scispace - formally typeset
Search or ask a question

Showing papers on "Image gradient published in 2004"


Proceedings ArticleDOI
27 Jun 2004
TL;DR: This paper examines (and improves upon) the local image descriptor used by SIFT, and demonstrates that the PCA-based local descriptors are more distinctive, more robust to image deformations, and more compact than the standard SIFT representation.
Abstract: Stable local feature detection and representation is a fundamental component of many image registration and object recognition algorithms. Mikolajczyk and Schmid (June 2003) recently evaluated a variety of approaches and identified the SIFT [D. G. Lowe, 1999] algorithm as being the most resistant to common image deformations. This paper examines (and improves upon) the local image descriptor used by SIFT. Like SIFT, our descriptors encode the salient aspects of the image gradient in the feature point's neighborhood; however, instead of using SIFT's smoothed weighted histograms, we apply principal components analysis (PCA) to the normalized gradient patch. Our experiments demonstrate that the PCA-based local descriptors are more distinctive, more robust to image deformations, and more compact than the standard SIFT representation. We also present results showing that using these descriptors in an image retrieval application results in increased accuracy and faster matching.

3,325 citations


Journal ArticleDOI
TL;DR: This paper presents a comprehensive approach to effectively and accurately extract coastlines from satellite imagery, in which the key component is image segmentation based on a locally adaptive thresholding technique and the positional precision of the resulting coastline is measured at the pixel level.
Abstract: This paper presents a comprehensive approach to effectively and accurately extract coastlines from satellite imagery. It consists of a sequence of image processing algorithms, in which the key component is image segmentation based on a locally adaptive thresholding technique. Several technical innovations have been made to improve the accuracy and efficiency for determining the land/water boundaries. The use of the Levenberg-Marquardt method and the Canny edge detector speeds up the convergence of iterative Gaussian curve fitting process and improves the accuracy of the bimodal Gaussian parameters. The result is increased reliability of local thresholds for image segmentation. A series of further image processing steps are applied to the segmented images. Particularly, grouping and labelling contiguous image regions into individual image objects enables us to utilize heuristic human knowledge about the size and continuity of the land and ocean masses to discriminate the true coastline from other object bo...

357 citations


Journal ArticleDOI
TL;DR: It is demonstrated that an acceptable, expedient solution of the energy functional is possible through a search of the image-level lines: boundaries of connected components within the level sets obtained by threshold decomposition.
Abstract: We propose a cell detection and tracking solution using image-level sets computed via threshold decomposition. In contrast to existing methods where manual initialization is required to track individual cells, the proposed approach can automatically identify and track multiple cells by exploiting the shape and intensity characteristics of the cells. The capture of the cell boundary is considered as an evolution of a closed curve that maximizes image gradient along the curve enclosing a homogeneous region. An energy functional dependent upon the gradient magnitude along the cell boundary, the region homogeneity within the cell boundary and the spatial overlap of the detected cells is minimized using a variational approach. For tracking between frames, this energy functional is modified considering the spatial and shape consistency of a cell as it moves in the video sequence. The integrated energy functional complements shape-based segmentation with a spatial consistency based tracking technique. We demonstrate that an acceptable, expedient solution of the energy functional is possible through a search of the image-level lines: boundaries of connected components within the level sets obtained by threshold decomposition. The level set analysis can also capture multiple cells in a single frame rather than iteratively computing a single active contour for each individual cell. Results of cell detection using the energy functional approach and the level set approach are presented along with the associated processing time. Results of successful tracking of rolling leukocytes from a number of digital video sequences are reported and compared with the results from a correlation tracking scheme.

222 citations


Patent
24 Dec 2004
TL;DR: In this article, a tone mapped image with HDR reconstruction data in the form of luminance ratios and color residual values is encoded and decoded to generate a restored HDR image free of the color alterations.
Abstract: Techniques are provided to encode and decode image data comprising a tone mapped (TM) image with HDR reconstruction data in the form of luminance ratios and color residual values. In an example embodiment, luminance ratio values and residual values in color channels of a color space are generated on an individual pixel basis based on a high dynamic range (HDR) image and a derivative tone-mapped (TM) image that comprises one or more color alterations that would not be recoverable from the TM image with a luminance ratio image. The TM image with HDR reconstruction data derived from the luminance ratio values and the color-channel residual values may be outputted in an image file to a downstream device, for example, for decoding, rendering, and/or storing. The image file may be decoded to generate a restored HDR image free of the color alterations.

212 citations


Book ChapterDOI
11 May 2004
TL;DR: Among image restoration literature, there are mainly two kinds of approach, one is based on a process over image wavelet coefficients, as wavelet shrinkage for denoising and the other is based over image gradient, which usually assumes that the image belongs to the space of functions of Bounded Variation.
Abstract: Among image restoration literature, there are mainly two kinds of approach. One is based on a process over image wavelet coefficients, as wavelet shrinkage for denoising. The other one is based on a process over image gradient. In order to get an edge-preserving regularization, one usually assume that the image belongs to the space of functions of Bounded Variation (BV). An energy is minimized, composed of an observation term and the Total Variation (TV) of the image.

196 citations


Journal ArticleDOI
TL;DR: The tracking experiments demonstrate that MGVF can be used to track both slow- and fast-rolling leukocytes, thus extending the capture range of previously designed cell tracking techniques.
Abstract: Recording rolling leukocyte velocities from intravital microscopic video imagery is a critical task in inflammation research and drug validation. Since manual tracking is excessively time consuming, an automated method is desired. This paper illustrates an active contour based automated tracking method, where we propose a novel external force to guide the active contour that takes the hemodynamic flow direction into account. The construction of the proposed force field, referred to as motion gradient vector flow (MGVF), is accomplished by minimizing an energy functional involving the motion direction, and the image gradient magnitude. The tracking experiments demonstrate that MGVF can be used to track both slow- and fast-rolling leukocytes, thus extending the capture range of previously designed cell tracking techniques.

