Tomas Bengtsson

Regularized optimization for joint super-resolution and high dynamic range image reconstruction in a perceptually uniform domain

This paper discusses resolution enhancement of a set of images with varying exposure durations, having a high combined dynamic range. So far, little has been said in relation to the Human Visual System when it comes to Super-Resolution and High Dynamic Range fusion, unlike the case for traditional Super-Resolution where errors are measured with respect to human perception in the pixel domain. We propose a Super-Resolution method in the L*a*b* domain to bridge that gap and present some image reconstruction results.

Optical flow estimation on image sequences with differently exposed frames

Abstract. Optical flow (OF) methods are used to estimate dense motion information between consecutive frames in image sequences. In addition to the specific OF estimation method itself, the quality of the input image sequence is of crucial importance to the quality of the resulting flow estimates. For instance, lack of texture in image frames caused by saturation of the camera sensor during exposure can significantly deteriorate the performance. An approach to avoid this negative effect is to use different camera settings when capturing the individual frames. We provide a framework for OF estimation on such sequences that contain differently exposed frames. Information from multiple frames are combined into a total cost functional such that the lack of an active data term for saturated image areas is avoided. Experimental results demonstrate that using alternate camera settings to capture the full dynamic range of an underlying scene can clearly improve the quality of flow estimates. When saturation of image data is significant, the proposed methods show superior performance in terms of lower endpoint errors of the flow vectors compared to a set of baseline methods. Furthermore, we provide some qualitative examples of how and when our method should be used.

On robust optical flow estimation on image sequences with differently exposed frames using primal-dual optimization

Optical flow methods are used to estimate pixelwise motion information based on consecutive frames in image sequences. The image sequences traditionally contain frames that are similarly exposed. However, many real-world scenes contain high dynamic range content that cannot be captured well with a single exposure setting. Such scenes result in certain image regions being over- or underexposed, which can negatively impact the quality of motion estimates in those regions. Motivated by this, we propose to capture high dynamic range scenes using different exposure settings every other frame. A framework for OF estimation on such image sequences is presented, that can straightforwardly integrate techniques from the state-of-the-art in conventional OF methods. Different aspects of robustness of OF methods are discussed, including estimation of large displacements and robustness to natural illumination changes that occur between the frames, and we demonstrate experimentally how to handle such challenging flow estimation scenarios. The flow estimation is formulated as an optimization problem whose solution is obtained using an efficient primal–dual method.

Super-resolution reconstruction of high dynamic range images in a perceptually uniform domain

Abstract. Super resolution is a signal processing method that utilizes information from multiple degraded images of the same scene in order to reconstruct an image with enhanced spatial resolution. The method is typically employed on similarly exposed pixel valued images, but it can be extended to differently exposed images with a high combined dynamic range. We propose a novel formulation of the joint super-resolution, high dynamic range image reconstruction problem, using an image domain in which the residual function of the inverse problem relates to the perception of the human visual system. Simulated results are presented, including a comparison with a conventional method, demonstrating that the proposed approach avoids some severe reconstruction artifacts.

Partial discharge characteristics of electrical treeing in XLPE insulation exposed to voltages of different rise times

The use of pulse width modulated (PWM) waveforms allows a more flexible energy applications and management. The downside is that under such operating conditions the voltage stress imposed on the insulation systems increases, particularly as the frequency content is considerably higher than at the conventional 50 Hz sinusoidal waveform. Among various degradation mechanisms, electrical treeing is one important processes and it can be linked with a presence of partial discharge (PD) activity. In this paper, the changes in PD characteristics at different stages of degradation by electrical treeing were investigated for XLPE based insulation exposed to rapidly changing voltages of varying magnitude. The tested material samples were subjected to two square voltage shapes, each characterized by their rise-time. Here the shorter was 1 ps and the longer 0.5 ms, all tests run at 414 Hz. To compare the voltage endurance, the voltage level was gradually increased until detectable PD activity and tree initiation could be observed, both electrically and optically. Typically continuous PD appearance was observed between 23 and 32 kVpp. The total number of PDs and their characteristics were monitored as the trees were gradually growing. Here the experimental results shows that the PDs appears at lower voltage magnitudes for the shorter rise times and that the trees had a significantly less branched appearance as compared to the longer voltage rise time. These results resemble the observations that recently been presented with tests on DC voltage pre-stressed material samples exposed to impulses of reversed polarity. Additionally the measured PD characteristics were non-symmetric, having a clear polarity dependence. This finding suggests a possibility to use the presented approach for a more detailed investigating of degradation processes in solid insulation systems suitable for both HVAC and HVDC applications.

Variational optical flow estimation for images with spectral and photometric sensor diversity

Motion estimation of objects in image sequences is an essential computer vision task. To this end, optical flow methods compute pixel-level motion, with the purpose of providing low-level input to higher-level algorithms and applications. Robust flow estimation is crucial for the success of applications, which in turn depends on the quality of the captured image data. This work explores the use of sensor diversity in the image data within a framework for variational optical flow. In particular, a custom image sensor setup intended for vehicle applications is tested. Experimental results demonstrate the improved flow estimation performance when IR sensitivity or flash illumination is added to the system.

Super-Resolution reconstruction of High Dynamic Range images with perceptual weighting of errors

Super-Resolution and High Dynamic Range image reconstruction are two different signal processing techniques that share in common that they utilize information from multiple observations of the same scene to enhance visual image quality. In this paper, both techniques are merged in a common model, and the focus is to solve the reconstruction problem in a suitable image domain, which relates to the perception of the Human Visual System. Simulated results are presented, including a comparison with a conventional method, demonstrating the benefits of the proposed approach, in this case avoiding some severe reconstruction artifacts.