Catálogo de publicaciones - libros

Complex images with low signal to noise ratio (SNR) appear in various applications. To recover the associated phase images, noise effects, as loss of contrast and phase residues that can deteriorate the phase unwrapping process, should be reduced. There are various methods for noise filtering in complex images, however most of them deal only with the magnitude image. Only few works have been devoted to phase image de-noising, despite the existence of important applications like Interferometric Synthetic Aperture Radar (IFSAR), Current Density Imaging (CDI) and Magnetic Resonance Imaging (MRI). In this work, a group of de-noising algorithms in the wavelet domain were applied to the complex image, in order to recover the phase information. The algorithms were applied to simulated and phantom images contaminated by three different noise models, including mixtures of Gaussian and Impulsive noise. Significant improvements in SNR for low initial values (SNR<5 dB) were achieved by using the proposed filters, in comparison to other methods reported in the literature.

We describe a method for objective and quantitative evaluation of image quality. The method represents a novel use of image enhancement concepts. It employs three new measures that evaluate the definition of contours, uniform intensity distribution, and noise rate in determining the image quality. Because the three measures have clear physical meanings, they can be selectively applied according to the viewer’s evaluation criteria. The three measures are relatively inexpensive to compute, making them suitable for automated ranking of image quality in personal digital imaging devices, such as digital cameras. However, the method is equally adept at evaluating other digital images such as those on the Internet. Experiments with the method show good correlation with visual quality assessment for various image subject types.

A new system—Monoscale Dual Ridgelet Frame (MDRF) is constructed in this paper, which can be viewed as a generalized version of Monoscale Ridgelet introduced by Candes. The MDRT takes the Dual Ridgelet Frame as its basic component. We show that localizing the Dual Ridgelet Frame into small squares, dyadic partition of [0, 1], constitutes a dual frame in [0,1] again. As an example, the high performance of the MDRF for image denoising is demonstrated experimentally.

Real time video streaming applications over the internet pose several challenging problems due to the time-varying loss/delay network characteristics. Multiple description video coding is one method to alleviate the detrimental effects caused by these channel variations. However the selection of the source that is right for the channel having a desired cost is fundamental to the success of the multiple description technique. The proposed approach selects the set of descriptions which minimize an objective function for the streaming distortion.

Portable digital cameras have widespread recently. Their image quality, low cost and portability have drastically changed the culture of photography, today. Many professionals start to take photos of documents, instead of photocopying them. This paper presents an algorithm designed to remove the background of document images acquired through portable digital cameras.

X-ray digital tomosynthesis (DT) is very useful to PCB inspection because it can obtain a cross-sectional image of a local inspection area quickly. The image intensifier, which is usually used in DT, distorts X-ray images in shape and intensity. Therefore, image distortion correction is one of the most important issues in realizing DT system. In this paper, two image distortion correction methods for an X-ray DT system are presented and their performances are compared. The first method is to use a simplified distortion model by a distance ratio function in intensity correction, and by 2D point mapping polynomials in shape correction. The second method is to use a general polynomial distortion model. The experimental results show a great improvement of the second method in compensation speed and accuracy.

This paper proposes an efficient video watermarking scheme using adaptive threshold and minimum modification of motion vectors. Our proposed algorithm guarantees the amount of embedded watermark data and minimizes modification of original motion vectors to avoid degradation of video quality without reducing payloads. Besides, our algorithm can embed and retrieve watermark data without any increment of bit rate and original video contents. Experimental result shows that the proposed scheme obtains better video quality than other previous algorithms by about 0.5~1.1dB. Our scheme can be useful to real-time video watermarking which must be compatible to MPEG video coding.

This paper presents a novel lossless compression technique to transmit correlated images or data within sensor networks of inexpensive devices by exploiting the temporal correlation under the distributed source coding paradigm where the complexity of the encoder is much lower than that of the decoder. The technique operates in pixel-domain to avoid any lossy transform and relies on syndrome decoding of trellis codes by innovatively encoding the final state of the trellis. Experimental results on standard test video sequences proved superiority of this technique against the entropy based LZW lossless coding as well as a recently developed asymptotically lossless distributed source coding technique.

Symmetry information is used as the basis of a compression algorithm for images of decorative tiles yielding a compact representation. This allows faster network transmission and less space for storage of tile images. This paper presents an algorithm capable of automatically detecting the patterns of symmetry of images of tiles. The methodology developed may apply to any sort of repetitive symmetrical colour images and drawings.

In this paper, we propose a fast video mixing method for reducing the computational complexity in the MCU (Multipoint Control Unit) used in the video conferencing. The conventional mixing method is based on the pixel-domain transcoder, of which computational complexity is linearly increased as the number of participants is increased. Basically the method requires many decoders and one huge encoder to mix the multiple bitstreams. To reduce the computational complexity, we propose a hybrid mixing method based on the syntax-based bitstream modification and pixel-domain transcoder. The proposed method reduces the computational complexity about 45% at the improved quality, compared with the conventional mixing method based on the pixel-domain transcoder.