An original method of higher order curve detection using the straight-line Hough transform (HT) is presented. This scheme is possible because curves in an image can be approximated by small straight-line segments. By combining the straight-line HT mapping
scheme with an associative memory (AM), the method can discrimmate and identify between different curvies. Curves such as circles, ellipses, and parabolas have been studied. This method is computationally very efficient and may yield to hardware implementation, thus making it possible to use the HT in fast real-time applications. The system consists of a parallel analog mapping from image plane to parameter plane and an AM pattern classifier.
Intravascular ultrasound imaging is a new technique that displays information on lumen and arterial walls, and is capable of providing real-time monitoring of cross-sectional high-resolution images. This technique has potential application for studying the dynamics of the arterial wall with respect to the presence or absence of pathology and the vascular response to physiological or pharmacological stimuli. Although the extraction of information related to coronary dynamics and wall pathologies is possible by manual procedures, it is very time consuming and influenced by intra- and interobserver errors. We developed an evaluation system for analyzing 3-D spaces defined by digitized cross-sectional ultrasound images of coronaries quantifying the vasomotion in relation to the morphology of the arterial wall. Sequences of echographic images were obtained and recorded as ordered stacks of 2-D frames on a VHS videotape. For each image, an automatic lumen edge segmentation was performed, then 3-D reconstruction was obtained to evaluate time-dependent lumen and vessel wall changes. These 3-D representations serve to demonstrate dynamic phenomena and to perform quantitative analyses (e.g., area/hemidiameter variations, projections, sections, "carving," etc.).
Currently, many low-cost computers can only simultaneously display a palette of 256 colors. However, this palette is usually selectable from a very large gamut of available colors. For many applications, this limited palette size imposes a significant constraint on the achievable image quality. We propose a method for designing an optimized universal color palette for use with halftoning methods such as error diffusion. The advantage of a universal color palette is that it is fixed and therefore allows multiple images to be displayed simultaneously. To design the palette, we employ a new vector quantization method known as sequential scalar quantization (SSQ) to allocate the colors in a visually uniform color space. The SSQ method achieves near-optimal allocation, but may be efficiently implemented using a series of lookup tables. When used with error diffusion, SSQ adds little computational overhead and may be used
to minimize the visual error in an opponent color coordinate system. We compare the performance of the optimized algorithm to standard error diffusion by evaluating a visually weighted mean-squared-error measure. Our metric is based on the color difference in CIE L*a*b*, but also accounts for the lowpass characteristic of human contrast sensitivity.
The rate-distortion trade-off in the discrete cosine transform- based coding scheme in ISO/JPEG is determined by the quantization table. To permit a different quality to be selected by a user, a common practice is to scale the standard quantization tables that have been empirically determined from psychovisual experiments. In this paper, an algorithm is presented to generate a quantization table that is optimized for a given image and for a given distortion. The computational complexity of this algorithm is reduced compared to other techniques. The optimized, image-adaptive quantization table typically yields an improvement of 15% to 20% in bit rate compared to the use of standard, scaled quantization tables. Once an optimized quantization table has been generated for a specific image, it can also be applied to other images with similar content with a small sacrifice in bit rate.
A new transform coder based on the zonal sampling strategy, which outperforms the JPEG baseline coder with comparable computational complexity, is presented. The primary transform used is the 8- × 8-pixel-block discrete cosine transform, although it can be replaced by other transforms, such as the lapped orthogonal transform, without any change in the algorithm. This coder is originally based on the Chen-Smith coder, therefore, we call it an improved Chen-Smith (ICS) coder. However, because many new features were incorporated in this improved version, it largely outperforms its predecessor. Key approaches in the ICS coder, such as a new quantizer design, arithmetic coders, noninteger bit-rate allocation, decimated variance maps, distance-based block classification, and human visual sensitivity weighting, are essential for its high performance. Image compression programs were developed and applied
to several test images. The results show that the ICS performs substantially better than the JPEG coder.
We propose an efficient algorithm that recognizes handwritten Korean and English characters in a low-resolution document. For a user-friendly input system for low-resolution documents consisting of two different sets of characters obtained by a facsimile or scanner, we propose a document-recognition algorithm utilizing several effective features (partial projection, the number of cross points, and distance features) and the membership function of the fuzzy set theory. Via computer simulation with several test documents, we show the effectiveness of the proposed recognition algorithm for both printed and handwritten and Korean and English characters in a low-resolution document.
A new edge-enhanced error diffusion algorithm, based on Eschbach's algorithm, is proposed. Thick-edged artifacts as well as small edge-enhancement effects for the bright or dark pixel values are observed in the previous algorithm. By analyzing the phenomena, a new improved algorithm is proposed by using the diffused error sum and input pixel value. An input pixel is classified into a normal- or edge-region pixel based on the error sum criterion. A new error calculation is then employed for the edge region pixel, while conventional error calculation is used for the normal-region pixel. The proposed method requires only a few additional calculations and provides edge-enhanced binary output images. The edges are influenced less by the brightness offset, and thick-edged artifacts are reduced.
Iterative halftoning algorithms offer great flexibility in adapting the halftoning process to specific demands. Constraints defined in the Fourier domain can be used to synthesize images with a wide variety of characteristics. Using such constraints, the halftoning process and the resulting image can be adapted to the characteristics of a processing system or the graytone original. Moreover, a control of image texture can be realized and combined with other constraints.
The Gibbs random field (GRF) has proved to be a simple and practical way of parameterizing the Markov random field, which has been widely used to model an image or image-related process in many image processing applications. In particular, the GRF can be employed to construct an efficient Bayesian estimation that often yields optimal results. We describe how the GRF can be efficiently incorporated into optimization processes in several representative applications, ranging from image segmentation to image enhancement. One example is the segmentation of computerized tomography (CT) volumetric image sequence in which the GRF has been incorporated into K-means clustering to enforce the neighborhood constraints. Another example is the artifact removal in discrete cosine transform-based low bit rate image compression where GRF has been used to design an enhancement algorithm that reduces the "blocking effect" and the "ringing effect" while still preserving the image details. The third example is the integration of GRF in a wavelet-based subband video coding scheme in which the high-frequency subbands are segmented and quantized with spatial constraints specified by a GRF, and the subsequent enhancement of the decompressed images is accomplished by smoothing with another type of GRF. With these diverse examples, we are able to demonstrate that various features of images can all be properly characterized by a GRF. The specific form of the GRF can be selected according to the characteristics of an individual application. We believe that the GRF is a powerful tool to exploit the spatial dependency in various images, and is applicable to many image processing tasks.
Image segmentation is one of the most important steps in computerized systems for analyzing geographic map images. We present a segmentation technique, based on fuzzy rules derived from the K-means clusters, that is aimed at achieving human-like performance. In this technique, the K-means clustering algorithm is first used to obtain mixed-class clusters of training examples, whose centers and variances are then used to determine membership functions. Based on the derived membership functions, fuzzy rules are learned from the K-means cluster centers. In the map image segmentation, we make use of three features, difference intensity, standard deviation, and a measure of the local contrast, to classify each pixel to the foreground, which consists of character and fine patterns, and to the background. A centroid defuzzification algorithm is adopted in the classification step. Experimental results on a database of 22 gray-scale map images show that the technique achieves good and reliable results, and is compared favorably with an adaptive thresholding method. By using K-means clustering, we can build a segmentation system of fewer rules that achieves a segmentation quality similar to that of using the uniformly distributed triangular membership functions with the fuzzy rules learned from all the training examples.