<div> <div> <div> <p>Estimation of physiological vital signs such as the Heart Rate (HR) has attracted a lot of attention due to the increase interest in health monitoring. The most common HR estimation methods such as Photoplethysmography(PPG) require the physical contact with the subject and limit the movement of the subject. Video-based HR estimation, known as videoplethysmography (VHR), uses image/video processing techniques to estimate remotely the human HR. Even though various VHR methods have been proposed over the past 5 years, there are still challenging problems such as diverse skin tone and motion artifacts. In this thesis we present a VHR method using temporal difference filtering and small variation amplification based on the assumption that HR is the small color variations of skin, i.e. micro blushing. This method is evaluated and compared with the two previous VHR methods. Additionally, we propose the use of spatial pruning for an alternative of skin detection and homomorphic filtering for the motion artifact compensation. </p><p><br></p> <p>Intelligent video surveillance system is a crucial tool for public safety. One of the goals is to extract meaningful information efficiently from the large volume of surveillance videos. Person re-identification (ReID) is a fundamental task associated with intelligent video surveillance system. For example, ReID can be used to identity the person of interest to help law enforcement when they re-appear in the different cameras at different time. ReID can be formally defined as establishing the correspondence between images of a person taken from different cameras. Even though ReID has been intensively studied over the past years, it is still an active research area due to various challenges such as illumination variations, occlusions, view point changes and the lack of data. In this thesis we propose a weighted two stream train- ing objective function which combines the Siamese cost of the spatial and temporal streams with the objective of predicting a person’s identity. Additionally, we present a camera-aware image-to-image translation method using similarity preserving Star- GAN (SP-StarGAN) as the data augmentation for ReID. We evaluate our proposed methods on the publicly available datasets and demonstrate the efficacy of our methods.</p></div></div></div>

Contact-less human detection and monitoring using radar technology has been recently applied in many areas including search-and-rescue for earthquake victims, fall detection, gait analysis and detection of other human activities. Radars can also provide important information about a persons state of health by monitoring the level of activities, heart and breathing rate. Also it can be used to generate warnings if some of the monitored parameters are outside of predefined limits. The major application of this work is for monitoring in-mates and their activities. This thesis deals with characterizing the performance of the radar system used for monitoring a single person in a contained environment. This thesis is experimentally based and during the thesis a large number of experiments were performed in order to monitor subjects in realistic conditions. The thesis explores feasibility of using the radar with a single radio-frequency channel input and two algorithms for breathing and heart rate estimation when the subject is at different relative orientation towards the radar as well as in different postures. Algorithm one is using Fast Fourier Transformation (FFT) and algorithm two is using Empirical Mode Decomposition (EMD) with Minkowski distance. We also detect the zones where the subject is when the subject is moving. Since this exploratory analysis provides initial features for classifications and algorithms for breathing and heart beat estimation, it can represent a foundation for future works on designing systems that track subjects and their breathing in real-time.

No description available.

Heart rate can say a lot about a person's health. While most conventional methods for heart rate measurement require contact with the subject, these are not always applicable. In this thesis, a non-invasive method for pulse detection is implemented and analyzed. Different signals from the color of the forehead—including the green channel, the hue channel and different ICA and PCA components—are inspected, and their resulted heart rates are weighted together according to the significance of their FFT peaks. The system is tested on videos with different difficulties regarding the amount of movement and setting of the scene. The results show that the approach of weighting measurements from different signals together has great potential. The system in this thesis, however, does not perform very well on videos with a lot of movement because of motion noise. Though, with better, less noisy signals, good results can be expected. / En människas puls säger en hel del om dennes hälsa. För att mäta pulsenanvänds vanligtvis metoder som vidrör människan, vilket iblandär en nackdel. I det här examensarbetet tas en metod för pulsmätningpå avstånd fram, som endast använder klipp från en vanlig videokamera. Färgen i pannan mäts och utifrån den genereras flera signalersom analyseras, vilket resulterar i olika mätvärden för pulsen. Genomatt värdera dessa mätvärden med avseende på hur tydliga signalernaär, beräknas ett viktat medelvärde som ett slutgiltigt estimat på medelpulsen. Metoden testas på videoklipp med varierande svårighetsgrad,beroende på hur mycket rörelser som förekommer och på vilketavstånd från kameran försökspersonen står. Resultaten visar att metodenhar mycket god potential och att man kan man förvänta sig finaresultat med bättre, mindre brusiga signaler.

