taken to stop or correct errors immediately to prevent a defective part. However, continuous monitoring of 3D printer through human vision is impractical for longer tasks and cannot be accurate. Human inspection is not infallible and could introduce its own errors. Given the multitude of application, continuous defect monitoring is extremely challenging, particularly for longer tasks. Based on these constraints a suitable remote monitoring system through open-source computer vision has been proposed. Incorporating the concept of Internet of things, the video of printing process could be viewed by a user or controller from a remote location enabling to monitor the printing process. With the use of Haar classifier in the program, the printing defects that occur in the 3D printed part could be identified immediately through contours drawn over the defective part. Depending upon the severity of defect indicated, the user could take respective action to control the defect formation. With Industry 4.0 in the implementation stages, this method of video monitoring of defects in 3D printers could be an effective method for widespread use of additive manufacturing systems in industries.

Facial Recognition Attendance Monitoring System using Deep Learning Techniques

The Facial Recognition Attendance Monitoring System employing Deep Learning Techniques represents a cutting-edge application of artificial intelligence in educational and corporate environments. The implementation of a Facial Recognition System can aid in identifying or verifying a person's identity from a digital image. Accurate attendance records are vital to classroom evaluation. However, manual attendance tracking can result in errors, missed students, or duplicate entries. The adoption of the Face Recognition-based attendance system could help eliminate these shortcomings. This innovative approach involves utilizing a camera to capture input images, detecting faces using algorithms such as Haarcascade, Eigen values, support vector machines, or the Fisher face algorithm, verifying the faces against a database of student profiles, and marking attendance in an Excel sheet. The use of OpenCV, an open-source computer vision library, ensures the efficient functioning of the system.

Deep Surveillance System

This study has been undertaken to investigate and implement multiple detection systems into a single surveillance system and check whether the input videos may comprise and capture a variety of realistic anomalies or not. In this paper, we propose to learn various anomalies by exploiting both normal and anomalous videos and implemented it to new model. Real time object detection is a vast, vibrant and sophisticated area of computer vision aimed towards object identification and recognition. Object detection detects the semantic objects of a class objects using Open source Computer Vision, which is a library of programming functions mainly trained towards real time computer vision in digital images and videos. The main aim behind this real time object detection is to help the peoples to overcome their difficulty. Real time object detection finds its uses in the areas like tracking objects, video surveillance, pedestrian detection, people counting, self-driving cars, face detection, tracking in sports and many more. This is achieved using Convolution, Probabilistic Neural Networks, etc. which are a representative tool of Deep learning. This project acts as an aiding tool for peoples who wants to take care of everything inside, outside, and around their house just for their full security expectations. Surveillance is a must for small houses to large-scale industries as they fulfil our safety aspects because theft and burglary have always been a problem. By combining this Surveillance idea to IoT and some Machine Learning stuff this will be a major product. The proposed project is a single autonomous surveillance system, based on analysis and detection technology. The proposed system is capable of monitoring all actions at once and alerts the concerned officials immediately and precisely.

Kinematics of the Upper Eyelid and the Globe During Downward Excursion With Comparative Analysis in Patients With Thyroid Eye Disease.

To analyze the kinematics of the upper eyelid and the globe on downward excursion for potential use in monitoring thyroid eye disease (TED) progression in an objective manner. Ten normal volunteers and 10 patients with TED were studied. A high-speed (240 fps) digital camera with a coaxial light source set at a constant distance from the subjects' eyes was used to record the excursion of the upper eyelid and the globe from extreme upgaze to extreme downgaze. Clinical data, including age, gender, race, thyroid function tests, Vision, Inflammation/Congestion, Strabismus/motility restriction, Appearance/exposure score (primary surgeons' preference of TED grading system), exophthalmometry, and eyelid measurements were collected for all patients with TED. Frame-by-frame analyses of the videos were performed using Python software (version 3.6) and the Open Source Computer Vision Library. Temporal resolution was obtained by measuring the number of frames from initiation of eyelid and globe movement from extreme upgaze (t 0 ) to extreme downgaze (t f ). Spatial resolution was obtained by measuring the number of pixels the eyelid margin and the globe traversed from t 0 to t f . The data were then plotted on a graph to calculate the velocity of the upper eyelid and the globe during downward excursion. Velocimetric calculations using high-speed photography suggests that downward excursion of the upper eyelid, and the globe occurs in 2 phases: the acceleration phase and the deceleration phase. Comparative analysis of slow-motion videography demonstrates that patients with TED were found to have attenuation in the early acceleration phase of upper eyelid excursion compared with normal subjects. In patients with TED, the difference in velocity between the eyelid and the globe occurs in the early deceleration phase. The upper eyelid normally synchronizes intimately with the globe during downward eye movement. Data from this study reveal that attenuation mostly in the early deceleration phase of eyelid movement relative to the globe accounts for the dynamic eyelid lag seen on clinical examination. Further analysis is needed to show if a quantified von Graefe sign can be used as an objective means of monitoring progression in TED.

