Showing papers on "Histogram equalization published in 1992"

Color image enhancement through 3-D histogram equalization

Abstract: The method of histogram equalization for grey-level image enhancement is extended to color images in the paper. A method of direct 3-D histogram equalization is proposed which results in a uniform histogram of the RGB values. Due to the correlation between the bands, the principle on which grey-level (1-D) histogram equalization is based is not valid in the case of color images (3-D). This problem is alleviated by employing a histogram specification method where a uniform histogram is specified. >

Anatomy of a color histogram

Abstract: It is shown that the color histogram has an even closer relationship to scene properties than has been previously described. Color histograms have identifiable features that relate in a precise mathematical way to scene properties. Object color and illumination color are the most obvious properties that are related to color distribution, and their extraction has already been described. It is shown here that the histogram of color variation may be further exploited to relate its shape to surface roughness and imaging geometry. An understanding of these features allows an improved estimate of illumination color and object color to be made. >

Color segmentation using perceptual attributes

Abstract: A general approach for achieving color image segmentation using uniform-chromaticity-scale perceptual color attributes is proposed. At first chromatic and achromatic areas in a perceptual IHS color space are defined. Then the image is separated into chromatic and achromatic regions according to the region locations in the color space. 1-D histogram thresholding for each color attribute is performed to split the chromatic and achromatic regions, respectively. Finally the region growing is used to solve the oversegmentation problem. In an experiment the power of the proposed approach is demonstrated. >

Method and apparatus for automatic tonescale generation in digital radiographic images

Abstract: A method and apparatus for automatically and adaptively generating tonescale transformation functions that are robust with respect to imaging systems, exposure conditions, and body parts. The technique uses the histogram of the digital input image, the cumulative distribution function of the histogram, and the entropy of subsections of the histogram to create the final tonescale transformation. Using these three functions, the histogram can be divided into a region of interest, a low-signal foreground region, and a high-signal background region. The tonescale is constructed to be substantially linear over the region of interest, joining smoothly with a nonlinear portion extending from the end of the low-signal foreground region to the start of the region of interest, and another nonlinear portion of the high-signal background region. The substantially linear region of interest and the entire image are subject to certain output density constraints to optimize the diagnostic utility of the final image. The technique depends entirely on the histogram of the input image, and, hence, it is adaptive and robust.

Sequential product code quantization of digital color image

Abstract: A digital color image quantization mechanism employs sequential product code vector quantization, to sequentially extract chrominance and luminance values from the vectors and quantizes chrominance and luminance features based upon a conditional distribution of these features within partitioned regions of chrominance/luminance color space. The mechanism sequentially partitions a histogram of the original digital color image in luminance, chrominance (Y,Cb,Cr) space coordinates into a plurality of sub-regions or color space cells, such that each partitioned color cell is associated with a color of the output palette through which the color composition of a reproduced color image is defined. A splitting criterion determines the sequential order of partitioning of an axis. Because of the increased sensitivity of the human visual system to contouring artifacts in regions of an image to low spatial activity, the splitting criterion along the luminance axis is scaled or weighted in inverse proportion to the average spatial activity of the luminance-chrominance region subject to be split. A map of chrominance and luminance output codes is generated for the respective pixels of the output color image in accordance with the axial splitting or quantization of the chrominance and luminance components of the histogram.

Automated tone correction apparatus and method using filtered histogram equalization

Abstract: An image processing system redistributes image luminance values in an image represented in u'v'L* space to achieve a near-optimal fit of luminance data to available dynamic range and to enhance the shape and detail of the resultant image Redistribution may be accomplished by passing the image through a filter having a peak response at a spatial frequency at which the human eye is highly sensitive, and then equalizing a cumulative histogram of filtered pixels An image may be transformed to compensate for tone compression which accompanies printing processes

Gradation compensation apparatus of video image

Abstract: A level of an input luminance signal for a pixel is compared with step values which change stepwise in level in a predetermined range of level, and distribution of the levels of the input luminance signal in the predetermined range of level is represented by a cumulative histogram. Subsequently, a histogram value is derived from the outline of the cumulative histogram, and a compensation value is produced on the basis of the histogram value, and the compensation value is added to the input luminance signal, and hence a compensated luminance signal is output to control the luminance of a video image.