In the context of health improving, the measurement of vital parameters such as heart rate (HR) can provide solutions for health monitoring, prevention and screening for certain chronic diseases. Among the different technologies for HR measuring, photoplethysmography (PPG) technique embedded in smart watches is the most commonly used in the field of consumer electronics since it is comfortable and does not require any user intervention. To be able to provide an all day and night long HR monitoring method, difficulties associated with PPG signals vulnerability to Motion Artifact (MA) must be overcome. Conventional signal processing solutions (power spectral density analysis) have limited generalization capability as they are specific to certain types of movements, highlighting the interest of machine learning tools, particularly deep learning (DL). Since DL models in the literature are trained on data from a different sensor than the internal sensor, transfer learning may prove unsuccessful. This work proposes a DL approach for estimating HR from wrist PPG signals. The model is trained on internal data with a greater demographic diversity of participants. This project also illustrates the contribution of multi-path and multi-wavelength PPG instead of the conventional single green PPG solution. This work presents several models, called DeepTime, with selected input channels and wavelengths: Mono_Green, Multi_Green, Multi_Wavelength, and Multi_Channel_Multi_Wavelength. They take temporal PPG signals as inputs along with 3D acceleration and provide HR estimation every 2 seconds with an 8-second initialization. This convolutional neural network comprised of several input branches improves the existing Withings internal method’s overall Mean Absolute Error (MAE) from 9.9 BPM to 6.9 BPM on the holdout test set. This work could be completed and improved by adding signal temporal history using recurrent layers, such as Long-Short-Term-Memory (LSTM), training the model with a bigger dataset, improving preprocessing steps or using a more elaborate loss function that includes a trust score. / I sammanhanget av förbättring av hälsouppföljning kan mätning av vitala parametrar som hjärtfrekvens (HR) erbjuda lösningar för förebyggande och screening av vissa kroniska sjukdomar. Bland olika tekniker för mätning av HR är fotoplethysmografi (PPG) integrerad i smartklockor den vanligast använda inom elektronikområdet eftersom den är bekväm och inte kräver något användaringripande. För att erbjuda en kontinuerlig HRövervakningsmetod utgör sårbarheten hos PPG-signaler för rörelseartefakter (MA) en stor utmaning. Konventionella signalbehandlingslösningar (analys av effektspektraltäthet) har begränsad generaliseringsförmåga eftersom de är specifika för vissa typer av rörelser, vilket betonar intresset för maskininlärningsverktyg, särskilt djupinlärning (DL). Eftersom DL-modeller i litteraturen tränas på data från en annan sensor än den interna sensorn kan överföringsinlärning vara misslyckad. Detta arbete föreslår en DL-ansats för att uppskatta HR från PPG-signaler på handleden. Modellen tränas på interna data med en större demografisk mångfald bland deltagarna. Detta projekt illustrerar även bidraget från flervägs- och flervågs-PPG istället för den konventionella enkla gröna PPG-lösningen. Detta arbete presenterar flera modeller, kallade DeepTime, med utvalda ingångskanaler och våglängder: Mono_Green, Multi_Green, Multi_Wavelength och Multi_Channel_Multi_Wavelength. De tar in temporära PPG-signaler tillsammans med 3D-acceleration och ger HR-uppskattning var 2:a sekund med en initialisering på 8 sekunder. Detta konvolutionella neurala nätverk, som består av flera ingångsgrenar, förbättrar den totala medelabsoluta felet (MAE) från 9,9 BPM (befintlig intern metod) till 6,9 BPM på testuppsättningen. Detta arbete kan kompletteras och förbättras genom att integrera den temporala historiken hos signalen med hjälp av återkommande lager (som LSTM), träna modellen på mer data, förbättra förbehandlingsstegen eller välja en mer sofistikerad förlustfunktion som inkluderar ett konfidensvärde.