Spatiotemporal fusion is important in providing high spatial resolution earth observations with a dense time series, and recently, learning-based methods have been attracting broad interest. These algorithms project image patches onto feature space the enforcement of simple mapping to predict fine from corresponding coarse ones. However, sophisticated projection, e.g., sparse representation, always computationally complex difficult be implemented on large patches, which cannot grasp enough...

Numerous state-of-the-art perceptual image quality assessment (IQA) algorithms share a common two-stage process: distortion description followed by effects pooling. As for the first stage, descriptors or measurements are expected to be effective representatives of human visual variations, while second stage should well express relationship among and quality. However, most existing (e.g., luminance, contrast, gradient) do not seem consistent with perception, pooling is often done in ad-hoc...

Abstract In this paper, a compensation control model of secondary cooling process billet continuous casting for quality steel has been presented. The effects on the spray various parameters such as superheat, speed, water temperature and chemical component were considered. determined to associate with two‐dimensional heat transfer equation solved by finite‐difference method. Effects superheat surface temperature, solidification structure solidifying end point discussed. Results indicate that...

Human vision is often adversely affected by complex environmental factors, especially in night scenarios. Thus, infrared cameras are leveraged to help enhance the visual effects via detecting radiation surrounding environment, but videos undesirable due lack of detailed semantic information. In such a case, an effective video-to-video translation method from domain visible light counterpart strongly needed overcoming intrinsic huge gap between and fields. To address this challenging problem,...

Cloud detection is an important task in remote sensing (RS) image processing. Numerous cloud algorithms have been developed. However, most existing methods suffer from the weakness of omitting small and thin clouds, inability to discriminate clouds photometrically similar regions, such as buildings snow. Here, we derive a novel algorithm for optical RS images, whereby test images are separated into three classes: thick noncloudy. First, simple linear iterative clustering adopted that able...

Noise that afflicts natural images, regardless of the source, generally disturbs perception image quality by introducing a high-frequency random element that, when severe, can mask content. Except at very low levels, where it may play purpose, is annoying. There exist significant statistical differences between distortion-free images and noisy become evident upon comparing empirical probability distribution histograms their discrete wavelet transform (DWT) coefficients. The DWT coefficients...

Diffusion models have achieved remarkable progress in low-light image enhancement. However, there remain two practical limitations: (1) existing methods mainly focus on the spatial domain for diffusion process, while neglecting essential features frequency domain; (2) conventional patch-based sampling strategy inevitably leads to severe checkerboard artifacts due uneven overlapping. To address these limitations one go, we propose a Multi-Domain Multi-Scale (MDMS) model In particular,...

Most of the existing image blurriness assessment algorithms are proposed based on measuring edge width, gradient, high-frequency energy, or pixel intensity variation. However, these methods content sensitive with little consideration variations, which causes variant estimations for images different contents but same degrees. In this paper, a content-insensitive blind metric is developed utilizing Weibull statistics. Inspired by property that statistics gradient magnitude (GM) follows...

The existing sparse representation-based visual trackers mostly suffer from both being time consuming and having poor robustness problems. To address these issues, a novel tracking method is presented via combining representation an emerging learning technique, namely extreme machine (ELM). Specifically, can be divided into two consecutive processes. Firstly, ELM utilized to find the optimal separate hyperplane between target observations background ones. Thus, trained classification...

Image fusion technology aims to obtain a comprehensive image containing specific target or detailed information by fusing data of different modalities. However, many deep learning-based algorithms consider edge texture through loss functions instead specifically constructing network modules. The influence the middle layer features is ignored, which leads between layers. In this article, we propose multidiscriminator hierarchical wavelet generative adversarial (MHW-GAN) for multimodal fusion....

Image distortion can be categorized into two aspects: content-dependent degradation and content-independent one. An existing full-reference image quality assessment (IQA) metric cannot deal with these different impacts well. Singular value decomposition (SVD) as a useful mathematical tool has been used in various processing applications. In this paper, SVD is employed to separate the structural (content-dependent) components. For each portion, we design specific model tailor for its...

Explicit visible videos can provide sufficient visual information and facilitate vision applications. Unfortunately, the image sensors of cameras are sensitive to light conditions like darkness or overexposure. To make up for this, recently, infrared capable stable imaging have received increasing attention in autonomous driving monitoring. However, most prosperous models still trained on massive clear data, facing huge gaps when deploying scenarios. In such cases, transferring video a...

A gradient composite coating of Cr–AlN with an outer Cr-rich layer and inner AlN-rich was deposited on the alumina-forming austenitic (AFA) steel surface by pack-cementation chromising. The phase composition, element distribution, microstructure were characterised analysed. microhardness friction wear performance at 600 °C tested. results show that mainly consists a polycrystalline Cr–Fe solid solution small amount Cr 2 C (Cr, Fe) N 1− x particles. contains plenty AlN particles different...

Infrared cameras are often utilized to enhance the night vision since visible light exhibit inferior efficacy without sufficient illumination. However, infrared data possesses inadequate color contrast and representation ability attributed its intrinsic heat-related imaging principle, which hinders application. Although, domain gaps between unpaired nighttime daytime videos even huger than paired ones that captured at same time, establishing an effective translation mapping will greatly...

A novel object tracking method is proposed that takes advantage of the fast learning capability extreme machine (ELM). Specifically, viewed as a binary classification problem, and ELM utilised for finding optimal separate hyperplane between backgrounds efficiently. To achieve more robust tracking, two constraints are introduced in training: (i) target visual changes across frames smooth (i.e. smoothness) (ii) probabilities to be true image samples around tracked trajectory preferred than...

In this paper, we propose a novel image quality assessment (IQA) metric based on nonnegative matrix factorization (NM-F). With nonnegativity and parts-based properties, NMF well demonstrates how human brain learns the parts of objects. This makes distinguished from other feature extraction methods like singular value decomposition (SVD), principal components analysis (PCA), etc. Inspired by this, adopt to extract features reference distorted images. Extreme learning machine (ELM) [10] is...

Multi-exposure image fusion is becoming increasingly influential in enhancing the quality of experience consumer electronics. However, until now few works have been conducted on performance evaluation multi-exposure fusion, especially colorful fusion. Conventional assessment methods for mainly focus grayscale information, while ignoring color components, which also convey vital visual information. We propose an objective method colored based saturation, together with texture and structure...

At present the visual tracking model based on sparse representation is mainly divided into two types: one to use template set reconstruct candidate samples, which called forward model; other project a space, reverse model. What models have in common compute correlation coefficient matrix of sample and set. Based this, paper establishes bidirectional cooperative Using L2-norm constraint item, coefficients could be uniformly convergent. In comparison conventional unidirectional model, fully...