Showing papers on "Histogram equalization published in 2000"

Adaptive image contrast enhancement using generalizations of histogram equalization

Abstract: This paper proposes a scheme for adaptive image-contrast enhancement based on a generalization of histogram equalization (HE). HE is a useful technique for improving image contrast, but its effect is too severe for many purposes. However, dramatically different results can be obtained with relatively minor modifications. A concise description of adaptive HE is set out, and this framework is used in a discussion of past suggestions for variations on HE. A key feature of this formalism is a "cumulation function," which is used to generate a grey level mapping from the local histogram. By choosing alternative forms of cumulation function one can achieve a wide variety of effects. A specific form is proposed. Through the variation of one or two parameters, the resulting process can produce a range of degrees of contrast enhancement, at one extreme leaving the image unchanged, at another yielding full adaptive equalization.

A hierarchical approach to color image segmentation using homogeneity

Abstract: In this paper, a novel hierarchical approach to color image segmentation is studied. We extend the general idea of a histogram to the homogeneity domain. In the first phase of the segmentation, uniform regions are identified via multilevel thresholding on a homogeneity histogram. While we process the homogeneity histogram, both local and global information is taken into consideration. This is particularly helpful in taking care of small objects and local variation of color images. An efficient peak-finding algorithm is employed to identify the most significant peaks of the histogram. In the second phase, we perform histogram analysis on the color feature hue for each uniform region obtained in the first phase. We successfully remove about 99.7% singularity off the original images by redefining the hue values for the unstable points according to the local information. After the hierarchical segmentation is performed, a region merging process is employed to avoid over-segmentation. CIE(L*a*b*) color space is used to measure the color difference. Experimental results have demonstrated the effectiveness and superiority of the proposed method after an extensive set of color images was tested.

Towards automatic image enhancement using genetic algorithms

Abstract: This paper introduces a new automatic image enhancement technique based on real-coded genetic algorithms (GAs). The task of the GA is to adapt the parameters of a novel extension to a local enhancement technique similar to statistical scaling, as to enhance the contrast and detail in the image according to an objective fitness criterion. We compared our method with other automatic enhancement techniques, like contrast stretching and histogram equalization methods. Results obtained, both in terms of subjective and objective evaluation, show the superiority of our method.

Locally adapted histogram equalization

Abstract: A system and method for improving the uniformity in exposure and tone of a digital image using a locally adapted histogram equalization approach. This approach involves first segmenting the digital image into a plurality of image patches. For each of these patches, a pixel brightness level histogram is created. The histogram for each patch is then optionally averaged with the histograms associated with a prescribed number of neighboring image patches. A normalized cumulative distribution function is generated for each patch based on the associated averaged histogram. This normalized-cumulative distribution function identifies a respective new pixel brightness level for each of the original pixel brightness levels. For each of the original pixel brightness levels, the 1s associated new pixel brightness levels from one or more of the image patches are blended. Preferably, this blending is accomplished using either a bilinear or biquadratic interpolator function. Finally, for each image patch, the original pixel brightness level of each pixel in the image patch is replaced with the blended pixel brightness level corresponding to that original brightness level. A further refinement can also be implemented to mitigate the effects of noise caused by areas of a single color in the scene depicted in patch. In one embodiment, this refinement entails employing a partially equalization approach. In another embodiment, the refinement entails limiting the gain exhibited by any of the blended pixel brightness levels associated with an image patch, in comparison to its associated original pixel brightness level, to a prescribed level.

An advanced contrast enhancement using partially overlapped sub-block histogram equalization

Abstract: In this paper, an advanced histogram equalization algorithm for contrast enhancement is presented. Histogram equalization is the most popular algorithm for contrast enhancement due to its effectiveness and simplicity. Global histogram equalization is simple and fast, but its contrast enhancement power is relatively low. Local histogram equalization, on the other hand, can enhance overall contrast more effectively, but the computational complexity is very high due to its fully overlapped sub-blocks. For high contrast and simple calculation, a low pass filter type mask is proposed. The low pass filter type mask is realized by partially overlapped sub-block histogram equalization (POSHE). With the proposed method, the computation overhead is reduced by a factor of about one hundred compared to that of local histogram equalization while still achieving high contrast.

