Improved Dot Diffusion by Diffused Matrix and Class Matrix Co-Optimization

Image data hiding is the hiding of invisible patterns in an image without degrading its visual quality. The hidden pattern can be visualized by overlaying the original image with watermarked ones or by applying an extraction algorithm. If the hidden pattern is visualized using the overlaying method, it can also be visualized using a simple XNOR operation. In the case that hidden data is extracted by a specific algorithm, it is commonly called steganography. In this paper we propose a method called dot diffusion with nonlinear thresholding (DDNT) to embed hidden data in halftone images, where the hidden data can be extracted by overlaying original and watermarked images or by performing the XNOR process on these two images. The main advantage of this method is that the hidden pattern's intensity can be adjusted by three parameters. The DDNT data hiding method utilizes the inherent parallelism in dot diffusion halftoning method, although it suffers from its lower visual quality than some other halftoning methods such as error diffusion

An improved version of Block Truncation Coding (BTC) termed Multitone Block Truncation Coding (MTBTC) is proposed. The presented framework exploits the digital multi toning strategy to obtain a superior quality BTC images. Previously, the halftone techniques such as ordered dithering, error diffusion and dot diffusion are extended to obtain different version of BTC. Though it performs good, there are still many scope for improvements. In recent times digital multitoning is gaining prominence and advanced multitone screens such as direct multibit search and clustered dot screens are developed. This multitone screen can enhance the visual quality significantly and offer more homogenous and noise free output for halftone images. Considering this properties, an upgradation to the halftone based BTC is proposed in this paper through the use of sophisticated multitone screens. From the simulation results, it is concluded that the proposed framework can provide very high quality output than any existing BTC techniques. Another highlighting aspect is that the proposed technique can facilitate higher block size processing and it is computationally very simple. In overall, the proposed scheme is an ideal candidate to implement compression for high resolution images and can be exploited further to adopt in artificial intelligent tasks.

A digital micromirror device (DMD) provides 2D- and 3D-scene reconstruction by displaying diffractive and holographic optical elements. The highest frame rates (tens of thousands of Hz) can be achieved if the displayed optical elements are binarized. Except for DMD applications, hologram binarization is useful for display creation, image encryption, information compression and storage, fast 3D printing, etc. Error diffusion is one of the most qualitative implementations of hologram binarization. In this paper, three group's weighting matrices of error diffusion are analyzed: 16 standard, eight dot and six diagonal matrices. Thus, 30 error diffusion methods were used for optically recorded off-axis digital hologram binarization. The quality of the image reconstruction from the binarized holograms was compared. Direct applications of error diffusion with large weighting matrices and dot diffusion provide the highest reconstruction quality. Seven metrics were used as the error diffusion threshold. Twelve bypass directions of error diffusion were analyzed. In addition, the joint application of Otsu threshold and complex bypass directions allows us to improve the quality of hologram binarization by 15%.

The error diffusion method can effectively reduce quality degradation by propagating thresholding errors to neighboring pixels in the conversion of a gray-scale hologram to a binary hologram. In previous works, the four weighting coefficients in error diffusion are mostly set as the Floyd-Steinberg coefficient, which was determined empirically and originally proposed for photograph processing. In this work, we point out that the Floyd-Steinberg coefficients can be suboptimal for hologram error diffusion binarization. Furthermore, the weighting coefficients are optimized for each different hologram adaptively. Compared with conventional coefficients, our optimized coefficients can better preserve the fidelity of a reconstructed image after a hologram is binarized.

Nowadays since the Internet is ubiquitous, the frequency of data transfer through the public network is increasing. Hiding secure data in these transmitted data has emerged broad security issue, such as authentication and copyright protection. On the other hand, considering the transmission efficiency issue, image transmission usually involves image compression in Internet-based applications. To address both issues, this paper presents a data hiding scheme for the image compression method called absolute moment block truncation coding (AMBTC). First, an image is divided into non-overlapping blocks through AMBTC compression, the blocks are classified four types, namely smooth, semi-smooth, semi-complex, and complex. The secret data are embedded into the smooth blocks by using a simple replacement strategy. The proposed method respectively embeds nine bits (and five bits) of secret data into the bitmap of the semi-smooth blocks (and semi-complex blocks) through the exclusive-or (XOR) operation. The secret data are embedded into the complex blocks by using a hidden function. After the embedding phase, the direct binary search (DBS) method is performed to improve the image quality without damaging the secret data. The experimental results demonstrate that the proposed method yields higher quality and hiding capacity than other reference methods.

The objective of this work is to reconstruct high quality gray-level images from halftone images, or the inverse halftoning process. We develop high performance halftone reconstruction methods for several commonly used halftone techniques. For better reconstruction quality, image classification based on halftone techniques is placed before the reconstruction process so that the halftone reconstruction process can be fine tuned for each halftone technique. The classification is based on enhanced 1D correlation of halftone images and processed with a three- layer back propagation neural network. This classification method reached 100 percent accuracy with a limited set of images processed by dispersed-dot ordered dithering, clustered-dot ordered dithering, constrained average, and error diffusion methods in our experiments. For image reconstruction, we apply the least-mean-square adaptive filtering algorithm which intends to discover the optimal filter weights and the mask shapes. As a result, it yields very good reconstruction image quality. The error diffusion yields the best reconstructed quality among the halftone methods. In addition, the LMS method generates optimal image masks which are significantly different for each halftone method. These optimal masks can also be applied to more sophisticated reconstruction methods as the default filter masks.