Combining vector quantization and histogram equalization

Abstract: Combined vector quantization and adaptive histogram equalization Pamela C. Cosman Eve A. Riskin Robert M. Gray tDurand Building, Department of Electrical Engineering Stanford University, Stanford, CA, 94305-4055 Department of Electrical Engineering, FT- 10 University of Washington, Seattle, WA 98195 ABSTRACT Adaptive histogram equalization is a contrast enhancement technique in which each pixel is remapped to an intensity proportional to its rank among surrounding pixels in a selected neighborhood. We present work in which adaptive histogram equalization is performed on the codebook of a tree-structured vector quantizer so that encoding with the resulting codebook performs both compression and contrast enhancement. The algorithm was tested on magnetic resonance brain scans from different subjects and the resulting images were significantly contrast enhanced. 1. INTRODUCTION Histogram equalization refers to a set of contrast enhancement techniques which attempt to spread out the intensity levels occurring in an image over the full available range.1 Histogram equalization is a competitor of interactive intensity windowing, which is the established contrast enhancement technique for medical images. In global histogram equalization, one calculates the intensity histogram for the entire image and then remaps each pixel's intensity proportional to its rank among all the pixel intensities. In adaptive histogram equalization (AHE), the histogram is calculated only for pixels in a context region, usually a square, and the remapping is done for the center pixel of the square. This can be called pointwise histogram equalization because, for each point in the image, one calculates the histogram for the square context region centered on that point. Because this is very computationally intensive, the bilinear interpolative version is an alternative that lowers the computational complexity.2 It calculates the histogram for only a set of non-overlapping context regions that cover the image and the reniapping of pixel intensity values is then exact for only the small number of pixels that are at the centers of these context regions. For all other pixels, a bilinear interpolation from the nearest context region centers determines the appropriate remapping function. With the bilinear interpolative version of AHE, the remapping function for a given pixel of intensity i at location (, y) is determined from the nearest 4 context regions as shown in figure 1. Ifm+_ denotes the mapping at the grid pixel (x+, y.) to the upper right of (x, y), and similar subscripts are used for the other surrounding context regions, then the interpolated AHE result is given by2: in(i) = a[bm(i) + (1 — b)m_(i)J + [1 — u]{bm_(i) + (1 — b)m__(i)], b= here y+—y- O-8194-0805-O/92/$4.QO SPIE Vol. 1653 Image Capture, Formatting, and Display (1992) / 213 Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/20/2014 Terms of Use: http://spiedl.org/terms

Method for constructing a histogram in digital image processing by statistical pixel sampling

Abstract: A method of creating a histogram of an image signal matrix, representative of a radiographic image obtained by scanning a stimulable phosphor sheet carrying a radiation image with stimulating rays, detecting the light emitted after stimulation and converting the detected light into electrical image signals, is described wherein the histogram is created by capturing only a fraction of the image pixels comprised in the image signal matrix, this fraction being determined on the basis of statical pixel sampling.

Automatic contrast adjustment for detail recognition in SEM images using on‐line digital image processing

Abstract: A new digital filter method has been developed for enhancement of detail recognition in SEM images of high signal-to-noise ratio using an on-line digital image processing system. The filter allows an automated improvement of the presentation of SEM image information and utilizes a median filter of very large mask size and histogram equalization. Since the method can be performed without input of any processing parameters, the user simply pushes a button for obtaining the processing result similar to conventional photo recording. The method utilizes digital signal processors for establishing high speed, hence, the processing results can be immediately assessed. When Applied to a variety of field-emission SEM images, there were no problems with inconvenient artifact encountered.

New results in color image quantization

Abstract: We investigate an efficient color image quantization technique that is based upon an existing binary splitting algorithm. The algorithm sequentially splits the color space into polytopal regions and picks a palette color from each region. At each step, the region with the largest squared error is split along the direction of maximum color variation. The complexity of this algorithm is a function of the image size. We introduce a fast histogramming step so that the algorithm complexity will depend only on the number of distinct image colors, which is typically much smaller than the image size. To keep a full histogram at moderate memory cost, we use direct indexing to store two of the color coordinates while employing binary search to store the third coordinate. In addition, we apply a prequantization step to further reduce the number of initial image colors. In order to account for the high sensitivity of the human observer to quantization errors in smooth image regions, we introduce a spatial activity measure to weight the splitting criterion. High image quality is maintained with this technique, while the computation time is less than half of that of the original binary splitting algorithm.