An efficient color re-indexing scheme for palette-based compression

Abstract: This paper presents a fast and efficient way for color re-indexing that tends to maximize the compression performance of a palette-based compression system. The proposed scheme relates the index difference of neighboring pixels to the potential cost of bits. It optimizes the assignment of index values to colors in a one-step look-ahead greedy fashion. Experimental results suggest that the proposed re-indexing scheme can reduce the bit rate by up to 43%, when compared to a previously proposed intensity-based color-indexing scheme. Furthermore, we show that with the proposed color re-indexing scheme, the palette-based JPEG-LS and palette-based JPEG-2000 can often outperform the graphical interchange format (GIF) significantly.

Video contrast enhancement through partial histogram equalization

Abstract: Methods and apparatus for video contrast enhancement are disclosed. A pixel level threshold is set for an input video frame in a video sequence. For pixel levels in the input video frame that are below the threshold, an adaptive contrast-enhancing function is applied. For other pixel levels, a scene-stable mapping function is applied. This contrast enhancement method can improve the contrast in darker areas of a scene depicted in the video sequence, without destroying the intended light levels for a scene or causing temporal brightness fluctuations in the enhanced video sequence.

Image representations for object detection using kernel classifiers

Abstract: This paper presents experimental comparisons of various image representations for object detection using kernel classifiers In particular it discusses the use of support vector machines (SVM) for object detection using as image representations raw pixel values, projections onto principal components, and Haar wavelets General linear transformations of the images through the choice of the kernel of the SVM are considered Experiments showing the effects of histogram equalization, a non-linear transformation, are presented Image representations derived from probabilistic models of the class of images considered, through the choice of the kernel of the SVM, are also evaluated Finally, we present a feature selection method using SVMs, and show experimental results

On “shapes” of colors for content-based image retrieval

Abstract: Color is a commonly used feature for realizing content-based image retrieval (CBIR). Towards this goal, this paper presents a new approach for CBIR which is based on well known and widely used color histograms. Contrasting to previous approaches, such as using a single color histogram for the whole image, or local color histograms for a fixed number of image cells, the one we propose (named Color Shape) uses a variable number of histograms, depending only on the actual number of colors present in the image. Our experiments using a large set of heterogeneous images and pre-defined query/answer sets show that the Color Shape approach offers good retrieval quality with relatively low space overhead, outperforming previous approaches.

Robust face detection against brightness fluctuation and size variation

Abstract: Proposes a face detection method intended to be used for a practical intelligent environment and human-interactive robot. Face detection and recognition are very crucial for such applications. However in the real situation, it is not easy to realize the robust detecting function because position, size, and brightness of the face image are very changeable. The proposed method solves these problems by combining correlation-based pattern matching, histogram equalization, skin color extraction, and multiple scale images generation. The authors have implemented a prototype system based upon the proposed method and conducted some experiments using the system. Results support the effectiveness of the proposed idea.