A halftone watermarking method of high quality, robustness, and capacity flexibility is presented in this paper. An objective halftone image quality evaluation method based on the human visual system obtained by least-mean-square is also introduced. In the encoder, the kernels-alternated error diffusion (KAEDF) is applied. This is able to maintain the computational complexity at the same level as ordinary error diffusion. Compared with Hel-Or (2001) using ordered dithering, the proposed KAEDF yields a better image quality through using error diffusion. We also propose a weighted lookup table (WLUT) in the decoder instead of LUT, as proposed by Pei and Guo (2003), so as to achieve a higher decoded rate. As the experimental results demonstrated, this technique is able to guard against degradation due to tampering, cropping, rotation, as well as print-and-scan processes in error-diffused halftone images.

Reproduction couleur par trames irrégulières et semi-régulières

This paper proposes an adaptive error diffusion algorithm for text enhancement followed by an efficient text segmentation that uses the maximum gradient difference (MGD). The gradients are calculated along with scan lines, then the MGD values are filled within a local window to merge text segments. If the value is above a threshold, the pixel is considered as potential text. Isolated segments are then eliminated in a non-text region filtering process. After the text segmentation, a conventional error diffusion method is applied to the background, while edge enhancement error diffusion is used for the text. Since it is inevitable that visually objectionable artifacts are generated when using two different halftoning algorithms, gradual dilation is proposed to minimize the boundary artifacts in the segmented text blocks before halftoning. Sharpening based on the gradually dilated text region (GDTR) then prevents the printing of successive dots around the text region boundaries. The method is extended to halftone color images to sharpen the text regions. The proposed adaptive error diffusion algorithm involves color halftoning that controls the amount of edge enhancement using a general error filter. However, edge enhancement unfortunately produces color distortion, as edge enhancement and color difference are trade-offs. The multiplicative edge enhancement parameters are selected based on the amount of edge sharpening and color difference. Plus, an additional error factor is introduced to reduce the dot elimination artifact generated by the edge enhancement error diffusion. In experiments, the text of a scanned image was sharper when using the proposed algorithm than with conventional error diffusion without changing the background.

In order to be able to print an image on a binary output device, the image has to be rendered as a sequence of on/off (black/white) states. The image quality of the printed binary image strongly depends on the actual distribution of the black/white pixels. In order to increase the fidelity of the reproduction it is desirable that a feedback algorithm in the form of error diffusion be used. This type of algorithm, however, produces patterns that are not reproducible by standard xerographic printers, where a clustering of pixels into dots is required. This talk demonstrates a method to combine error diffusion with standard halftoning by first using a pixel based error correction followed by a halftone cell based error correction. This combination allows the generation of printable patterns that incorporate characteristics of feedback algorithms.

In this paper, a halftoning-based watermarking method is presented. This method enables high pixel-depth watermark embedding, while maintaining high image quality. This technique is capable of embedding watermarks with pixel depths up to 3 bits without causing prominent degradation to the image quality. To achieve high image quality, the parallel oriented high-efficient direct binary search (DBS) halftoning is selected to be integrated with the proposed orientation modulation (OM) method. The OM method utilizes different halftone texture orientations to carry different watermark data. In the decoder, the least-mean-square-trained filters are applied for feature extraction from watermarked images in the frequency domain, and the naïve Bayes classifier is used to analyze the extracted features and ultimately to decode the watermark data. Experimental results show that the DBS-based OM encoding method maintains a high degree of image quality and realizes the processing efficiency and robustness to be adapted in printing applications.

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

We developed a newhalftoning method, adaptive cell, which generates unstructured dots as by error diffusion method, but the produced binary image is less granular and more stable than that of error diffusion. The basic idea is covering the input image with cells which vary in size and shape depending on the input image data. Then appropriate number of pixels at the center of each cell are set on, to preserve the local average of the input image. The characteristics of the generated binary image shows that this method is suitable for high density electro photographic binary printers.

An analysis of the error diffusion method is presented which is based on the terminology of filter theory. The calculation of diffractive elements with the error diffusion method is shown. The stagnation of the iterative Fourier transform algorithm with a bandwidth constraint is overcome by error diffusion.

In this paper, we give a brief introduction of the parallel error diffusion and classify various approaches into three classes - partitioned area, concurrent, and inter-area error diffusions. These approaches, including intra-dot diffusion, space-filling curve traversal, Fibonacci-like sequence, multi-center dot diffusion, dispersed-dot diffusion, concurrent processing, neural network, and inter-dot diffusion are discussed and examples are given. Comparisons are made with other deterministic error filters (e.g. Floyd-Steinberg, Schroeder, Stucki, and Shiau-Fan) and with the corresponding clustered-dot and dispersed-dot ordered dithers. We also provide several extensions to the existing techniques. The conditions for parallel line error diffusions of horizontal, vertical, and diagonal lines are recognized and new parallel error diffusions are developed.