Letteratura scientifica selezionata sul tema "Camera monitoring"

This thesis was made in collaboration with the Baltic Seabird Project with the purpose of evaluating the possibilities of using cameras to monitor the fledging success of Common Guillemots on Stora Karlsö. Fledging success is measured from the time the egg is laid, hatched, until the chick is 15 days old and considered ready to leave the nest. Camera monitoring means that the breeding area where the birds reside is photographed at a predetermined interval to attempt to capture the offspring to determine its survival. The study was conducted on Stora Karlsö by installing cameras in the artificial breeding shelf, Auk Lab, and determine how well the method works. Whether the offspring can be caught on camera was the main question, with a secondary question addressing the differences from the traditional monitoring method which is done by daily controls by one person on site. The results show that camera monitoring can be a useful method. It is possible to observe the off-spring, but distance and angle are important factors affecting image quality. Power supply and the memory card size are important factors that decide whether time savings can be made compared to the traditional method. Further development of the routines for camera monitoring is necessary to create a reliable data collection<br>Studien gjordes i samarbete med Baltic Seabird Project med syftet att undersöka möjligheten att använda kameror för att övervaka häckningsframgång hos Sillgrissla på Stora Karlsö. Häckningsframgång definieras som tiden från att ägget läggs, kläcks, och ungen överlever i 15 dagar då den anses vara redo att lämna boet. Övervakning med kameror innebär att häckningsytan där fåglarna sitter fotograferas med ett bestämt tidsintervall för att försöka fånga avkomman på bild och därmed se att den har överlevt. Undersökningen har genomförts på Stora Karlsö genom att installera kameror i den konstgjorda häckningshyllan Auk Lab och bedöma hur väl metoden fungerar. Huruvida avkomman fångas på bild eller inte var den huvudsakliga frågeställningen, med en sekundär frågeställning som rör skillnaden mellan den traditionella insamlingsmetoden som genomförs genom att en person dagligen kontrollerar fåglarna. Resultaten visar att kameraövervakning kan vara en användbar metod, då det går att fånga avkomman på bild men avstånd och vinkel är några av de avgörande faktorerna för att framgångsrikt kunna observera avkomman. Strömtillförsel och minneskortets storlek är avgörande faktorer för att göra vinningar tidsmässigt jämfört med den traditionella metoden. Fortsatt utveckling av rutinerna för kameraövervakning är nödvändigt för att få en pålitlig datainsamling.  2

Measuring physiological parameters or vital sign using camera has become popular in recent years. Contact-less monitoring and extraction of vital signs can be important source of information in situations like medical care system and safety control system. This paper presents the implementation of real-time, non-contact method for extraction of vital signs, heart rate in this case. A better face tracking method is used for efficient face detection. This study extends some of the previous works done and have a comparison study with several methods. The developed system used filtering with window over the green channel of the signal and then Converted to frequency domain to analyze the signal to detect heart rate. The developed system achieved high correlation and showed small error while referencing with actual heart signal from ECG. This method delivers better result in better light condition but gives fairly good result on lower light as well.

Patients in hospital can be prone to sudden, life-threatening changes in their cardiovascular state. Haemodynamic parameters such as blood pressure, pulse transit time (PTT) and perfusion can be monitored in clinical situations to identify these changes as early as possible. Continuous blood pressure is usually monitored using a catheter placed into a major artery, but this is invasive and involves risk to the patient. In the last decade, the field of non-contact vital sign monitoring has emerged, with growing evidence that the remote photoplethysmogram (rPPG) signal can be used to estimate vital signs using video cameras. If the analysis of the rPPG signal can be expanded to include the estimation of haemodynamic parameters, it could result in methods for the continuous, non-contact monitoring of a subject's haemodynamic state. In a physiology study, a series of video recordings were made of 43 healthy volunteers. The subjects sat in a purpose-built chamber, and the composition of the air was carefully adjusted to cause the subjects to experience large, controlled changes in blood oxygen levels. To validate the video camera algorithms, reference data were also collected. Along with the volunteer study, a clinical study was performed to acquire data in a challenging clinical environment. Data were collected from patients on haemodialysis in the Renal Unit, a population likely to experience sudden changes in haemodynamics. The reference data from the Renal Unit study were analysed to determine the extent to which PTT and mean arterial pressure (MAP) are related. The correlation coefficients and linear fits were found on a global and a per-subject basis. In addition, the video recordings from the Physiology study were processed to derive rPPG signals, and these signals were analysed to obtain estimates for PTT. Local rPPG signals were also derived for different regions of interest, and the waveforms were analysed using a novel application of the technique of signal averaging to produce spatial maps of perfusion and blood flow. The correlation between conventionally measured PTT and MAP was found to be weaker in the haemodialysis population than has been shown elsewhere in the literature, except for a sub-set of patients. The results of the video analysis showed that PTT could be estimated robustly and consistently, although direct validation of these estimates was not possible because of the different method used to calculate the reference PTT. For most subjects, the spatial mapping methods produced robust maps that were consistent over time. These results suggest that it is possible to detect changes in haemodynamics using a video camera, and that this could have applications in healthcare, providing that challenges such as subject movement and clinical validation can be overcome.