A Method of Color Histogram Creation for Image Retrieval

Abstract: The traditional method of color histogram creation is toequally subdivide a color space (e.g. RGB, HSI) into acertain number of bins and then count the number of pix-els each bin contains. This strategy results in a quite largenumber of bins with trivial color differences betweenadjacent bins. Consequently small changes in the scene(e.g. changes in the illumination conditions, presence ofnoise) may cause important modifications of the histo-gram. We proposed a new method of color histogram cre-ation based exclusively on the hue component in thechromatic image region and on the intensity componentin the achromatic image region. The color appearance ofthe image is described using a relatively small number ofbins. The proposed method of histogram creation hasbeen evaluated based on the performances achieved inretrieving similar images in a heterogeneous image col-lection. The experimental results reveal that the proposedmethod is less sensitive to small changes in the sceneachieving higher retrieval performances than the tradi-tional method of histogram creation.1. INTRODUCTIONThe most popular technique for image retrieval in a heter-ogeneous collection of images is the comparison ofimages based on their histograms. The histogramdescribes the gray-level or color distribution for a givenimage. It is a global feature which can be used to performa fast but no so reliable indexing process. The histogramfeature can be used as a preliminary step for databaseindexing in order to reduce the number of candidateimages for the next steps which could use other features(e.g. shape, texture, orientation) to compare the databaseimages with a given query image. The major advantageoffered by the histogram feature consists in its small sen-sitivity to scale, rotation and translation [1]. An appropri-ate color space, a color quantization scheme, a histogramrepresentation, and a similarity metric are the main ingre-dients required for the design of a histogram basedretrieval system [2]. The RGB color space is inappropri-ate for image retrieval due to the fact that it is not relatedwith the way humans perceive colors. Other color spaceslike opponent color space [1], HSI or YIQ are generallyused for retrieval proposes [2], [3]. The Lu*v* space isalso used because it yields a perceptually uniform spac-ing of colors [4].Once a certain color space is subdivided in a number ofbins, the histogram is created by simply counting thenumber of pixels each bin contains. This strategy usuallyresults in a very large number of bins, and hence thecolors represented by adjacent bins would reveal onlytrivial differences. Consequently small changes in thescene (e.g., change in the illumination conditions) or thepresence of noise usually determine large number of pix-els to drift from one bin to another. As a result twoimages which are quite similar one to each other mayhave very different histogram representations.In our method a relatively small number of bins is used inorder to describe the most prominent colors which maybe perceived in the image. The histogram is created basedon the hue and intensity components. The two compo-nents are weighted according with their relevance in dif-ferent image regions based on the value of standarddeviation of the RGB tristimuli.The paper is organized as follows. The proposed methodof histogram creation is described in Section 2. Someexperimental results and comparisons are shown in Sec-tion 3, and some concluding remarks are then presentedin Section 4.2. THE PROPOSED METHODThe hue (H) component is the most suitable one to use inorder to describe the color content of a digital image. Itcontains most of the color information and hence it isalmost constant regardless of the changes in the illumina-tion conditions (e.g., shadows which usually occlude theobjects in a natural image) [5]. However, natural imagesoften contain achromatic regions where the hue compo-

Method and apparatus for estimating feature values in a region of a sequence of images

Abstract: A method and apparatus are disclosed that estimate the brightness or other feature values of unchanging or slowly changing regions of an image in a sequence of video images even when the regions is obscured by objects over large portions of the video sequence. The apparatus and method generate a histogram for each image region position over a plurality of image frames in the sequence. The mode, or most frequently occurring value, of the image region as indicated by the histogram is selected as representing the unchanging portion of the image. The mode values of all of the regions are then assembled to form a composite image of the unchanging or slowly changing feature values. According to one method, the histogram is generated using a recursive filter. In order to process images that exhibit some motion from frame to frame, the images in the video sequence may be aligned before generating the histogram. If the camera produces artifacts such as variations in the image caused by an automatic gain control (AGC) function, each image in the sequence of video images may be filtered either temporally or spatially before performing the histogramming operation to remove these artifacts. To reduce processing time, the image processing may be spaced in time such that only every nth image is processed. Alternatively, each region of an image sequence may be processed at random irregular intervals in order to obtain the histogram. In one embodiment of the invention, the histogram is applied over relatively small groups of frames in order to generate a noise reduced image.

