Poisson noise is an important type of electronic that present in a variety photon-limited imaging systems. Different from the Gaussian noise, depends on image intensity, which makes restoration very challenging. Moreover, complex geometry images desires regularization capable preserving piecewise smoothness. In this paper, we propose denoising model based fractional-order total variation (FOTV). The existence and uniqueness solution to are established. To solve problem efficiently, three...

Integrating a low-resolution hyperspectral image and high-resolution multispectral is widely acknowledged as an effective approach for generating image. Recent studies have highlighted the nuclear norm efficient method this problem through utilization of low-rankness. However, standard has limitation due to treating singular values equally. To address issue, we incorporated concept weighted from denoising into fusion, ensuring retention crucial data components. Furthermore, propose unified...

Timing jitter as a key performance of single-photon avalanche diode (SPAD) detectors plays significant role in determining the fast temporal response behavior SPAD device. Nevertheless, few analytic models are developed to directly calculate characteristic timing for its modeling difficulty. In this paper, we propose simple method, which can predict SPADs, without using time-consuming Monte Carlo simulation. Model investigation incorporates current, buildup time, and tail under different...

The automation edge extraction of glomerulus is an important step for analyzing kidney-tissue image in the computer aided diagnosis system kidney disease. According to characteristic these medical images, this paper proposes a method based on wavelet transformation and watershed algorithm. First, LOG filter applied low-resolution image, which corresponds low frequency sub band after transformation, so that rough information can be obtained. After labeling remove noises thinning, genetic...

Abstract High order total variation (TV2) and ℓ 1 based (TV2L1) model has its advantage over the TVL1 for ability in avoiding staircase; a constrained unconstrained counterpart simplicity estimating parameters. In this paper, we consider solving TV2L1 magnetic resonance imaging (MRI) signal reconstruction problem by an efficient alternating direction method of multipliers. By sufficiently utilizing problem's special structure, manage to make all subproblems either possess closed-form...

Abstract The quality of Chinese characters written by foreign students is reflected not only in the result writing, but also writing movement. 25 characteristics handwriting movement were selected from 39 parameters, such as time, space, and pressure. Based on feature motion, 6 classifiers, Decision Tree Artificial Neural Network, are to compare performance them. Experiments show that accuracy DT algorithm best regardless whether target known or not. Handwriting features effective...

In this paper, a new relaxation model based on mean curvature for adaptive image restoration is proposed. To solve the problem efficiently, an alternating direction method of multipliers (ADMMs) Furthermore, rigorous convergence theory proposed algorithm established. We also give complexity analysis our method. Experimental results are provided to demonstrate effectiveness and efficiency over state-of-the-art synthetic natural images.

We propose a matched-field processing framework for tracking problems in shallow water environments where the conventional plane-wave assumptions do not hold. Multiple passive acoustic sensors are employed to collect observation data, and sequential Monte Carlo techniques used due high nonlinearity dynamic state formulation. In order enhance performance, we design frequency selection algorithm which adaptively chooses optimal at each time instant. The improved performance is demonstrated...

In noisy condition, the accurate measurement of Lipschitz exponent is very important for many actual applications.All modulus maxima are utilized to estimate in each maximum line traditional wavelet transform (WTMM) method, but as positions and coefficients corrupted some scales by noise, so performance estimation poor.For this proposed unpolluted filtrated line, then these corresponding used obtain a pre-estimation exponent.Based on pre-estimation, different weighted operations applied...