Image smoothing is a prerequisite for many computer vision and graphics applications. In this article, we raise an intriguing question whether a dataset that semantically describes meaningful ...structures and unimportant details can facilitate a deep learning model to smooth complex natural images. To answer it, we generate ground-truth labels from easy samples by candidate generation and a screening test and synthesize hard samples in structure-preserving smoothing by blending intricate and multifarious details with the labels. To take full advantage of this dataset, we present a joint edge detection and structure-preserving image smoothing neural network (JESS-Net). Moreover, we propose the distinctive total variation loss as prior knowledge to narrow the gap between synthetic and real data. Experiments on different datasets and real images show clear improvements of our method over the state of the arts in terms of both the image cleanness and structure-preserving ability. Code and dataset are available at https://github.com/YidFeng/Easy2Hard .
This study explores the smoothing effect on the uncertainty of the wind power fluctuations that affect a power system at points of common coupling. Set pair analysis is applied to evaluate the ...similarity between the power fluctuations of a single wind turbine and those of all aggregated wind turbines, based on which a quantitative index describing the wind power smoothing effect is proposed. The smoothing effect characteristics of a wind farm cluster are investigated for different numbers of wind turbines, different wind speeds, different seasons, and multiple sampling intervals. A significant smoothing effect is usually observed at a shorter sampling interval and a higher wind speed, and the smoothing effect index increases with an increase in the numbers of wind turbines and farms. Additionally, the correlation between the smoothing effect of aggregated wind farms and the forecast accuracy for the corresponding aggregated power output is examined. The experimental results indicate that the wind power forecast accuracy varies with the smoothing effect index, which is influenced by the number of wind farms. Furthermore, the aggregated output from a wind farm cluster with a higher smoothing effect index exhibits better forecasting performance than that from a single wind farm, showing that the trend of the wind power series becomes smoother due to the smoothing effect, thus enabling one-step-ahead wind power forecasting with higher accuracy.
In this note we shall present a new Gaussian approximation based framework for approximate optimal smoothing of non-linear stochastic state space models. The approximation framework can be used for ...efficiently solving non-linear fixed-interval, fixed-point and fixed-lag optimal smoothing problems. We shall also numerically compare accuracies of approximations, which are based on Taylor series expansion, unscented transformation, central differences and Gauss-Hermite quadrature.
This article presents a novel ramp rate control and active power smoothing and shifting methodology for net-load profiles in large power distribution networks where high amount of photovoltaic (PV) ...penetration levels exist. The novelty is that the proposed methodology uses dynamic state-of-charge (SoC) management, energy storage optimal use of any given size and fast-Fourier transform-based reference curves that adequately fit energy storage system (ESS) active and reactive power capabilities. Then, a least square minimization-based optimization method is applied taking both ancillary applications into consideration and to maintain SoC within limits based on ESS capacity. The methodology is applied on a 8500 nodes US power grid model using CYMEDist simulation platform and applied on a real US power distribution system. Results show significant level of net-load ramp rate reduction as well as smoothing of fast fluctuations (up to 60%) due to both PV intermittency and load changes. The approach is also capable of breaking large systems into virtual local net-load locations where ESS's are installed to mitigate the severity of the anticipated shape of system net-loads.
Subspace-based methods suffer from the rank loss of the noise free data covariance matrix in the context of direction of arrival (DOA) estimation of coherent sources. The well-known spatial smoothing ...techniques are then widely employed to create a rank restored data covariance matrix. However, conventional spatial smoothing techniques, such as the spatial smoothing pre-processing (SSP), modified spatial smoothing pre-processing (MSSP), and improved spatial smoothing (ISS), do not make full use of the available information in the data covariance matrix. In this paper, an enhanced spatial smoothing (ESS) technique is proposed to exploit both the covariance matrices of individual subarrays and the cross-covariance matrices of different subarrays. Besides, the proposed method can work directly on the signal subspace (ESS-SS), since the signal subspace contains all the information of the DOAs of incoming signals. After de-correlation, the subspace method ESPRIT is adopted to estimate the DOAs. Compared with conventional approaches, the proposed method is more powerful to de-correlate the correlation between signals, and also more robust to the noise impact. The proposed method is tested on numerical data in coherent scenarios, and compared with conventional approaches. Simulation results show that the proposed method has an enhanced resolving capability and a lower signal-to-noise ratio threshold.