Optical system controller, optical system control method, and recording medium

Abstract: PROBLEM TO BE SOLVED: To quickly perform an autofocus, even if the main object has the bias of the color. SOLUTION: An AF control part of a digital camera is provided with a histogram generating circuit 251 which generates the histogram of the width of the edge in AF area, a noise removal part 263 which removes the noise component from the histogram, a histogram evaluation part 264 which determines the evaluation value showing the degree of the focus from the histogram, and an object color selection part 281 which selects the color component utilized for AF control. Furthermore, the AF control part is provided with a driving quantity determining part 265 which determines the driving quantity of the focus lens. In the object color selection part 281, the color component from which edge is most detected is determined as an object color from the image for every color component which constitutes the color image, and the result is inputted into the histogram evaluation part 264. In the histogram evaluation part 264, the evaluation value about the object color is obtained, and the driving quantity determining part 265 quickly positions the focus lens in the focal position, while changing the driving quantity by using the evaluation value of the object color.

Color image watermarking in HSI space

Abstract: For graylevel images, the histogram based watermarking approach considers the image histogram as its signature and, by exact histogram specification, images are marked accordingly. This paper extends to color images the histogram based watermarking. Image marking is performed in the hue, saturation, intensity (HSI) space by creating notches in the intensity component histogram. The number of notches and their locations are the features which define each watermark. The detection is blind and of considerably low complexity. Robustness of the proposed method is investigated-tests performed using StirMark show that marking with notches ensures high robustness against many attacks: JPEG compression, geometric distortions (cropping, scaling, rotation, row and column removal, translation, etc.,), linear and nonlinear filtering.

A composite histogram for image retrieval

Abstract: In this paper, we propose a composite histogram for image indexing and retrieval. The histogram represents the composition of edge and luminance features as well as colors in images. As a result of combining the different features of the image together, we can reduce the size of the histogram drastically without degrading the retrieval performance. Since the proposed histogram contains only 14 bins, it has great advantage in terms of memory and computational complexity in the image retrieval. Experimental results show that the proposed histogram with 14 bins yields better retrieval performance than the RGB color histogram with 256 bins and the HSV color histogram with 166 bins.

Image retrieval using flexible image subblocks

Abstract: In this research, we propose a flexible subblock image retrieval algorithm which is robust to object translation, lighting change, object appearance or disappearance in an image which is divided into 9 non-overlapping subblocks. Furthermore, using Ohta-color moments and biorthogonal wavelet frames from each subblock, we can reduce the dimension of color space and improve the performance. Also, as the two features are applied to the multistep k-nearest neighbor algorithm, this system clearly outperforms the global color histogram, R G B moments, and HS1 moments using the same method. In addition, we provide another retrieval environment for the user which is a relative block location search as well as an absolute block location search. In the case of the block location search, the user selects a block the user wants to search.

Image retrieval system and method using image histogram

Abstract: An image retrieval system and method using an image histogram for determining central points and dispersion values as well as quantity information of color about respective histogram bins, thereby using these as mapping information for image retrieval. The image retrieval method using an image histogram includes the following steps. A first step of computing an image histogram bin when an image is inputted, and accumulating values of x, y, x 2 , y 2 to compute central points and dispersion values. A second step of normalizing the respective central points and dispersion values through dividing these by size of whole image, and storing these. A third step of generating a value of model to be retrieved by drawing a feature vector when a query image is inputted, and computing the difference between the generated value of model and central points and dispersion values of an image histogram, count, and number of corresponding bins within the data stored in the second step. A fourth step of specifying a similarity value of an image using the values computed in the third step.

Approximating the statistical distribution of color histogram for content-based image retrieval

Abstract: With the rapid development of multimedia technologies, the problem of how to retrieve a specified image from a large amount of image databases becomes an important issue. We propose a new method by approximating the statistical distribution of color histogram for image indexing and comparison. It becomes simple to compute the distance measure of two images by the proposed method. Also, our proposed method tolerates the generally considered problems when retrieving images such as transition, scaling and rotation. The represented features of the approximated distribution of the color histogram can also be used as the indices for the image database. Experimental results show that our proposed method is quite effective not only for the performance but also better results for image indexing and retrieval.

