This paper presents an adaptive demosaicing algorithm. Missing green samples are first estimated based on the variances of color differences along different edge directions. The missing red and blue components then interpolated plane. algorithm can effectively preserve details in texture regions and, at same time, it significantly reduce artifacts. As compared with latest algorithms, proposed produces best average performance both objectively subjectively

When imaging through a semi-reflective medium such as glass, the reflection of another scene can often be found in captured images. It degrades quality images and affects their subsequent analyses. In this paper, novel deep neural network approach for solving problem is presented. Traditional removal methods not only require long computation time different optimization functions, performance also guaranteed. As array cameras are readily available nowadays devices, we first suggest paper...

A unified approach to the realization of forward and inverse discrete cosine transforms is proposed. With this approach, an odd prime length DCT/IDCT with two half-length convolutions can be realized without extra overhead in terms number multiplications. The formulation most suitable for using distributed arithmetic, which case typical convolvers used as core unit hardware implementation transforms. Hence, efficient chip proposed demonstrate superiority formulation. architecture easily meet...

In most digital cameras, Bayer color filter array (CFA) images are captured and demosaicing is generally carried out before compression. Recently, it was found that compression-first schemes outperform the conventional demosaicing-first in terms of output image quality. An efficient prediction-based lossless compression scheme for CFA proposed this paper. It exploits a context matching technique to rank neighboring pixels when predicting pixel, an adaptive difference estimation remove...

In daily photography, it is common to capture images in the reflection of an unwanted scene. This circumstance arises frequently when imaging through a semi-reflecting material such as glass. The will affect visibility background image and introduce ambiguity that perturbs subsequent analysis on image. It very challenging task remove since problem severely ill-posed. this paper, we propose novel algorithm solve removal based light field (LF) imaging. For proposed algorithm, first show strong...

Multiscale error diffusion is superior to conventional methods in digital halftoning as it can eliminate directional hysteresis completely. However, there a bias favor particular type of dots the course process. A new multiscale method proposed improve performance by reducing aforementioned bias. The pattern noise flat regions and boundary effect found some other methods. At same time, preserve local features input image output. This critical quality, especially when resolution output...

This letter presents a new approach to reduce coding artifacts in transform image coding. We the problem an estimation of each coefficient from its quantized version with local mean and variance. The proposed method can significantly low bit-rate coded images, at same time guarantee that resulting images satisfies quantization error constraint.

Presents two new fast discrete cosine transform computation algorithms: a radix-3 and radix-6 algorithm. These algorithms are superior to the conventional algorithm as they (i) require less computational complexity in terms of number multiplications per point, (ii) provide wider choice sequence length for which DCT can be realized and, (iii) support prime factor-decomposed realize 2/sup m/3/sup n/-point DCT. Furthermore, mixed-radix is also proposed such that an optimal performance achieved...

Multiscale error diffusion (MED) is superior to conventional algorithms as it can eliminate directional hysteresis completely and possesses a good blue noise characteristic. However, due its filter design, not suitable for systems with poor isolated dot generation instable gain. In this paper, we propose MED algorithm produce halftones of desirable green characteristics. This allows one adjust the cluster size freely through single parameter supports linear relationship between input gray...

In a fringe projection profilometry (FPP) process, the captured images can be modeled as superimposition of projected patterns on texture objects. Extracting from is an essential procedure in FPP, but traditional single-shot FPP methods often fail to perform if objects have highly textured surface. this paper, new algorithm which allows object and pattern estimated simultaneously, proposed. The heart proposed enhanced morphological component analysis (MCA) tailored for problems. Conventional...

Reversible color-to-grayscale conversion (RCGC) is a method that embeds the chromatic information of full color image into its grayscale version such original can be reconstructed in future when necessary. In practical applications, it required to provide means authenticate an information-embedded integrity guaranteed. However, none current RCGC algorithms take this factor account. paper, address issue, we develop information-embedding framework based on vector quantization-based (VQ-based)...

This paper presents a low complexity joint color demosaicking and digital zooming algorithm for single-sensor cameras. The proposed directly extracts edge information from raw sensor data interpolation in both to preserve features its output. allows the extracted be exploited consistently stages also efficiently, as no separate extraction process is required different stages. can produce zoomed full-color image well Bayer filter array with outstanding performance compared conventional...

A digital halftone of blue noise characteristics is preferred as dots in the a constant input should be isotropically and homogeneously distributed. In practice, placement constrained by sampling grid hence aliasing happens when gray level middle range. To solve this problem, Lau et al. suggested replacing isolated dot clusters to maintain principal frequency output 1/2 happens. However, model does not take into account that due stochastic nature distribution there considerable amount energy...

We generalize a formulation for converting length-2/sup n/ discrete cosine transform into n groups of equations, then apply novel technique its implementation. The sizes the are 2/sup n-1/, n-2/, ...2/sup 0/ respectively, while their structures extremely regular. realization can be converted recursive filter form, which is particularly simple practical implementation.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

A digital halftoning method is proposed based on a multiscale error diffusion technique for halftoning. It can improve the performance by effectively removing pattern noise and eliminating boundary "blackhole" effects. dot overlap compensation scheme also to eliminate bias in gray scale of printed images.

This paper studies the restoration of images which are color-quantized with error diffusion. Though there many reported algorithms proposed for restoring noisy blurred color and inverse halftoning, is rarely addressed in literature especially when halftoning. Direct application existing techniques generally inadequate to deal this problem. In paper, a algorithm based on projection onto convex sets proposed. makes use available palette mechanism halftoning process derive useful priori...

Restoring an image from its convolution with unknown blur function is a well-known ill-posed problem in processing. Many approaches have been proposed to solve the and they shown good performance identifying restoring original image. However, actual implementation, various problems incurred due large data size long computational time of these are undesirable even current computing machines. In this paper, efficient algorithm for blind restoration based on discrete periodic Radon transform...

Purpose – This paper aims at presenting the benefits from implementing IT service management (ITSM) in an organization for managing campus‐wide operations. In order to improve fault correlation business perspectives, we proposed a framework automate network and system alerts with respect its impact proactive notification operations management.

The conventional blue-noise model that specifies the desired noise characteristics of an ideal halftone has been updated recently, and simulation results showed can serve as a better guideline for developing algorithms. At moment, only feature-preserving multiscale error diffusion-based algorithm was developed based on model. As does not support real-time applications, tone-dependent diffusion (TDED) is To proposed TDED algorithm, we optimize filter quantizer threshold each possible input...

Nowadays Location Based Services applications are increasingly useful. However, problems like floor identification for multi-buildings and adverse effects of devices diversity needed to be resolved. In this paper we propose a new approach using cosine similarity computed by Wi-Fi fingerprints radio map Convolutional Neural Network (CNN) model achieve multi-floor classification. We in use locations-based as the feature vector instead conventional Access Point sets. also timesaving walk-survey...

The authors propose using the correlated cosine structure (CCS) for computation of discrete transform (DCT). This has circulant property and is most suitable hardware realization. They show that there exists a close relationship between CCS DCT. In such case, 2/sup m/ length DCT can be decomposed recursively into shorter new approach results in very simple straightforward gives minimum number multiplications its realization.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML"...