Gary E. Ford

Directional interpolation of images based on visual properties and rank order filtering

The goal of this research is to develop interpolation techniques which preserve or enhance the local structure critical to image quality. Preliminary results are presented which exploit either the properties of vision or the properties of the image in order to achieve the goals. Directional image interpolation is considered which is based on a local analysis of the spatial image structure. The extension of techniques for the design of linear filters based on properties of human perception reported previously to enhance the perceived quality of interpolated images is considered.<<ETX>>

Radiometric equalization of nonperiodic striping in satellite data

Abstract Images acquired by remote sensing contain radiometric errors caused by variations in the sensor response. In this note, we present a unified treatment of the correction of periodic or nonperiodic errors, which provides some insight into the relation of correction algorithms to the type of radiometric degradation. Successful correction of a NOAA Very High Resolution Radiometer (VHRR) thermal infrared image is demonstrated.

The position-orientation masking approach to parametric search for template matching

A new search method over (x,y,/spl theta/), called position-orientation masking is introduced. It is applied to vertices that are allowed to be separated into different bands of acuteness. Position-orientation masking yields exactly one /spl theta/ value for each (x,y) that it considers to be the location of a possible occurrence of an object. Detailed matching of edge segments is performed at only these candidate (x,y,/spl theta/) to determine if objects actually do occur there. Template matching is accelerated dramatically since the candidates comprise only a small fraction of all (x,y,/spl theta/). Position-orientation masking eliminates the need for exhaustive search when deriving the candidate (x,y,/spl theta/). Search is guided by correlations between template vertices and distance transforms of image vertices. When a poor correlation is encountered at a particular position and orientation, nearby positions at that orientation and nearby orientations at that position are masked out. Position and orientation traversal are by quadrant and binary decomposition. >

The evolution of mean curvature in image filtering

A new formulation for inhomogeneous image diffusion is presented in which the image is regarded as a surface in 3-space. The evolution of this surface under diffusion is analyzed by classical methods of differential geometry. A nonlinear filtering theory is introduced in which only the divergence of the direction of the surface gradient is averaged. This averaging preserves edges and lines, as their direction is non-divergent, while noise is averaged since it does not have non-divergent consistency. Our approach achieves this objective by evolving the surface at a speed proportional to mean curvature leading to the minimization of the surface area and the imposition of regularity everywhere. Furthermore, we introduce a new filter that renders corners, as well as edges, invariant to the diffusion process. Experiments demonstrating the adequacy of this new theory are presented.<<ETX>>

Nonlinear adaptive image filtering based on inhomogeneous diffusion and differential geometry

The inadequacy of the classic linear approach to edge detection and scale space filtering lies in the spatial averaging of the Laplacian. The Laplacian is the divergence of the gradient and thus is the divergence of both magnitude and direction. The divergence in magnitude characterizes edges and this divergence must not be averaged if the image structure is to be preserved. We introduce a new nonlinear filtering theory that only averages the divergence of direction. This averaging keeps edges and lines intact as their direction is nondivergent. Noise does not have this nondivergent consistency and its divergent direction is averaged. Higher order structures such as corners are singular points or inflection points in the divergence of direction and also are averaged. Corners are intersection points of edges of nondivergent direction (or smooth curves of small divergence in direction) and their averaging is limited. This approach provides a better compromise between noise removal and preservation of image structure. Experiments that verify and demonstrate the adequacy of this new theory are presented.

On mean curvature diffusion in nonlinear image filtering

Mean curvature diffusion is shown to be a position vector diffusion, tending to scalar diffusion as a flat image region is approached, and providing noise removal by steepest descent surface minimization. At edges, it switches to a nondiffusion state due to two factors: the Laplacian of position vanishes and the magnitude of the surface normal attains a local maximum.

Optimum and reduced complexity multiuser detectors for asynchronous CPM signaling

Maximum likelihood detector algorithms are developed for the matrix of transmitted symbols in a multiuser system in which the received signal is the sum of K cochannel continuous phase modulated (CPM) signals and additive white Gaussian noise. We illustrate that the maximum likelihood matrix detector, which provides optimum detector performance, consists of K sets of front-end matched filters followed by a Viterbi algorithm. We also derive two reduced complexity receivers, demonstrating through simulation that they perform within 1-2 dB of the optimal while substantially reducing complexity. The paper demonstrates how performance can be traded off against complexity, giving particular attention to cochannel Gaussian minimum shift keyed (GMSK) signals

Network model for invariant object recognition

Abstract A network model is proposed for invariant object recognition. The overall complexity of the weight connection is lower than that of several established networks. The performance of the proposed model is comparable to that obtained with Zernike moments.

Development of a Computer-Aided Age Determination System: Evaluation Based on Otoliths of Bank Rockfish off California

Abstract We have developed a computer-aided system (Bony Parts) to analyze periodic bands in fish otoliths (or other structures) for age estimation. The image analysis program first scans the image of a thin otolith section, perpendicular to the bands specified by the user. Adjacent scans are averaged and filtered with Fourier transformation or spatial domain convolution. Bands of higher density are detected and are marked and summed on the screen. We evaluated this new technique using subsamples of thin-sectioned otoliths from the bank rockfish Sebastes rufus. The time and effort for cleaning, preparation, sectioning, and mounting are the same for both traditional and computer-aided techniques. The computer-aided technique reduced the time and tedium of counting bands, yet still allowed the user to interactively make subjective decisions about aging criteria. Both approaches produced similar readings, but computer-aided estimates were more precise than traditional readings and required less analysis time...

Invariance of edges and corners under mean-curvature diffusions of images

We have recently proposed the use of geometry in image processing by representing an image as a surface in 3-space. The linear variations in intensity (edges) were shown to have a nondivergent surface normal. Exploiting this feature we introduced a nonlinear adaptive filter that only averages the divergence in the direction of the surface normal. This led to an inhomogeneous diffusion (ID) that averages the mean curvature of the surface, rendering edges invariant while removing noise. This mean curvature diffusion (MCD) when applied to an isolated edge imbedded in additive Gaussian noise results in complete noise removal and edge enhancement with the edge location left intact. In this paper we introduce a new filter that will render corners (two intersecting edges), as well as edges, invariant to the diffusion process. Because many edges in images are not isolated the corner model better represents the image than the edge model. For this reason, this new filtering technique, while encompassing MCD, also outperforms it when applied to images. Many applications will benefit from this geometrical interpretation of image processing, and those discussed in this paper include image noise removal, edge and/or corner detection and enhancement, and perceptually transparent coding.