Ante Poljičak

Discrete Fourier transform–based watermarking method with an optimal implementation radius

In this paper, we evaluate the degradation of an image due to the implementation of a watermark in the frequency domain of the image. As a result, a watermarking method, which minimizes the impact of the watermark implementation on the overall quality of an image, is developed. The watermark is embedded in magnitudes of the Fourier transform. A peak signal-to-noise ratio is used to evaluate quality degradation. The obtained results were used to develop a watermarking strategy that chooses the optimal radius of the implementation to minimize quality degradation. The robustness of the proposed method was evaluated on the dataset of 1000 images. Detection rates and receiver operating characteristic performance showed considerable robustness against the print-scan process, print-cam process, amplitude modulated, halftoning, and attacks from the StirMark benchmark software.

Portable real-time DCT-based steganography using OpenCL

Abstract In this paper, a steganographic method for real-time data hiding is proposed. The main goal of the research is to develop steganographic method with increased robustness to unintentional image processing attacks. In addition, we prove the validity of the method in real-time applications. The method is based on a discrete cosine transform (DCT) where the values of a DCT coefficients are modified in order to hide data. This modification is invisible to a human observer. We further the investigation by implementing the proposed method using different target architectures and analyze their performance. Results show that the proposed method is very robust to image compression, scaling and blurring. In addition, modification of the image is imperceptible even though the number of embedded bits is high. The steganalysis of the method shows that the detection of the modification of the image is unreliable for a lower relative payload size embedded. Analysis of OpenCL implementation of the proposed method on four different target architectures shows considerable speedups.

Copyright protection of images on a social network

In this paper we developed a steganographic method for image protection on Facebook social network. The method hides additional information such as a logo or serial number. Uploading the images on the social network has induced image degradation on the original images. Based on the analysis of this attack we developed a method based on the discrete cosine transform - DCT where additional information is hidden by modifying the coefficients in the cosine domain. The method was evaluated and it is shown that bit error rate (BER) for different attacks was below 1%.

Colorimetrically accurate gray component replacement using the additive model

New gray component replacement method for use in four color printing is developed.The methods performance is tested on a large dataset and image database.The method achieves great ink saving while preserving the visual appearance.Further research key points for improving the methods versatility are identified. In four color printing, gray component replacement (GCR) is a method of replacing the achromatic component of a mixture of chromatic inks by a corresponding amount of the achromatic ink. Simple approaches to GCR lead to unacceptable color shifts. In this paper, the method using multiple processing steps is proposed. The novelty of the proposed method is augmentation of masking equations, which allows finding solutions for a pre-set amount of the black ink. The method performance was evaluated and compared with state of the art commercial solution. The experiments have shown that the proposed model is on average capable of achieving 4055% ink savings with median colorimetric difference of less than 0.3, thus preserving the visual appearance of images.

An optimized radial basis function model for color characterization of a mobile device display

This paper presents an optimized color characterization model based on Radial Basis Functions (RBFs). The performance of the proposed model was tested on a number of different mobile devices and compared with the performance of other state of the art color characterization models. We compared the accuracy of models using the CIELAB color difference. Four different models were discussed in detail: Piecewise Linear Model Assuming Variation in Chromaticity, Polynomial regression, Artificial Neural Network, and proposed Radial Basis Function model. For training and evaluation of the models we measured a large number of color samples on various mobile device displays. Results have shown that our optimized RBF model has superior accuracy over other models with median color difference of 0.39. In addition, it has particularly good accuracy for colors on the boundary of device’s gamut with maximum color difference of 0.87, where other models shown unacceptably high (>10) color difference.

Applicability of the GCR masking in an image watermarking method

There exist continuous effort to develop new or improve existing watermarking methods. The main goal of this paper is to give preliminary insight of the applicability of the Gray Component Replacement (GCR) masking to frequency domain based watermarking scheme. The use of GCR masking can potentially increase the performance of a watermarking method by improving robustness or imperceptibility of an embedded watermark. The applicability of GCR masking is tested by extending well known watermarking method with GCR masking capability and comparing the imperceptibility of the embedded watermark before and after GCR masking. Results show that GCR masking can substantially increase the PSNR value of the watermarked image.

Adjustment of visual information for visually impaired people

The aim of this work is to test algorithms for image enhancement with different values in the algorithm and to test those algorithms on the different accelerators that could be used for medical aids. The results of this research show advantages and disadvantages of equipment that we used and also can indicate implantation of algorithm into portable devices. Five selected images were filtered through proposed algorithm and compared with golden standard through metrics MSE, PSNR and SSIM. Speeup for edge detector was tested on different machines.

The Model of Degradation of Color Values in a Reproduction Process

This paper deals with the problem of the degradation of color information that happens when color image is converted from one medium to another. The purpose of experiment was to determine the behaviour between two closely related colors when transferred on paper and then digitalized back using scanner. The aim of research was to examine the possibility of information hiding by means of modification of RGB channels of a digital color image.

Case study: Digital offset printing- resolution and color reproduction

In this paper, the influence of resolution on the color quality reproduction in digital offset printing were investigated. In our experiment special test charts were printed with different resolutions and compared. The results show that high resolution improve image contrast but not the color reproduction.