PC Automation using Hand Gestures using Machine Learning

Pattern recognition and Gesture recognition are the developing fields of research.Hand gestures can add another dimension to contactless computer operating.Hence in this project, we develop a software which presents a system prototype that is able to automatically recognize gestures and runs pc commands for particular gestures.We are using python to develop our solution by using the integrated modules like cv2(OpenCV(cv2)-Open source computer vision is a library of programming functions mainly aimed at real-time computer vision.) and Mediapipe(detects hand and fingers data points.)

Improving chest X-ray report generation by leveraging warm starting

Automatically generating a report from a patient’s Chest X-rays (CXRs) is a promising solution to reducing clinical workload and improving patient care. However, current CXR report generators—which are predominantly encoder-to-decoder models—lack the diagnostic accuracy to be deployed in a clinical setting. To improve CXR report generation, we investigate warm starting the encoder and decoder with recent open-source computer vision and natural language processing checkpoints, such as the Vision Transformer (ViT) and PubMedBERT. To this end, each checkpoint is evaluated on the MIMIC-CXR and IU X-ray datasets. Our experimental investigation demonstrates that the Convolutional vision Transformer (CvT) ImageNet-21K and the Distilled Generative Pre-trained Transformer 2 (DistilGPT2) checkpoints are best for warm starting the encoder and decoder, respectively. Compared to the state-of-the-art (M2 Transformer Progressive), CvT2DistilGPT2 attained an improvement of 8.3% for CE F-1, 1.8% for BLEU-4, 1.6% for ROUGE-L, and 1.0% for METEOR. The reports generated by CvT2DistilGPT2 have a higher similarity to radiologist reports than previous approaches. This indicates that leveraging warm starting improves CXR report generation. Code and checkpoints for CvT2DistilGPT2 are available at https://github.com/aehrc/cvt2distilgpt2.

Find Missing Person using HAAR CASCADE

Abstract: The search and rescue operations for finding missing persons present significant challenges, including vast areas to cover, limited resources, and the need for timely and accurate identification. This research paper introduces a comprehensive solution that integrates Haar cascade, face recognition, and OpenCV (Open Source Computer Vision Library) to enhance the search and rescue process. By leveraging computer vision and artificial intelligence techniques, this approach aims to improve the efficiency and accuracy of locating missing individuals.

A video processing and machine learning based method for evaluating safety-critical operator engagement in a motorway control room

In safety-critical automatic systems, safety can be compromised if operators lack engagement. Effective detection of undesirable engagement states can inform the design of interventions for enhancing engagement. However, the existing engagement measurement methods suffer from several limitations which damage their effectiveness in the work environment. A novel engagement evaluation methodology, which adopts Artificial Intelligence (AI) technologies, has been proposed. It was developed using motorway control room operators as subjects. Openpose and Open Source Computer Vision Library (OpenCV) were used to estimate the body postures of operators, then a Support Vector Machine (SVM) was utilised to build the engagement evaluation model based on discrete states of operator engagement. The average accuracy of the evaluation results reached 0.89 and the weighted average precision, recall, and F1-score were all above 0.84. This study emphasises the importance of specific data labelling when measuring typical engagement states, forming the basis for potential control room improvements. Practitioner summary: This study demonstrates an automatic, real-time, objective, and relatively unobtrusive method for measuring dynamic operator engagement states. Computer vision technologies were used to estimate body posture, then machine learning (ML) was utilised to build the engagement evaluation model. The overall evaluation shows the effectiveness of this framework. Abbreviations: AI: Artificial Intelligence; OpenCV: Open Source Computer Vision Library; SVM: Support Vector Machine; UWES: Utrecht Work Engagement Scale; ISA Engagement Scale: Intellectual, Social, Affective Engagement Scale; DSSQ: Dundee Stress State Questionnaire; SSSQ: Short Stress State Questionnaire; EEG: electroencephalography; ECG: Electrocardiography; VMOE: Video-based Measurement for Operator Engagement; CMU: Carnegie Mellon University; CNN: Convolutional Neural Network; 2D: two dimensional; ML: Machine learning.

