This PDF file contains the editorial “Editorial: The Changing of the Guard” for JEI Vol. 3 Issue 04

The computational procedure introduced by Chen, Smith, and Fralick is in widespread use for computing the discrete cosine transform. It is popular because of its simplicity-it is a cascade of smaller (2x 2) orthogonal transforms. We show ways to further exploit the simplicities of such procedures and introduce a new family of transform (the generalized Chen transform or GCT). One parameterization of the GCT is essentially optimal for image compression but has a lower computational cost than other transforms.

Imaging bar-code scanners may use an edge-detection algorithm as the first step to locate a bar code in the image it acquires. This is also likely a step that consumes a significant amount of the total bar-code detection time. A simple but efficient special-purpose edge-detection algorithm is introduced that makes use of the special properties expected from a bar-code image.

The first step for human face recognition is to locate the head boundary in a head-and-shoulders image. An approach that uses adaptive contour models or "snakes" is described to solve this problem. Since we have a priori knowledge of the shape of a head, this active contour model is tailor-made for representing the head boundary. In this paper, a reliable method to locate the approximate position of the head and to estimate the head boundary is proposed. The effect of the parameters for snakes is investigated by locating the head boundary, and a best set of the parameters is suggested. A fast algorithm based on the greedy algorithm for active contour modeling is also presented. The computational complexity of this new algorithm is analyzed and compared with the greedy algorithm. This fast algorithm has a performance capability comparable to the greedy algorithm and reduces the execution time by more than 30% on the average.

A new approach to adaptive-neighborhood or region-based noise filtering is presented. The basic idea in this technique is to identify contextually related features in the image and to carry out statistical filtering operations using the pixels in these areas. Neighborhoods in the image are identified as sets of pixels that are 8-connected to the reference or seed pixel and are within a specified gray-level tolerance of the seed pixel. Thus, operations are based on contextual details in the image rather than an arbitrary grouping ofpixels (as in 3 x 3 filtering). These methods are applied to synthesized and natural images, and it is shown both quantitatively and qualitatively that adaptive-neighborhood filtering techniques are superior to analogous fixed-neighborhood filtering techniques.

A new technique is presented for the restoration of images degraded by a linear, shift-invariant blurring point-spread function in the presence of additive white Gaussian noise. The algorithm uses overlapping variable-size, variable-shape adaptive neighborhoods (ANs) to define stationary regions in the input image and obtains a spectral estimate of the noise in each AN region. This estimate is then used to obtain a spectral estimate of the original undegraded AN region, which is inverse Fourier transformed to obtain the space-domain deblurred AN region. The regions are then combined to form the final restored image. Mathematical derivation and implementation of the adaptive-neighborhood deblurring (AND) filter is discussed, and experimental results are presented with an analysis of the performance of the AND filter as compared to the fixed-neighborhood sectioned deblurring (FNSD) Wiener and power spectrum equalization filters. It is shown that using the AND algorithm for image deblurring enables the identification of relatively stationary regions. This improves the restoration process and produces results that are superior to those obtained using the FNSD method both visually and in terms of quantitative error measures.

Image restoration is an estimation process that attempts to recover an ideal high-quality image from a given degraded version. The Wiener filter method derived from the minimum mean square error criterion is widely used in image restoration to restore degraded images. In this method the constant Γ, which is an a priori representation of the signal-to-noise ratio for the complete image plane, is unknown and its value is supplied by the user and adjusted by the trial-and-error approach. A new estimation process of Γ is proposed. First of all, a second image is constructed from the given degraded image (referred to as the first image) using Lagrange's interpolation technique. Lagrange's interpolation technique used here is actually a modified version of the original approach. Secondly, an expression for Γ[u,v], the ratio of the power spectrum of the noise to the power spectrum of the first image is obtained using the power spectra of the first and the second images. However, the Wiener filter only needs Γ for the complete image plane. Therefore an arithmetic mean of a selected set of Γ[u,v] values is calculated. This arithmetic mean is then used as Γ in the Wiener filter to restore the first image.

Separation of characters and lines in color map images, especially when they are connected or overlapped, is a very challenging task in image analysis. We present a method to tackle this problem using robust line tracing, connected component analysis, and color clustering algorithms. Good results have been obtained using this method with real test images.

We propose a technique for the segmentation of color map images by means of an algorithm based on fuzzy clustering and prototype optimization. Its purpose is to facilitate the extraction of lines and characters from a wide variety of geographical map images. In this method, segmentation is considered to be a process of pixel classification. The fuzzy c-means clustering algorithm is applied to a number of training areas taken from a selection of different color map images. Prototypes, generated from the clustered pixels, that satisfy a set of validation criteria are then optimized using a neural network with supervised learning. The image is segmented using the optimized prototypes according to the nearest neighbor rule. The method has been verified to work efficiently with real geographical map data.

An adaptive Bayesian segmentation algorithm for color images is presented, which extends the adaptive clustering approach of Pappas to multichannel images. A scalar segmentation label field is generated for the multichannel data, which is modeled as a vector field, where the components of the vector field (each individual channel) are assumed to be conditionally independent given the segmentation labels. The class conditional probability model for the vector image field is taken as a multivariate Gaussian with a space-varying mean function. A Gibbs random field is employed as the a priori probability model for the segmentation label field that imposes a spatial connectivity constraint on the labels. The space-varying class means associated with the image segments can be used to form an estimate of the actual image from noisy observations. Experimental results are provided to demonstrate the benefits of using adaptivity via the space-varying means and the spatial connectivity constraint. We also discuss the effects of the color space within which the clustering is performed on resulting segmentations.

To achieve an ideal colorimetric calibration for CMYK (cyan, magenta, yellow, and black) printers, we propose a novel CMYK determination technique under the restrictions of a colorimetric match, use of the entire color gamut, and a smooth gradation of each primary color. The algorithm proposed here includes the following steps: (1) obtain the two extreme conditions for CMYK combinations to be able to utilize the entire CMYK gamut, namely maximum black technique and minimum black technique; (2) determine an initial black amount that lies between the two conditions; (3) apply a smoothing technique for the black amount; and (4) determine the remaining colors, CMY, for colorimetric match. An iterative smoothing technique in a uniform color space is introduced to obtain visually "smoothed" black gradations. The gradation quallty for each primary color is evaluated with a hypothetical unstable printer. The smoothed CMYK technique eliminates sudden changes for each primary color, so that a printer using this technique becomes robust against a change of characteristic curves of the printer such as dot gain. Also, it is found that a CMYK combination using the technique is suitable for a lookup table and interpolation technique in practical conversions.

This PDF file contains the editorial “Book Rvw: Stereo Computer Graphics and Other True 3D Technologies," edited by David F. McAllister for JEI Vol. 3 Issue 04