Estimating scene properties by analyzing color histograms with physics-based models

Abstract: The goal of machine vision is to allow intelligent systems to describe the world around them by the interpretation of images. The difficulty is that vision is a very complex process, since images may contain shadows, highlights, interreflections, and other phenomena. Images are created through the interaction of light with the world; therefore, any vision system that is to understand images must have a model of those interactions. By using physics-based models to describe image formation, we can analyze images in a systematic way. In applying physical models to machine vision, one of the key tools has been color histogram analysis. A color histogram shows the variation of colors observed within the scene. Analysis of color histograms may be extended to images that contain interreflection. Applying the dichromatic reflection model to interreflection between two objects shows how color, imaging geometry, and surface roughness affect the color histogram. By measuring the dimensions of the resulting clusters, additional information about scene characteristics may be obtained that is not available in simpler scenes. This thesis describes an algorithm for analyzing color histograms that yields estimates of the properties of object roughness, object albedo, and object color, as well as illumination position, illumination intensity and illumination color. The relationship between these scene parameters and the histogram measurements is complex and cannot be solved analytically. Instead, the solution is obtained by interpolating between histogram measurements from images with known scene parameters. The algorithm is tested on simulated images, and the values recovered by the program compare very favorably with the actual parameters used to simulate the scene. The algorithm is also tested on real images, and the calculated values are reasonably close to estimates of the scene parameters obtained by other means. The algorithm presented here works very quickly, and requires only a single color image. This method may be applied to such varied tasks as surface inspection and object recognition. It demonstrates the importance of considering laws of reflection in explaining color variation, and shows why color is such a valuable feature. This type of analysis of color appearance brings us closer to the goal of allowing machines to interpret images. (Abstract shortened by UMI.)

Hierarchical Hough transform based on pyramidal architecture

Abstract: A hierarchical Hough transform (HT) based on pyramidal architecture is described, being a main component of the low-to-medium spatial vision subsystem for a mobile robot. The sequence of processing in the system originally conceived to be essential to the extraction of line features in indoor scenes consisted of: histogram equalization, smoothing with the use of a medial filter, edge detection using the Sobel edge detectors, binarization to extract the edges detected, labeling, rebinarization and thinning to refine the edges to thin lines, and line extraction using a hierarchical approach to the HT method. The task was to establish the importance of each step for the success of the hierarchical HT. It was implemented on a 386-based personal computer with 640 K memory and proved to give results of high quality as compared with the standard HT implementation. >

Real-time thresholding using histogram concavity (image processing)

Abstract: In this paper, the authors present a method of generating a real-time threshold for each incoming image frame using the histogram concavity technique. A real-time histogram hardware unit is built to facilitate such approach. The system can determine the optimal threshold value once the frame is obtained, and it takes about 10 ms to work out this value. >

Smooth vertical motion via color palette manipulation

Abstract: Smooth vertical movement of a graphics image on a raster display is performed via color palette manipulation at any speed. A color look-up table has a color palette position for each color within the graphics image plus unassigned color palette positions that are assigned to pixels on the top of horizontal color edges of the graphics image. During movement of the graphics image the color values of the assigned color palette positions are updated each display cycle as a function of a desired rate of vertical movement of the graphics image.

Multichannel histogram equalization with applications in the hsi color space

Abstract: This paper presents a novel method for multichannel and color image equalization. The equalization can be performed on the three channels RGB simultaneously, by using the joint pdf. Alternatively, equalization at the HSI domain can be performed, in order to avoid changes in digital image hue. A parallel algorithm is proposed for color image histogram calculation and equalization.

Combined vector quantization and adaptive histogram equalization

Abstract: Adaptive histogram equalization is a contrast enhancement technique in which each pixel is remapped to an intensity proportional to its rank among surrounding pixels in a selected neighborhood. We present work in which adaptive histogram equalization is performed on the codebook of a tree-structured vector quantizer so that encoding with the resulting codebook performs both compression and contrast enhancement. The algorithm was tested on magnetic resonance brain scans from different subjects and the resulting images were significantly contrast enhanced.

