Showing papers on "Histogram equalization published in 1994"

Efficient color histogram indexing

Abstract: In image retrieval based on color, the weighted distance between color histograms of two images, represented as a quadratic form, may be defined as a match measure However, this distance measure is computationally expensive (naively O(N/sup 2/) and at best O(N) in the number N of histogram bins) and it operates on high dimensional features (O(N)) We propose the use of low-dimensional, simple to compute distance measures between the color distributions, and show that these are lower bounds on the histogram distance measure Results on color histogram matching in large image databases show that pre-filtering with the simpler distance measures leads to significantly less time complexity because the quadratic histogram distance is now computed on a smaller set of images The low-dimensional distance measure can also be used for indexing into the database >

System and method for diagnosis of living tissue diseases using digital image processing

Abstract: A system and method for diagnosis of living tissue diseases is described. The system includes a computer device for controlling its operation. An operator control device is coupled to the computer device. A viewing screen is coupled to the computer device for displaying digitized images of the living tissue. The operator, using the control device, selects desired portions of the digitized image for further image enhancement according to a desired image enhancement feature selectable from a plurality of image enhancement features. The image enhancement features include any combination of grey scale stretching, contrast enhancement based on logarithmic histogram equalization, spot enhancement and magnification. The system further includes means for visualization and quantification of micro-calcifications, and means for visualization and quantification of mass spiculations.

Cascaded image processing using histogram prediction

Abstract: A method of improving processing of a color natural scene image defined by color density signals processed by multiple processing elements, wherein after an exposure processor incorporates a luminance signal histogram, a modifying TRC curve is derived and applied to the luminance histogram signal. The resultant modified luminance histogram signal is blurred prior to directing it to a contrast processor. The modifying TRC curves determined at each processor are combined and applied to the color density signals only after combination.

Image line-segment extracting apparatus

Abstract: An image line-segment extracting apparatus which in processing steps to conduct Hough transformation of a sequence of edge pixels composing an image, to plot a histogram based on Hough function values, to successively detect peaks from the histogram and to extract a plurality of line segments existing in the image, can detect the only true peaks, effectively eliminating false peaks The process also includes a feature by which region is defined and the amount of influence of a remarkable one of the peaks in a histogram upon frequency distribution on the histogram are determined and then histogram frequency values of other peaks in the defined region are corrected according to the determined amount of influence to partially rewrite the histogram

Image extraction method and apparatus, and image recognition method and apparatus for extracting/recognizing specific images from input image signals

Abstract: A circumscribing rectangle is obtained for a black continuous image part using a two-tone image signal. If it is determined that the image part is possibly one which is obtained as a result of rotating the object, lengths of sides of the image part are compared with threshold values. If it is determined that the image part is not one which is obtained as a result of rotating the object, the height and horizontal length of the circumscribing rectangle are compared with the threshold values. Thus, it is determined whether or not the image part is identical to the object. An RGB chromaticity histogram is produced for each of small regions of an input color image. Each of the chromaticity histograms of the small regions is compared with reference ones. As a result of the comparison, an identification number of the reference histogram having the highest similarity to that of the small region among those constituting the input image is given to the small region. A histogram of the identification numbers thus is produced for the input color image. The thus-produced histogram is used to determine which one of a plurality of objects is identical to the input color image.

Color-image quantization with use of a fast binary splitting technique

Abstract: We investigate an efficient color-image quantization technique that is based on an existing binary splitting algorithm [ IEEE Trans. Signal Process. 39, 2677 ( 1991)]. The algorithm sequentially splits the color space into polytopal regions and picks a palette color from each region. As originally proposed, the complexity of this algorithm is a function of the image size. We introduce a fast histogramming step so that the complexity will depend only on the number of distinct image colors. Data structures are employed that permit the storage of a full-color histogram at moderate memory cost. In addition, we apply a prequantization step that reduces the number of initial image colors while preserving image quality along visually important color coordinates. Finally, we incorporate a spatial-activity measure to reflect the increased sensitivity of the human observer to quantization errors in smooth image regions. This technique preserves the quantitative and qualitative performance of the original binary splitting algorithm while considerably reducing the computation time.

A robust morphological algorithm for automatic radiation field extraction and correlation of portal images.