Automated invoice data extraction using image processing

Manually processing invoices which are in the form of scanned photocopies is a time-consuming process. There is a need to automate the task of extraction of data from the invoices with a similar format. In this paper we investigate and analyse various techniques of image processing and text extraction to improve the results of the optical character recognition (OCR) engine, which is applied to extract the text from the invoice. This paper also proposes the design and implementation of a web enabled invoice processing system (IPS). The IPS consists of an annotation tool and an extraction tool. The annotation tool is used to mark the fields of interest in the invoice which are to be extracted. The extraction tool makes use of opensource computer vision library (OpenCV) algorithms to detect text. The proposed system was tested on more than 25 types of invoices with the average accuracy score lying between 85% and 95%. Finally, to provide ease of use, a web application is developed which also presents the results in a structured format. The entire system is designed so as to provide flexibility and automate the process of extracting details of interest from the invoices.

Method and analysis of color changes of facial skin after applying skin makeup

This study investigated facial skin color differences before and after makeup. Toward this goal, a photo gauge, devised with a pair of color checkers as a reference, collected face images. In addition, color calibration and a deep-learning method extracted the color values of representative areas of facial skin. The photo gauge photographed 516 Chinese females before and after applying makeup. Then, the collected images were calibrated by referencing skin color patches, and the lower cheek regions’ pixel colors were extracted using open-source computer vision libraries. Following the visible color spectrum of humans, the color values were computed in L*, a*, and b* of CIE1976L*a*b*. The results showed that the facial colors of the Chinese females changed to become brighter, less reddish, and less yellowish after applying the makeup, resulting in a paler skin tone. During the experiment, subjects were given five varieties of liquid foundation to choose one sample that best fits their skin. However, we failed to find any noticeable relationship between the individual’s facial skin color characteristics and the liquid foundation selected. In addition, 55 subjects were identified according to their makeup use frequency and skill, but their color changes did not differ from the other subjects. This study provided quantitative evidence of makeup trends in the Shanghai region in China, and the method proposes a novel approach toward remote skin color research.

YONGʻINNI VIDEOTASVIRDA RANGLI FILTRLASH BILAN INTENSIVLIK OʻZGARISHI ASOSIDA ANIQLASH

The paper reviews achievements made in the field of video analytics and computer vision, which enable automatic detection of fires, based on video data. Various algorithms have been implemented in sought for effective fire detection methods using video. As one of these, a color-based fire detection algorithm is being described. However, using only one color model for fi re detection is an inefficient approach. The paper proposes a method for estimating temporal changes in pixel intensity, which is used to retrieve information about fires from fire-like objects in a video image. This method helps to calculate the average intensity value in a sequence of shots. A special software has been developed in the Python programming language using the OpenCV (Open Source Computer Vision Library) library, and the corresponding findings have been gained in view to demonstrate the effectiveness of the proposed method.

Facial Emotion Recognition Using Deep Neural Network

The major part in the process of humanization of systems is the capability of distinguishing the emotions of the person. In this research paper we represent the composition of an instinctive system that is capable of detecting the emotion by using their facial expressions. Three techniques of neural network are tailored, educated and subordinated to different jobs, after this the performance of the network was improved. A live videotape operation that can currently simulate the person’s emotion depicts how well the model connects to the world. Since the invention of computers many technologists and masterminds are introducing instinctively intelligent systems that are very helpful to humans mentally and physically. In the previous decades the usage of computer has increased rapidly which helps in developing fast literacy systems, where internet has provided vast quantum of data for teaching the machine. These two enlargements elevated the exploration on intelligent learning systems by using neural networks in favorable ways. The facial emotion detection machine needs to be trained to get the system ready. The installation of OpenCV(Open Source Computer Vision) is essential for this machine. OpenCV is a library that is required for computer vision.

Online attendance system based on facial recognition with face mask detection.