Contrast Enhancement of Infrared Image via Wavelet Transform

Abstract: A new scheme of contrast enhancement of infrared image via wavelet transform is proposed. By means of wavelet transformation of the low contrast infrared image, its distribution of multiscale gradient can be achieved .We improve image contrast by modifying the modulus of the gradient image on multiple scales rationally and enlarging its dynamic scope in scale space directly, and we can control noise magnification. Experimental results demonstrate this method is better than histogram equalization under this circumstance.

Method for searching multimedia data using color histogram

Abstract: The present invention relates to a multimedia search method for searching multimedia objects using color histogram. The present invention provide a multimedia retrieval method and a multimedia feature structure capable of accelerating accurate and efficient multimedia object search by adopting various meaning reflection level according to each bin value of the color histogram. In the present invention, the color groups are classified into the solid, neutral, and gray levels, and the binarization standard point is determined at each color group such that the color distribution can be effectively represented in the binary histogram, resulting in improving the retrieval performance of binary color histogram which is a shortcoming in the conventional binary histograms in addition to remaining the high efficiency of the binary color histogram.

Representing an image with a posterized joint histogram

Abstract: Representing an image includes extracting information reflecting one or more characteristics of an image that then is used to compute a first histogram vector. The first histogram vector includes one or more vector elements, each representing information for a different characteristic extracted about the image. Multiple subsets are identified within at least one of the vector elements included in the first histogram vector and a second histogram is created. The second histogram vector includes a vector element for each of the subsets identified within the vector elements included in the first histogram vector. Data within the vector elements of the second histogram vector represents the extracted image characteristics.

Image detector, image detection method, digital camera and printer

Abstract: PROBLEM TO BE SOLVED: To quickly and fully detect whether an image similar at least to a given image searched for is present in an input image as a search object or not. SOLUTION: An image detector 10 comprises an image inputting part 1, a color converting part 2, a search region setting part 3, a hue/modified saturation (HQ) histogram preparing part 4, an HQ histogram comparing part 5, a similarity determining part 6, a region position storing part 7 and a similar region information outputting part 8, and searches to determine whether an image searched for is present in an input image or not. A rough search for determining similarity according to a color histogram with low gradation resolution first computes a candidate region wherein the image searched for is possibly present, and a detailed search for determining similarity according to a color histogram with high gradation resolution next detects a region wherein the image searched for is present in the candidate region.

Device and method for detecting image, digital camera and printer

Abstract: PROBLEM TO BE SOLVED: To exactly detect whether or not an image similar to a prescribed retrieving image at least exists in an input image as a retrieval target even when there are a little image data. SOLUTION: An HQ histogram preparing part 4 decides whether the number of pixels in a partial area and the retrieving image is less than a prescribed value or not and when it is less than the prescribed value, a color histogram, to which smoothing processing is applied, is prepared. An HQ histogram comparing part 5 compares the partial area of the input image with the color histogram of the retrieving image. A similarity deciding part 6 finds the similarity of compared color histograms, decides whether similarity thereof is higher than a prescribed level or not and defines the partial area higher than the prescribed level as a detection area, where the retrieving image exists.

Digital camera, image processing method and recording medium

Abstract: PROBLEM TO BE SOLVED: To reduce the time required for displaying a histogram on the basis of image data obtained by image pickup from the image pickup as much as possible. SOLUTION: The image processing method has image pickup steps A01, A04 where an object is picked up to obtain image data, a thumbnail image generating step A05 where thumbnail image data are generated from the obtained image data, a histogram collection step A08 where data for a histogram are collected on the basis of the generated thumbnail image data, and a histogram display step A09 where the obtained histogram data that are superimposed on the thumbnail image and the resulting image is displayed. COPYRIGHT: (C)2002,JPO

Hiding Information in Color Images Using Small Color Palettes

Abstract: The small size of a color palette is utilized to include hidden information in the true-color image. Adaptive vector quantization is performed to define color palette, which is used for purposes to extract the hidden information. The resulting Voronoi tessellation is approximated by 3D cubes to define regions centered in the prototype vectors that are used to code hidden information. The regions are partitioned into cells the identity of which encode the hidden information in each pixel.