Since forward collision is one of the most common and dangerous types of traffic accidents, many studies and researches have been conducted to develop forward collision avoidance system. To facilitate the tradeoff between comfort and safety for forward collision avoidance, the driver's state needs to be monitored and estimated. Such support is necessary for Forward Collision Warning (FCW) system given human-involved control. Due to the advances of Driver Monitoring System (DMS), the demand for camera-based driver's state estimation has increased. This master thesis project, conducted at Zenuity AB, investigates a method to estimate driver's awareness based on DMS. The estimation of a driver's awareness is expected to help adapt FCW system based on visual attention when facing the unpredictable braking of the leading vehicle. The project consists of three tasks: gaze estimation, Gaze-to-Object Mapping (GTOM), and awareness estimation. A combined Kalman Filter was developed in gaze estimation for compensation of missing data and outliers and reducing the difference to “ground truth” data. The uncertainty matrix from gaze estimation was utilized to extract a gaze-to-object probability signal in GTOM, while the corresponding fixation duration was also obtained in GTOM. The two extracted new features were used in awareness estimation with two methods: Logistic Regression and two-Hidden Markov Model. The comparison between the two methods reveals whether a complex method is preferred or not. Based on the results of this project, Logistic Regression seems to perform better in driver's state estimation, with 92.0% accuracy and 76.3% True Negative rate. However, further research and improvements on the two-Hidden Markov Model are needed to reach a more comprehensive conclusion. The main contribution of this project is an investigation of an end-to-end method for driver's awareness estimation and thereby an identification of challenges for further studies.<br>Frontkollision (forward collision) är en av de vanligaste och farligaste typerna av trafikolyckor. Många studier och undersökningar har genomförts för att utveckla system för att undvika kollisioner. För att underlätta avvägningar mellan komfort och säkerhet för att undvika Frontkollision måste förarens tillstånd övervakas och skattas. Ett sådant stöd är nödvändigt för Forward Collision Warning (FCW) systemet, som involverar interaktion med människor. Efterfrågan på kamerabaserad uppskattning för föraren har ökat på grund av framsteg Driver Monitoring System (DMS). Det här examensarbete genomfördes på Zenuity AB och undersökte en metod för att skatta förarens medvetenhet baserad på Driver Monitoring System. Uppskattningen av förarens medvetenhet förväntas bidra till att anpassa FCW-systemet. Detta FCW-system är baserat på visuell uppmärksamhet om när oförutsägbar bromsning av det framförvarande fordonet sker. Examensarbetet består av tre uppgifter: blickuppskattning, Gaze-to-Object Mapping (GTOM), och medventenhetsuppskattning. Ett kombinerat Kalman-filter har utvecklats i gaze uppskattning för att kompensera saknade data och outliers samt reducera skillnaden till “ground truth” data. Osäkerhetesmatrisen från gaze uppskattningen användes för att extrahera en gaze-to-object sannolikhetssignal i GTOM. Den motsvarande fixeringsvaraktigheten erhålls också i GTOM. De två extraherade nya egenskaperna användes i medvetenhetsanalys med hjälp av två metoder: logistic regression och two-Hidden Markov Model. Jämförelsen mellan de två metoderna avslöjar om en komplex metod är att föredra eller inte. Resultatet av detta examensarbet visar att logistic regression fungerar bättre i förarens statusuppskattning med 92% noggrannhet och 76.3% True Negative rate. Vidare forskning och förbättringar av den two-hidden Markov modell behövs för att dra en mer omfattande slutsats. Det huvudsakliga bidraget av examensarbetet är en utforskning av en end-to-end metod för att uppskatta förarens medvetenhet och därmed kunna identifiera utmaningar för framtid studie.