This paper presents an online system for recording attendance based on facial recognition incorporating facial mask detection. The main objective of this project is to develop an effective attendance system based on face recognition and face mask detection, and to provide this service online through a browser interface. This would allow any user to use this system without the need to install special software. They simply need to open the interface of this system in a browser through any terminal. Recording attendance information online allows data to be easily recorded in a centralized online database. Since faces are used as biometric signatures in this project, all users registered in the system will have their profiles loaded with their face-images samples. Initially, before face recognition can be done, the model training phase based on SVM will be carried out, mainly to develop a trained model that can perform face recognition. A set of synthetic data will also be used to train the same model so that it can perform identification for users wearing face masks. The server application is coded in Python and uses the Open-Source Computer Vision (OpenCV) library for image processing. For web interfaces and the database, PHP and MySQL are used. With the integration of Python and PHP scripting programs, the developed system will be able to perform processing on online servers, while being accessible to users through a browser from any terminal. According to the results and analysis, an accuracy of about 81.8% can be achieved based on a pre-trained model for face recognition and 80% for face mask detection.

Seeing wildlife behavior in a new way: Novel utilization of computer vision for focal reptile videography behavior study

AbstractTechnology has the potential to assist and directly address ecological topics. Our study compared efficiency and efficacy of foreground‐detection computer vision and manual review methods for understanding behavior, natural history, and ecology of green pit vipers. Between 2015 and 2020 at the Sakaerat Biosphere Reserve in Thailand, 18,871 scans (pictures and timelapse recordings) and 1,507 minutes of continuous videos were recorded for 6 adult female, big‐eyed pit vipers (Trimeresurus macrops). MotionMeerkat, an open source computer vision tool for finding ecological events in long video, appeared to be less efficient and overestimated presence of behaviors. Our study suggested that foreground‐detection computer vision methods, exemplified by the program MotionMeerkat, can be free and open‐source as well as easy to learn and use, but may not be as time efficient or fully address the complexity of ecological topics compared to review of images by human observers.

Visual Analysis of Agricultural Workers using Explainable Artificial Intelligence (XAI) on Class Activation Map (CAM) with Characteristic Point Data Output from OpenCV-based Analysis

In this study, we use explainable artificial intelligence (XAI) based on class activation map (CAM) techniques. Specifically, we use Grad-CAM, Grad-CAM++, and ScoreCAM to analyze outdoor physical agricultural (agri-) worker image datasets. In previous studies, we developed body-sensing systems to analyze human dynamics with the aim of enhancing agri-techniques, training methodologies, and worker development. These include distant, visual data-based sensing systems that capture image and movie datasets related to agri-worker motion and posture. For this study, we first obtained the aforementioned image datasets for researcher review. Then, we developed and executed Python programs with Open-Source Computer Vision (OpenCV) libraries and PyTorch to run XAI-oriented systems based on CAM techniques and obtained heat map-pictures of the visual explanations. Besides, we implement optical flow-based image analyses using our Visual C++ programs with OpenCV libraries, automatically set and chase the characteristic points related to the video datasets. Next, we analyze the dataset features and compare experienced and inexperienced subject groups. We investigate the output’s features, accuracies, and robustness to be able to make recommendations for real agri-workers, managers, product-developers, and researchers. Our findings indicate that the visualized output datasets are especially useful and may support further development of applied methods for these groups.

A Novel Segmentation Recognition Algorithm of Agaricus bisporus Based on Morphology and Iterative Marker-Controlled Watershed Transform

Accurate recognition of Agaricus bisporus is a prerequisite for precise automatic harvesting in a factory environment. Aimed at segmenting mushrooms adhering together from the complex background, this paper proposes a watershed-based segmentation recognition algorithm for A. bisporus. First, the foreground of A. bisporus is extracted via Otsu threshold segmentation and morphological operations. Then, a preliminary segmentation algorithm and a novel iterative marker generation method are proposed to prepare watershed markers. On this basis, a marker-controlled watershed algorithm is adopted to segment and recognize A. bisporus individuals. All the algorithms are implemented based on OpenCV (Open Source Computer Vision) libraries. Tests on images of A. bisporus collected at the cultivation bed show that the average correct recognition rate of the proposed algorithm is 95.7%, the average diameter measurement error is 1.15%, and the average coordinate deviation rate is 1.43%. The average processing time is 705.7 ms per single image, satisfying the real-time constraints based on 1 image/s. The proposed algorithm performed better than the current Circle Hough Transform (OpenCV’s implementation). It is convenient and easy to operate, providing a sound basis for subsequent research on mechanized harvesting equipment for A. bisporus.

