It may be difficult to locate keywords in noisy document images because of degraded OCR performance. A new technique for word image matching has the potential to select those. word images in a document that . represent potential keywords and to generate improved prototypes for those keywords. No explicit recognition is pe~formed in this process, but better OCR performance will occur on the impr...

This paper describes the performance of the MPEG-4 still texture image codec in coding noisy images. As will be shown, when using the MPEG-4 still texture image codec to compress a noisy image, increasing the compression rate does not necessarily imply reducing the peak-signal-to-noise ratio (PSNR) of the decoded image. An optimal operating point having the highest PSNR can be obtained within t...

Quantization of a noisy source is a classic problem. Although, optimal solutions have been shown to exist under certain assumptions, they are usually either extremely difficult to implement or computationally very intensive. The vector quantization (VQ) based strategy we present in this paper is simple, efficient, and capable of joint compression and denoising of images corrupted by additive wh...

This paper focuses on the iris recognition without the cooperation of the subject. The main objective of this analysis is to implement a system that can identify fake identification and deal with noisy images. Current iris recognition system lacks the capability to deal with noisy images and produces increased error rates. In these conditions, we propose a system that can overcome these barrier...

In this paper, we present the problem of noisy images recognition and in particular the stage of primitives selection in a classification process. We suppose that segmentation and statistical features extraction on documentary images are realized. We describe precisely the use of concept lattice and compare it with a decision tree in a recognition process. From the experimental results, it appe...

We present a new blur identification procedure based on a hybrid, two-level evolutionary-deterministic scheme which allows dynamic optimization of the size and the pattern of the support of the point spread function. We demonstrate that the proposed techniques can be generalized to the case of multi-band identification. Numerical simulations demonstrate the overall priority of the proposed iden...

The use of Wiener filters to restore noisy HREM images is discussed, exploiting the fact that the noise spectrum can be rather simply estimated. Conventional Wiener, Cannon, parametric versions of these and a random-phase form are found to be very effective. Quantitative analysis indicates that the signal-to-noise ratio is improved by a factor of 3-7, with better results for larger pictures. Su...