This thesis presents a novel self-powered infrastructural traffic monitoring approach that estimates traffic information by combining three detection techniques. The traffic information can be obtained from the presented approach includes vehicle counts, speed estimation and vehicle classification based on size. Two categories of sensors are used including IR Lidar and IR camera. With the two sensors, three detection techniques are used: Time of Flight (ToF) based, vision based and Laser spot flow based. Each technique outputs observations about vehicle location at different time step. By fusing the three observations in the framework of Kalman filter, vehicle location is estimated, based on which other concerned traffic information including vehicle counts, speed and class is obtained. In this process, high reliability is achieved by combing the strength of each techniques. To achieve self-powering, a dynamic power management strategy is developed to reduce system total energy cost and optimize power supply in traffic monitoring based on traffic pattern recognition. The power manager attempts to adjust the power supply by reconfiguring system setup according to its estimation about current traffic condition. A system prototype has been built and multiple field experiments and simulations were conducted to demonstrate traffic monitoring accuracy and power reduction efficacy.<br>Master of Science

Technical surface cleanliness is a mandatory requirement for many production lines. It is part ofthe quality control process to ensure that the surfaces of the manufactured components are freeof contaminations and are ready for the next step in the production line. The project has takeninto account two different requirements related to the component cleanliness: first, thecleanliness is measured as the level of contamination on a sample and then, the particles sizes aredetermined. An important factor in the cleanliness is the presence of particles larger than acertain limit, called the critical particles, which have to be detected. This thesis is inspired from anissue SCANIA has (2018) in its production plant in Södertälje, Stockholm. SCANIA has a qualitycontrol system that analyzes the contamination level of few components through a timeconsuming process that takes hours and requires expensive microscopes, human intervention anda dedicated measurement room in the plant. The aim of this thesis is to investigate an alternativetechnical cleanliness monitoring method through the image analysis of the contaminationsamples. Through a simple digital camera, pictures from the contaminated samples are taken,processed and analyzed in order to obtain the cleanliness level and the particles size of thesamples. In contract to the current solution, the proposed method has the possibility of beingimplemented in the production line, providing larger sampling rate.<br>Teknisk ytrenhet är ett obligatoriskt krav för många produktionslinor. Det är en del avkvalitetsstyrningsprocessen för att säkerställa att ytorna på de tillverkade komponenterna är friafrån föroreningar och är redo för nästa steg i produktionslinan. Projektet har tagit hänsyn till tvåolika krav relaterade till komponents renhet: för det första mäts renheten som föroreningsnivå påett prov och sedan bestäms partikelstorlekarna. En viktig faktor vid renhetsbestämning ärnärvaron av partiklar som är större än en viss gräns, som kallas de kritiska partiklarna, som måstedetekteras. Avhandlingen är inspirerad av ett problem SCANIA har (2018) i sinproduktionsanläggning i Södertälje, Stockholm. SCANIA har ett kvalitetskontrollsystem somanalyserar föroreningsnivån av få komponenter genom en tidskrävande process vilken tar timmaroch kräver dyra mikroskop, mänskligt ingripande och ett dedikerat mätrum i anläggningen. Syftetmed denna avhandling är att undersöka en alternativ teknisk renhetsövervakningsmetod genombildanalys av förorenade proverna. Med en enkel digitalkamera tas bilder på de förorenadeproverna, bearbetas och analyseras för att uppnå renhetsnivån och partikelstorleken på proven. Iöverensstämmelse med den nuvarande lösningen finns det möjlighet att implementera denföreslagna metoden i produktionslinan, vilket ger större samplingsfrekvens.