Crack detection and crack segmentation in concrete beams undergoing mode I fracture using computer vision and convolutional neural network

The study presents two crack detection techniques for a beam undergoing mode I fracture: (i) convolutional neural network approach and (ii) image processing using opensource computer vision library OpenCV and pixel count of cracks approach. The second method also includes crack segmentation and masking for visualization of the cracks. The study attempted to evaluate the accuracy of both methods at different crack mouth opening displacement of seven simply supported concrete beams being tested using three-point bend test. The novelty of the present work lies in quantification of the crack growth from captured images and evaluating the potential of applying computer vision techniques as a replacement for sensitive crack measuring instruments to create a robust computer vision based health monitoring system. Results suggest that both methods identified the cracks in the beam and are capable of generating a warning before the collapse.

Studi Pendahuluan Sistem Keamanan Parkir Berbasis Opencv Di Kampus Universitas Indo Global Mandiri

Indo Global Indo Global Mandiri University, is one of the universities in Palembang City which has a fairly large parking area, so it is necessary to have a concept of comfort and security for the academic community and also guests in placing their vehicles in the parking area within the IGM University Campus. This study tries to provide an alternative how to build a vehicle parking concept, especially cars, by detecting vehicle plates and who owns it, so that it can be known early on vehicles that enter an unknown parking area, so that it can be known whether guests are or are treated as people who are not clear. the goal. The research model was built using the Python Programming Language with the OpenCV (Open Source Computer Vision) library as an image processor which with a process or several stages can change the car license plate in the form of an image into a string (word), then by connecting it to the vehicle owner's database. In order to detect who owns the car that will park. The results of the preliminary research, by detecting the vehicle plate in the form of a car image, the results obtained that the area code and plate number can be detected, but other images are also detected. The results show that the plate number is detected as BG 1359 UJ, converted to BG' 13:19 Q UJ. There were 2 conversion failures, namely the number 5 was changed to :1 and the small white circle (bolt) was changed to Q.Keywords: Parking, Detection, Vehicle Plate and OpenCV

A Novel Approach for Exam E-assessment Utilizing Image Processing

There is a brand-new feature called Exam (Infinity Exam) that supports paper-based exams and speeds up the entire process while maintaining all of their beneficial qualities and minimizing their drawbacks, notably in higher education. The method is very different from those employed in the earlier 10+ years, which were implemented in a way that prevented them from replicating and supplanting the conventional paper-based examination format. The article's core relies on the image processing flow, which is the most crucial component of the software. Multiple Choice Questions (MCQ) have been a more common method of testing someone's knowledge over time. The use of multiple choice questions in exams is becoming more widespread in the education sector (including in schools and colleges). It is employed even when conducting interviews. The current scenario involves either manually correcting the test or using OMR technology. Having OMR at all times in real time is rather challenging, and manually correcting it takes a lot of effort and could result in a mistake. We address this issue by applying a digital image processing technique in our proposed system to correct the response using multiple-choice questions written in Python. Here, we are processing data using Open-Source Computer Vision Library (OpenCV).

Identification and Counting of Coffee Trees Based on Convolutional Neural Network Applied to RGB Images Obtained by RPA

Computer vision algorithms for counting plants are an indispensable alternative in managing coffee growing. This research aimed to develop an algorithm for automatic counting of coffee plants and to determine the best age to carry out monitoring of plants using remotely piloted aircraft (RPA) images. This algorithm was based on a convolutional neural network (CNN) system and Open Source Computer Vision Library (OpenCV). The analyses were carried out in coffee-growing areas at the development stages three, six, and twelve months after planting. After obtaining images, the dataset was organized and inserted into a You Only Look Once (YOLOv3) neural network. The training stage was undertaken using 7458 plants aged three, six, and twelve months, reaching stability in the iterations between 3000 and 4000 it. Plant detection within twelve months was not possible due to crown unification. A counting accuracy of 86.5% was achieved with plants at three months of development. The plants’ characteristics at this age may have influenced the reduction in accuracy, and the low uniformity of the canopy may have made it challenging for the neural network to define a pattern. In plantations with six months of development, 96.8% accuracy was obtained for counting plants automatically. This analysis enables the development of an algorithm for automated counting of coffee plants using RGB images obtained by remotely piloted aircraft and machine learning applications.