Image smoothing is a fundamental procedure in applications of both computer vision and graphics. The required smoothing properties can be different or even contradictive among different tasks. ...Nevertheless, the inherent smoothing nature of one smoothing operator is usually fixed and thus cannot meet the various requirements of different applications. In this paper, we first introduce the truncated Huber penalty function which shows strong flexibility under different parameter settings. A generalized framework is then proposed with the introduced truncated Huber penalty function. When combined with its strong flexibility, our framework is able to achieve diverse smoothing natures where contradictive smoothing behaviors can even be achieved. It can also yield the smoothing behavior that can seldom be achieved by previous methods, and superior performance is thus achieved in challenging cases. These together enable our framework capable of a range of applications and able to outperform the state-of-the-art approaches in several tasks, such as image detail enhancement, clip-art compression artifacts removal, guided depth map restoration, image texture removal, etc. In addition, an efficient numerical solution is provided and its convergence is theoretically guaranteed even the optimization framework is non-convex and non-smooth. A simple yet effective approach is further proposed to reduce the computational cost of our method while maintaining its performance. The effectiveness and superior performance of our approach are validated through comprehensive experiments in a range of applications. Our code is available at https://github.com/wliusjtu/Generalized-Smoothing-Framework .
We present a new efficient edge-preserving filter-"tree filter"-to achieve strong image smoothing. The proposed filter can smooth out high-contrast details while preserving major edges, which is not ...achievable for bilateral-filter-like techniques. Tree filter is a weighted-average filter, whose kernel is derived by viewing pixel affinity in a probabilistic framework simultaneously considering pixel spatial distance, color/intensity difference, as well as connectedness. Pixel connectedness is acquired by treating pixels as nodes in a minimum spanning tree (MST) extracted from the image. The fact that an MST makes all image pixels connected through the tree endues the filter with the power to smooth out high-contrast, fine-scale details while preserving major image structures, since pixels in small isolated region will be closely connected to surrounding majority pixels through the tree, while pixels inside large homogeneous region will be automatically dragged away from pixels outside the region. The tree filter can be separated into two other filters, both of which turn out to have fast algorithms. We also propose an efficient linear time MST extraction algorithm to further improve the whole filtering speed. The algorithms give tree filter a great advantage in low computational complexity (linear to number of image pixels) and fast speed: it can process a 1-megapixel 8-bit image at ~ 0.25 s on an Intel 3.4 GHz Core i7 CPU (including the construction of MST). The proposed tree filter is demonstrated on a variety of applications.
The junction temperature fluctuation of an insulated-gate bipolar transistor (IGBT) is the most important factor of its aging failure, and smoothing the fluctuation is an effective way to improve the ...life of an IGBT. The existing methods for smoothing the fluctuation by active junction temperature control are not yet ready wide application, and exploring the different approaches to active junction temperature control is a hot topic. This paper presents a method of active junction temperature control that shifts the turn-off trajectory of an IGBT to adjust the IGBT turn-off loss for smoothing the junction temperature. The relationship between parameters of the adjusting circuit and turn-off loss is analyzed. On the basis of this analysis, a method of estimating the smoothing ability for the proposed active junction temperature control is deduced. Using an IGBT installed in a 1.2-MW direct-drive wind power converter as an example, the evaluation result shows that the proposed method can completely smooth the junction temperature fluctuation caused by a 40% rated load fluctuation. Finally, a low-power experiment is carried out.
Diversity smoothing has been widely used for angle estimation with multiple input multiple output (MIMO) radar in the presence of coherent or correlated targets, and the parameter identifiability is ...an interesting issue. Previous works have shown sufficient conditions for some special cases, such as the coherent targets with conventional MIMO radar, and the array size are derived as a function of the target number and target structure. In this article, we further introduce the interelement spacings to build a complete parameter identifiability scheme. For monostatic MIMO radars, the antenna numbers and interelement spacings of transmit and receive arrays are derived as functions of the target number and the target structure. The optimization models are built to calculate the maximum number of detectable targets for a given target structure. Additionally, the conditions for the bistatic MIMO radar are derived from two-dimension viewpoint. It is shown that the new results improve upon previous spatial smoothing or diversity smoothing methods and recover them in special cases. Simulation results are presented that corroborate our theoretical findings.
Conventional wind power smoothing control adopts hard-coded filtering algorithms to produce smoothed power output without considering the actual system need. In this letter, we propose a novel ...smoothing control paradigm in context of performance-based regulation service, in which the actual balancing need is considered and formulated as an automatic generation control regulation mileage. The proposed smoothing objective is to alleviate the regulation mileage while maximize wind energy harvesting. The effectiveness of the proposed framework is demonstrated through comparable case studies, through which the simulation results suggest a high potential for practical applications.