Power transformers are essential components within the power distribution system. Transformer failures having a high economic impact on the distribution operators and the industrial and domestic customers. Dielectric mineral oil is used in transformers for electrical insulation between live parts, cooling and protection of the insulation papers in the transformer. Oil contamination and changes in the chemical structure of the oil result in the decay of insulation paper and reduced insulation and cooling which can lead to a transformer failure. The general approach to oil monitoring has been for an operator to examine the colour index (ASTM) of the oil, electrical strength, acidity, water contents and dissolved gas analysis results and form an opinion as to the extent of oil degradation. Chromatic techniques enable data from di↵erent sources to be combined to give an overall evaluation about the condition of a system being monitored. One of the main goals for this work was to use chromatic techniques for integrating the oil data from the di↵erent sources and sensors. In addition the chromatic approach enables liquids to be monitored optically so a second aim was to apply chromatic optical oil monitoring using portable system by transmitting polychro- matic light through the oil sample, which is contained in a transparent cuvette and imaged using a mobile phone camera. A number of oil samples were optically analysed with portable chromatic sys- tem and the optical data was compared with the colour index and chromatically companied with the dissolved gas and other oil data to give overall evaluation of oil degradation. The chromatic optical result compared favourably with the colour index. It was also possible to classify the oil samples chromatically into categories of low, medium and high degradation. This enabled the chromatic data combination approach to be implemented as a prototype system in Matlab software that an operator could use to get a classification of an oil sample. Essential experiment was introduced to monitor di↵erent oil particles by obtaining the result of di↵erent filtered samples through the filter paper. Beside the ability to analyse data and distinguish between fresh and contam- inated oil samples the chromatic technique has the ability to track the history of di↵erent degraded oil samples which can give an indication about failure faults and it could give a prediction of any future faults. Therefore a commercially viable reliable system can be developed to extend the service life and extend the maintenance schedules. These monitoring systems could lead to extending the service life of the transformers, making the electricity supply more reliable and giving the consumer a better quality of life.

Background Camera-based photoplethysmography (cbPPG) is a measurement technique which enables remote vital sign monitoring by using cameras. To obtain valid plethysmograms, proper regions of interest (ROIs) have to be selected in the video data. Most automated selection methods rely on specific spatial or temporal features limiting a broader application. In this work, we present a new method which overcomes those drawbacks and, therefore, allows cbPPG to be applied in an intraoperative environment. Methods We recorded 41 patients during surgery using an RGB and a near-infrared (NIR) camera. A Bayesian skin classifier was employed to detect suitable regions, and a level set segmentation approach to define and track ROIs based on spatial homogeneity. Results The results show stable and homogeneously illuminated ROIs. We further evaluated their quality with regards to extracted cbPPG signals. The green channel provided the best results where heart rates could be correctly estimated in 95.6% of cases. The NIR channel yielded the highest contribution in compensating false estimations. Conclusions The proposed method proved that cbPPG is applicable in intraoperative environments. It can be easily transferred to other settings regardless of which body site is considered.

Camera traps have taken off one of the most popular tools in ecology. This thesis aims to develop existing camera trap methodology in order to better assess the distribution and abundance of deer in the UK. Particular focus was made on the invasion history of muntjac to help elucidate their invasion pattern. The number of founding females was estimated to be 4 or 5 individuals. The effect of covariates on the camera detection zones were considered to help improve density estimates resulting from camera trap research. Flash type and individual passing speed proved to be two important covariates adding weight to the recommendation that camera detection zones should be survey specific and that activity patterns should be considered when determining detection zones. Eight deer population densities were estimated from across the UK using both thermal imaging distance sampling and random encounter model (REM) techniques. A higher density was found with the REM, thought the two methods appeared more comparable in open woodlands. A low quality thermal imagine camera may have bias the results, but this study also emphasises the need to ensure other parameters, such as daily travel distance are site specific and as accurate as possible. Muntjac sightings, within Northern Ireland, were collated and verified using a scoring system and survey combination. The REM was trialled in one site, finding a minimum population of 5 muntjac deer. This baseline result can be used in any future population monitoring. These verified sightings alongside others from Ireland were used to test a muntjac species distribution model with different sampling bias approaches. The random background model was the most parsimonious model suggesting, in this case, that the additional bias controlling techniques may not always be necessary.