Dual Reversible Data Hiding in Encrypted Halftone Images Using Matrix Encoding

Face recognition is done gray or color images. Face recognition can be also done in binary images. Halftone images are binary images having gray shade information. These images are more informative as compared to binary images. Halftone images can be generated from different kernels. The quality of a halftone image depends on the kernel used to generate it. Halftone images can be used to generate human recognizable features for face recognition. The human recognizable features are fed to multiclass SVM classifiers for face recognition. The performance of using different human recognizable features generated from halftone images of different standard kernels is studied. It is found that some of kernels give higher recognition rate than the other. Also, the performance is also dependent on size of the window used for feature extraction. In most cases, the minimum size window 3 × 3 gives higher recognition rate in most of the cases. Of the standard kernels Frankie Siera-3, Stucki kernels give higher recognition rate and of the proposed kernels P2 kernel gives higher recognition rate than other proposed kernels.

We presents a new method for color printers.Our algorithm applies to both a color printer model and a model for the human visual system (HVS).The method strives to minimize the perceived error between the continuous original image and the halftone image.The color printer models can account for a variety of printer characteristics.We propose a specific printer model that accounts for overlap between neighboring dots of ink and spectral absorption properties of the inks.The color image is partitioned into several regions using clustering segmentation method, whose performance depends on the selection of distance metrics.Each clustering uses the least-squares model-based(Lsmb) algorithm to obtain halftone image.A performance measure for halftone images is used to evaluate our algorithm.Analysis and simulation results show that the proposed algorithm produces better color halftone image quality when we increase the number of clustering with a certain range.

We present general and unified algorithms for lossy/lossless coding of bilevel images. The compression is realized by applying arithmetic coding to conditional probabilities. As in the current JBIG standard the conditioning may be specified by a template. For better compression, the more general free tree may be used. Loss may be introduced in a preprocess on the encoding side to increase compression. The primary algorithm is a rate-distortion controlled greedy flipping of pixels. Though being general, the algorithms are primarily aimed at material containing half-toned images as a supplement to the specialized soft pattern matching techniques that work better for text. Template based refinement coding is applied for lossy-to-lossless refinement. Introducing only a small amount of loss in half-toned test images, compression is increased by up to a factor of four compared with JBIG. Lossy, lossless, and refinement decoding speed and lossless encoding speed are less than a factor of two slower than JBIG. The (de)coding method is proposed as part of JBIG2, an emerging international standard for lossless/lossy compression of bilevel images.

In this paper, we present an image authentication scheme for halftone images based on the error diffusion technique. Our scheme converts a gray-level image into a halftone image by using the error diffusion technique and embeds the authentication codes in the obtained halftone image at the same time. From the experimental results, our proposed scheme can detect the tampered places according to the embedded authentication codes. Besides, the visual quality of the stego halftone image in our scheme is satisfactory.

In this paper, a low computational complexity method is proposed to embed a multi-tone visual pattern using the concept of secret sharing. One halftone image is obtained using regular error diffusion, and the other image is obtained using the proposed Generalized Noise-Balanced Error Diffusion (GNBEDF) in considering the properties of the watermark and the first halftone image in the same time. With the proposed method, the bi-level watermark embedding becomes a special case to this technology. The bi-level watermark decoding can be simply achieved via printing the two halftone images onto two transparencies and then superimposing them together to reveal the watermark pattern. However, as the multi-tone watermark is involved, the proposed Gaussian Difference (GD) approach is needed to perform the decoding to produce a decoded multi-tone watermark. Throughout this work, a lowpass filter obtained by Least-Mean-Square is employed to assess the image quality.

Digital halftone is the technique used to convert gray-scale images into binary ones, simulating gray shades by scattering appropriately black and white pixels. Sometimes, there arises the necessity of increasing the resolution of a halftone image. Some recent works have proposed a number of learning-based techniques to zoom binary images. However, they cannot consider a large neighborhood to decide the colors of the resolution-increased pixels, because their running time skyrockets with the growth of the window and the sample images sizes. The use of large window and samples is required to zoom halftone images accurately. This paper presents a new technique to zoom quickly and precisely images generated by any locally-decided halftone algorithm. It is based on decision-tree learning and it is very fast, even using a large window or large samples. The zoomed images obtained by this technique are incredibly sharp and accurate.