Abstract: Accurate extraction of the radiation field boundary from portal images is required in selective histogram equalization techniques for contrast enhancement of portal images, and is required in portal‐simulator image correlation. Local edge detectors are so noise sensitive that delineation of the field from anatomy edges and noise is very difficult. Optimization in edge localization and noise suppression can be achieved with optimal edge detectors, but the required calculation is very time consuming. A portal‐image segmentation algorithm based on morphological edge detection and morphological segmentation techniques is presented. Relying on two predefined criteria that are nonsensitive to variation of portal‐image types, the algorithm can automatically search the optimal threshold value, which is sensitive to the variation of the type and quality of portal images. With two stages of the searching procedure, the algorithm can accommodate a large variation of single and double exposure portal images obtained from different therapy machines. Results of our morphological edge detector are also compared to that of an optimal edge detector. Portal‐simulator image correlation using the inertia moments of the radiation field mask is investigated, and compared to correlation obtained using inertia moments of the radiation field contour. It is shown that the mask method is less sensitive to small variations and distortions in the field shape, resulting in a more accurate correlation. This would substantially simplify the task of treatment verifications.

Bounds for the discrimination power of color indexing techniques

Abstract: Most color indexing techniques proposed in the literature are similar: images are represented by color histograms, and a metric on the color histogram space is used to determine the similarity of images. In this paper we determine the limits of these color indexing techniques. We propose two functions to measure the discrimination power of indexing techniques: the capacity (how many distinguishable histograms can be stored) and the maximal match number (the maximal number of retrieved images). We derive bounds for these functions. These bounds have two practical aspects. First, they help a user to decide whether color histograms effectively index database images from a given domain. Second, they facilitate the choice of a good threshold for the distance below which histograms are considered similar. Our arguments are based on an analysis of the metrical properties of the histogram space and results from coding theory. The results show that over a large range of reasonable parameters the capacity is very large. Thus, the set of parameters for which color indexing works well can be described as the set of parameters for which the maximal match number is below an application-dependent maximum.

Method for estimating scene parameters from color histograms

Abstract: One of the key tools in applying physics-based models to machine vision has been the analysis of color In the mid-1980’s it was recognized that the color histogram for a single inhomogeneous surface histograms. with highlights will have a planar distribution in color space. It has since been shown that the colors do not in a plane but form clusters at specific points. Physics-based models of reflection predict that fall randomly shape of the histogram is related not only to the illumination color and the object color but also to such the properties as surface roughness and imaging geometry. We present an algorithm for analyzing color noncolor histograms that yields estimates of surface roughness, phase angle between the camera and the light source, and illumination intensity. These three scene parameters are related to three histogram measurements. However, the relationship is complex and cannot be solved analytically. Therefore we developed a method for estimating these properties that is based on interpolation between histograms that come from images of known scene properties. We present tests of our algorithm on simulated data, and the results compare well with the known simulation parameters. We also test our method on real images, and the results compare favorably with the actual parameters estimated by other means. Our method for estimating scene properties is very fast and requires only a single color image.

Application of adaptive histogram equalization to x-ray chest images

Abstract: This paper uses adaptive histogram equalization to enhance the contrast of X-ray chest image. Several methods are presented to eliminate the noise caused by the enhancement. The mainly used methods are frame averaging, bandstop zero-phase filtering based on 2D FFT. The cause of noise is also analyzed.

The Capacity and the Sensitivity of Color Histogram Indexing

Abstract: Color histogram matching has been shown to be a promising way of quickly indexing into a large image database. Yet, few experiments have been done to test the method on truly large databases, and even if they were performed, they would give little guidance to a user wondering if the technique would be useful with his or her database. In this paper we de ne and analyze measures relevant to extending color histogram indexing to large databases: capacity (how many distinguishable histograms can be stored) and sensitivity (how the average number of retrieved images depends on the retrieval threshold). The theoretical results lead to a practical test procedure which enables a user to determine the performance of color histogram indexing on a large database by looking at a small, randomly-chosen subset of the images. We suggest how our analysis can be extended to other feature-based indexing techniques. Capacity and Sensitivity of Color Histogram Indexing 1

Using Illumination Invariant Color Histogram Descriptors for Recognit ion

Abstract: Color pixel distributions provide a useful cue for object recognition. Recently, for example, a technique called color indexing due to Swain and Ballard [6] used color histograms for the efficient recognition of objects from a large database in the presence of changes in scene geometry and occlusion. The effectiveness of this and other approaches that match color distributions, however, depends on the approximate constancy of the scene illumination. In this paper, we develop color histogram descriptors that are invariant to changes in the intensity and spectral distribution of the illumination. We present a set of experiments that dernonstrate the effectiveness of these descriptors for object recognition in the presence of changes in illuminant spectral power distribution.

A system and method for processing images of living tissue

Abstract: A system and method for processing images of living tissue diseases is described. The system includes a computer device for controlling its operation. An operator control device is coupled to the computer device. A viewing screen is coupled to the computer device for displaying digitized images of the living tissue. The operator, using the control device, selects desired portions of the digitized image for further image enhancement according to a desired image enhancement feature selectable from a plurality of image enhancement features. The image enhancement features include any combination of grey scale stretching, contrast enhancement based on logarithmic histogram equalization, spot enhancement and magnification. The system further includes means for visualization and quantification of micro-calcifications, and means for visualization and quantification of mass spiculations.