Color image enhancement by three-dimensional histogram modification

Abstract: Histogram-based color image enhancement is usually accomplished by transforming from RGB coordinates to another coordinate system, modifying the components represented in that system, and converting the results back to RGB. Although a few methods function directly in RGB space, they also attempt to reduce dimensionality from the histogram's original three dimensions. Such methods seldom yield images that use the full extent of RGB color range. A new method called "histogram explosion" has been developed to perform true multivariate enhancement directly in RGB color space. Discussed are the algorithm's operational parameters, behavior, implementation, and possible improvements. Results show histogram explosion to be very effective and flexible in enhancing color images. Finally, two iterative methods are suggested as possible approaches for the extension of the histogram equalization algorithm to operate upon three dimensional histograms.

A connectionist approach for color image segmentation

Abstract: A connectionist clustering strategy is presented for segmenting color images. First the local peaks in the 3-D red, green, blue histogram are located. Then using these as the prototypes other patterns are classified into one of them. The prototype selection method employs only neuronal dynamics and therefore is faster than existing clustering neural networks. The classification network takes into account the distribution of the data and hence is less prone to misclassifications. Experimental results obtained by applying the network for segmenting one color image are presented. >

An interdisciplinary course in digital image processing

Abstract: This paper documents the development and first offering of an interdisciplinary undergraduate course in Digital Image Processing at the University of Wyoming. The course itself was designed to serve majors from a wide range of academic disciplines, although in its initial offering, it was attented mainly by students majoring in Computre Science and Electrical Engineering. National Foundation funding for equipment for the course was used to purchase a high speed image processing system and six state-of-the-art graphics workstations with software that supported basic and intermediate level image processing operations. Students in the course were required to perform a standard set of image processing sequences such as histogramming and histogram equalization, edge detection and evaluation, image smoothing, region growing, Fourier filtering, and image warping. Each student, in consultation with the instructor, then pursued a specific topic in image processing which involved either combining several image processing operations to produce a desired result or developing special code to implement image processing algorithms that were discussed in the text but not included in the provided software. The nature of the course and its impact on education at the University of Wyoming is discussed in the paper that follows.

Adaptive representation of histogram using interval tree

Abstract: Proposes a new data structure, called histogram interval tree, to represent a distribution of gray levels in terms of intervals over a given gray-level range. The proposed structure will allow a user to refine the intervals adaptively as needed in each individual application. The author demonstrates that by using this new representation, the interactive specification of desired histogram becomes particularly easy and effective. Results from the proposed technique is compared with that from existing techniques for image enhancement using histogram transformation. The applications of the histogram interval tree can be extended when certain properties associated with histograms are attached to the tree. The author illustrates an example of such extension by building a threshold hierarchy for image segmentation. >

Low-light-level images reconstruction

Abstract: We propose a series of procedures to construct an image as similar as possible to that detected in good illumination conditions (standard image), starting from a low light level (L3) image. In L3 conditions, only a small number of photopulses are detected in the whole image area. An image taken in these conditions appears like a few isolated light points over a dark background. This makes it nearly impossible to recognize an object represented on it. We have developed a method based on the L3 image statistics in order to estimate the intensity received by each pixel. This method consist of a spatial average performed by a photon counting mask and can be used to construct a standard image from only one L3 image. As a second step, we have studied some histogram operations to eliminate the heavy statistics dependence that remains in the post-mask image. The best results correspond to the histogram specification but, to perform it, it would be necessary to know the standard image histogram. The last step of our work is the development of a fitting method to obtain this standard image histogram. This fitting is based on the statistical behavior of the L3 image and can be done using only a post-mask histogram as data.

Color image sense reversal

Abstract: Replication of a two-color original image with foreground and background colors exchanged is effected on a received signal representing the color information of each successive pixel of a two-color original image by generating a histogram of the color data distribution for each color plane of the original image, determining a translation point within each histogram, and performing a histogram translation of the color plane image data about the translation point. The translation point may be determined by selecting the approximate midpoint between the foreground information and the background information in each histogram as the translation point, or by selecting the statistical average of each histogram as the translation point. Once the translation point is determined, a new color data value for each pixel in a color plane is selected by substracting the pixel's original color data value from twice the color data value of the translation point.

Method and apparatus for color image processing

Abstract: A method for displaying a color image and a color palette including the steps of providing an analog representation of a color image, the analog representation corresponding to a digital representation of the color image in a first color space, providing a digital representation of a color palette in a second color space, and employing the digital representation of the color palette in the second color space in order to provide an analog representation of the color palette, the analog representation corresponding to a digital representation of the color palette in the first color space.