This paper presents an improved XOR-based Data Hiding Scheme (XDHS) to hide a halftone image in more than two halftone stego images. The hamming weight and hamming distance is a very important parameter affecting the quality of a halftone image. For this reason, we proposed a method that involves minimizing the hamming weights and hamming distances between the stego image and cover image in 2×2-pixel grids. Moreover, our XDHS adopts a block-wise operation to improve the quality of a halftone image and stego images. Furthermore, our scheme improves security by using a block-wise operation with A-patterns and B-patterns. Our XDHS method achieves a high quality with good security compared to the prior arts. An experiment verified the superiority of our XDHS compared with previous methods.

Reproduction couleur par trames irrégulières et semi-régulières

This paper presents new results in the halftoning of color images. The first automatic generation of the Wall Street Journal type halftones of gray-scale images was done using the M-lattice system and reported in Sherstinsky and Picard [1994]. The M-lattice system was derived from the reaction-diffusion model, first proposed by Turing [1952] in order to explain mammal coat patterns. The M-lattice is a non-linear dynamical system that is well-suited for a variety of applications formulated as constrained non-linear optimization. In particular, it can perform image processing operations that emphasize oriented patterns. The present study uses this property to extend the special-effects halftoning of gray-scale images to that of color images. The duality metric comes from the directionality information extracted by steerable filters from the gray-scale version of the original color image. The binary requirement is stated as an explicit constraint, and all three (red, green, and blue) halftone components are synthesized simultaneously by the M-lattice.

This work employs the well known weighted least squares method to optimization to produce halftone images using improved K-means clustering theory. Our algorithm applies to both a printer model and a model for the human visual system (HVS). In this algorithm, the improved K-means clustering method is used to segment an image several regions. In the halftone process, each clustering uses the weighted least-squares model-based(WLSMB) algorithm by use of direct binary search iterative method to obtain halftone image. Analysis and simulation results show that the proposed algorithm produces better gray-scale halftone image quality when we increase the number of clustering with a certain range and outperforms least-squares model-based algorithm in the PSNR (Peak Signal Noise Ratio), WSNR (Weighted Signal Noise Ratio) criteria.

In this paper, we propose a noble watermark embedding and extraction method for printed images. Watermark bits are hidden at pseudo-random locations within a printed halftone image during halftoning process. To remove the artifacts and distortion due to the inserted watermark bits, iterative error minimizing technique is used. During each iteration the pattern which decreases the halftone error is selected as the next pattern where the halftone error is defined as the difference of HVS-filtered printed halftone image and grayscale image. The inserted watermarks are extracted from the scanned images after a series of image processing operation which is described in detail. In order to be robust to cropping and rotation, the watermark is inserted periodically in halftone images. Experiments using real scanned images show that the proposed method is feasible method to hide the large amount of data within a halftone image without noticeable distortion and the watermark is robust to cropping and rotation.

Watermarking in halftone images with parity-matched error diffusion

Cluster dot dithering is one of the most common halfton- ing techniques. It is fast, low in complexity and allows for variability and inconsistencies in point spreads in printer outputs. Determina- tion of the basic dither cell size is critical for the quality of the half- toning. There is a basic tradeoff between large and small cell sizes: spatial resolution versus gray tone resolution. Large dither cell sizes produce good tone resolution but poorly reproduce spatial details in the image. Small dither cells, on the other hand, produce fine spatial resolution but lack the tone resolution which produces smooth gray tone gradients in halftone images. Typically, cluster dot dithering assumes a predefined dither cell size that compromises between fine detail reproduction and good gray tone reproduction. It is clearly advantageous to allow variability in the dither cell size using small cell sizes in image regions of fine details and using large cell sizes in image regions where gray tones are to be accurately reproduced. In this paper, we introduce and discuss several adaptive dithering techniques based on cluster dot dithering. © 1999 SPIE and IS&T. (S1017-9909(99)00402-X)

Security analysis of a reversible data hiding scheme in encrypted images by redundant space transfer

Screening techniques are widely used in the binary display of continuous-tone images by digital output devices. The quality of the halftone image resulting from such a nonlinear transformation is dependent on the dot profile employed. In the Fourier domain, aliasing degrades the halftone image. The relationship between image quality and dot profile is studied from this point of view. For a well-known dot profile, it is shown analytically how aliasing is suppressed, and that quantization contouring may be eliminated without an appreciable increase in aliasing.