Image processing suitable for changing the number of colors in an image

Abstract: An image processing apparatus converts color image data into hue, saturation and luminosity data. When a specific color portion and a range are designated in the hue data, processing for reducing tones is executed based on the designated data. Thereafter, a reverse-conversion into color image data is performed, thereby making it possible to output a high-quality image of an illustration. Further, a histogram of designated hue data is generated, processing for reducing tones is executed based on the histogram, and the data is subsequently reverse-converted into color image data. This makes it possible to output a higher quality image of an illustration.

Subband coding of color images with limited palette size

Abstract: An approach for compression of color images using a sorted color palette and a subband source coder is presented. When the decoded image is to be displayed on an eight bit monitor this method significantly reduces the decoder computational complexity. To avoid requantization before displaying the decoded image, the encoding and decoding are performed on a pseudo greyscale image which represents the color space defined by a sorted palette. To sort the palette, a measure for color similarity is needed. Here an l/sub 2/ distance in the RGB and luminance-chrominance color space has been studied, and a new color correlation measure is proposed. For comparison a transform source coder (JPEG) is used to encode the pseudo greyscale image. Simulation results show that the subband source coder gives better results than the transform source coder. The quality of the decoded color images is surprisingly good. >

Image synthesizing system and video game apparatus using the same

Abstract: An image synthesizing system is formed to perspectively transform and project a 3-D object in a virtual 3-D space represented by a combination of polygons onto a projection plane in a view-point coordinate system and to calculate color specifying data for each pixel to display an image on a display 18. The image synthesizing system has a depth information transforming unit 30 for transforming a depth information Z in the view-point coordinate system of each polygon into a depth information Z' for the depth cueing, a color palette 28 responsive to the color specifying data for each pixel for outputting a color signal as a front-color signal, a depth cueing information setting unit 26 for calculating a depth cueing back-color signal 320 depending on the displayed scene, and a color signal calculation unit 36 responsive to the depth information Z' for the depth cueing, front-color signal and back-color signal for forming and outputting a depth-cued color signal 350 for each pixel.

Image processing apparatus and method in which multi-level color signal with removed edge component is encoded

Abstract: An image processing apparatus in which a multi-level color signal is input and converted into a binary color signal. An edge component is removed from the multi-level color signal based on the binary color signal and the binary color signal and the multi-level color signal of which the edge component is removed are encoded.

Effects of luminance quantization error on color image processing

Abstract: A common approach to color image processing is to apply monochrome techniques to the quantized intensity component. However, the quantization error in the intermediate intensity image propagates to the final processed image. This can lead to significant distortion, depending on the input RGB values and on the particular image processing function being applied. A theoretical analysis of the worst-case quantization error is presented. Experimental results with histogram equalization demonstrate how the histogram resolution affects the performance. >

Image thresholding by histogram transformation

Abstract: A method for threshold selecting based on histogram transformation is proposed. New formulae for histogram transformation are introduced in the paper. After doing histogram transformation, the histogram became smooth. So, the optimal threshold is easy to detect. It is computationally fast and efficient. Experimental results for a set of images are presented.© (1994) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

Histogram equalization of the saturation component for true-color images using the C-Y color space

Abstract: Histogram equalization is a well known tool for enhancing the contrast and brightness of grayscale images. Grayscale histogram equalization has been extended to color images with limited success. One common method is to equalize the illumination component, while leaving the saturation and hue components unchanged. This method doesn't improve the overall color saturation of the image. Another approach is to apply equalization techniques in the RGB color space. The difficulty in using the RGB color space is that it does not correspond to human interpretation of color. This paper describes a method for histogram equalization of the saturation component using the color difference (or C-Y) color space. Since equalization of the saturation component alone leads to color artifacts, attention is given to the relationship that exists between saturation and intensity.

Explosion of multidimensional image histograms

Abstract: Enhancement of multispectral and true-color images is frequently accomplished by histogram modification. The color components are generally separately adjusted within an appropriate, often perceptually based, color coordinate system. However, independent component modification is seldom able to exploit the full extent of the RGB gamut unless some color values are clipped at the RGB boundaries. Further improvement in color contrast is sometimes more important than preserving perceptual attributes. For example, color composites derived from multispectral images have little basis in human perception and thus need not be perceptually constrained. "Histogram explosion", a new concept for multivariate enhancement, enables the full RGB extent to be exploited without clipping. While the algorithm is not generally based upon a perceptual model, original hue values may be preserved with proper parameter choices. The paper presents details of the histogram explosion method, an analysis of its computational complexity, and some experimental results. >

Transmission of the color information using quad-trees and segmentation-based approaches for the compression of color images with limited palette

Abstract: The compression of color images is usually performed independently along each of the 3 axis of a luminance-chrominance color space. When applied to images using a limited color palette, it generates images which take on more values than those found in the original color palette. These images must be color quantized before they can be displayed with a limited palette. In this paper, we present two new approaches for the lossy compression of color quantized images that does not require color quantization of the decoded images. The algorithms restrict the pixels of the decoded image to take values only in the original color palette. The first algorithm does so by using lists of colors taken by pixels in variable block sizes. The second one uses a color segmentation of the image. These two approaches improve the performance of previously proposed algorithms. For comparable quality and similar bit rate, the proposed approaches have lower decoding complexity than standard DCT-based coding algorithms. >

Color palette restoration

Abstract: When designing hardware, it is often desirable to represent images as economically as possible. Due to this, algorithms have been developed to create reduced palette images. Much better viewing results can be obtained by first reconstructing a full color image from the reduced palette image. This creates a need for a palette restoration algorithm. This paper develops an algorithm to reconstruct high resolution color image data from reduced color palette images. The algorithm is based on stochastic regularization using a non-Gaussian Markov random field model for the image data. This results in a constrained optimization algorithm that is solved using an iterative constrained gradient descent computational algorithm. During each iteration the potential update must be projected onto the constraint space. In this paper a projection operator that maps a vector onto a quantized constraint space is developed. Results of the proposed palette restoration algorithm have indicated that it is effective for the reconstruction of palettized images. Quantitative as well as visual results of the experiments are presented.© (1994) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

Address predictive color quantization image compression for multimedia applications

Abstract: Most modem workstations and personal computers are using 8-bit color mapped graphics architecture for displaying color images. To display a decompressed color image of the conventional DCT-based compression algorithms on these palette-based display systems, the decoded image must first be color quantized. To avoid the computational intensive color quantization process and provide a fast decoding process, new address predictive color quantization image compression schemes for multimedia applications are proposed in this paper. A closest pairs color palette ordering technique is also proposed for effectively exploited the redundancy of the palettized image. Computer simulation results are given to show the performance of the new compression algorithms in terms of bit rate and signal to noise ratio. >

Preprocessing of training set for backpropagation algorithm: histogram equalization

Abstract: This paper introduces a data preprocessing algorithm that can improve the efficiency of the standard backpropagation (BP) algorithm. The basic approach is transforming input data to a range that associates high-slopes of sigmoid where relatively large modification of weights occurs. This helps escaping of early trapping from prematured saturation. However, a simple and uniform transformation to such desired range can lead to a slow learning if the data have a heavily skewed distribution. In order to improve the performance of BP algorithm on such distribution, the authors propose a modified histogram equalization technique which enhances the spacing between data points in the heavily concentrated regions of skewed distribution. The authors' simulation study shows that this modified histogram equalization can significantly speed up the BP training as well as improving the generalization capability of the trained network. >

Method and apparatus for extracting and recognizing a specific image from an imput image signal

Abstract: A circumscribing rectangle is obtained for a black continuous image part using a two-tone image signal. If it is determined (in S105) that the image part is possibly one which is obtained as a result of rotating the object, lengths of sides of the image part are compared with threshold values (in S106 and S108). If it is determined (in S105) that the image part is not one which is obtained as a result of rotating the object, the height and horizontal length of the circumscribing rectangle are compared with the threshold values (in S107 and S108). Thus, it is determined (in S115 and S116) whether or not the image part is identical to the object. An RGB chromaticity histogram is produced for each of small regions of an input color image (in S202A). Each of the chromaticity histograms of the small regions is compared with reference ones (in S202A). As a result of the comparison, an identification number of the reference histogram having the highest similarity to that of the small region among those constituting the input image is given to the small region (in S204). A histogram of the identification numbers thus is produced for the input color image (in S204). The thus-produced histogram is used to determine (in S205) which one of a plurality of objects is identical to the input color image.

Adaptive Energy Function for Active Net.

Abstract: This paper proposes an algorithm to extract a target region from a color image. The algorithm utilizes the active net for stable target extraction. In order to extend the original active net for color image analysis, the image energy function of the net is defined by the function, called the Ratio Histogram, of the histogram backprojection method. Since ratio histogram is computed from both color histograms of a target and an input image, the adaptive image energy function for the input image can be created. Then, since our proposed method uses target colors as the index of the target, the net can extract a target regardless of the target size and of its pose. Additionally, we refine the proposed image energy dynamically during iterations of the net to improve the net convergency. Experimental results for color images taken by a video camera show the validity of our approaches.