156 citations


Journal ArticleDOI
TL;DR: An enhanced, region-aided, geometric active contour that is more tolerant toward weak edges and noise in images is introduced and the partial differential equation (PDE) resulting from this integration of image gradient flow and diffused region flow is implemented using a level set approach.
Abstract: An enhanced, region-aided, geometric active contour that is more tolerant toward weak edges and noise in images is introduced. The proposed method integrates gradient flow forces with region constraints, composed of image region vector flow forces obtained through the diffusion of the region segmentation map. We refer to this as the region-aided geometric snake or RAGS. The diffused region forces can be generated from any reliable region segmentation technique, greylevel or color. This extra region force gives the snake a global complementary view of the boundary information within the image which, along with the local gradient flow, helps detect fuzzy boundaries and overcome noisy regions. The partial differential equation (PDE) resulting from this integration of image gradient flow and diffused region flow is implemented using a level set approach. We present various examples and also evaluate and compare the performance of RAGS on weak boundaries and noisy images.

115 citations


Patent
04 May 2004
TL;DR: In this paper, an image processing apparatus for generating a wide dynamic range image to enable contrast to be maintained in low luminance image areas and high luminance area when the image is displayed by a narrow density range display system having an image data buffer in which short-time exposure image data is stored.
Abstract: An image processing apparatus for generating a wide dynamic range image to enable contrast to be maintained in low luminance image areas and high luminance image areas when the image is displayed by a narrow density range display system having: an image data buffer in which short-time exposure image data is stored; an image area segmenting circuit for fractionating long-time exposure image data into areas of proper and improper exposure; a segmented area image information extracting circuit for segmenting the properly exposed area of the long-time exposure image data on the basis of the segmented information and segmenting the improperly exposed area as a properly exposed area by applying the short-time exposure image data; a gradation correcting circuit for gradation-correcting image data in the properly exposed areas of the long-time exposure and the short-time exposure images which have been segmented by the segmented area image information extracting means, respectively; and an image synthesizer for synthesizing the properly exposed areas after gradation correction to form a composite wide dynamic range image.

112 citations


Reference BookDOI
01 Dec 2004
TL;DR: The need for stereology is examined, and how grids for measurement and markers for measurement, combined with Boolean operations, can be used as selection criteria for stereological techniques.
Abstract: INTRODUCTION STEREOLOGY The need for stereology Unfolding size distribution Volume fraction Surface area Design of experiments Topological properties Other stereological techniques IMAGE ACQUISITION Scanners Digital cameras Scanning microscopes File formats Color adjustment Color space coordinates Color channels Optimum image contrast Removing noise Nonuniform illumination Image distortion and focus PROCESSING AND ENHANCEMENT Improving local contrast Image sharpening Rank-based filters Edge finding Texture Directionality Finding features in images Image combinations Thresholding Automatic threshold settings using the histogram Automatic threshold settings using the image Other thresholding approaches Color image thresholding Manual marking BINARY IMAGES Erosion and dilation The Euclidean distance map Separating touching features Boolean combinations Using grids for measurement Using markers to select features Combined Boolean operations Region outlines as selection criteria Skeletonization Fiber images Skeletons and feature shape Measuring distance and location with the EDM Summary FEATURE MEASUREMENT Counting Size measurement Size distributions Comparisons Edge correction Brightness and color measurement Location Gradients Shape Identification Conclusions COLOR PLATES (insert)

108 citations


Patent
16 Jul 2004
TL;DR: In this article, a video encoding/decoding apparatus and method for color images include a first motion prediction unit that generates a first prediction residue image of an input image on a basis of a first-motion prediction result of the input image.
Abstract: A video encoding/decoding apparatus and method for color images include a first motion prediction unit that generates a first prediction residue image of an input image on a basis of a first motion prediction result of the input image. An image information detection unit sets a predetermined color component of an R-G-B image to a reference color component and determines whether the input image is a Y-Cb-Cr image or an R-G-B image and whether a color component is the reference color component. A second motion prediction unit performs motion prediction for the first prediction residue image and generates a second prediction residue image if the input image is the R-G-B image and the color component of the input image is not the reference color component.

97 citations


Proceedings ArticleDOI
01 Dec 2004
TL;DR: In this paper, a nonparametric image blur measure is presented based on edge analysis and is suitable for various image processing applications, which is obtained by combining the standard deviation of the edge gradient magnitude profile and the value of the edges gradient magnitude using a weighted average.
Abstract: A nonparametric image blur measure is presented. The measure is based on edge analysis and is suitable for various image processing applications. The proposed measure for any edge point is obtained by combining the standard deviation of the edge gradient magnitude profile and the value of the edge gradient magnitude using a weighted average. The standard deviation describes the width of the edge, and its edge gradient magnitude is also included to make the blur measure more reliable. Moreover, the value of the weight is related to image contrast and can be calculated directly from the image. Experiments on natural scenes indicate that the proposed technique can effectively describe the blurriness of images in image processing applications.

Proceedings ArticleDOI
23 Aug 2004
TL;DR: This work presents a background light intensity normalization algorithm suitable for historical document images that adaptively captures the background with a "best fit" linear function and normalized with respect to the approximation.
Abstract: This work presents a background light intensity normalization algorithm suitable for historical document images. The algorithm uses an adaptive linear function to approximate the uneven background due to the uneven surface of the document paper, aged color and light source of the cameras for image lifting. Our algorithm adaptively captures the background with a "best fit" linear function and normalized with respect to the approximation. The technique works for both gray scale and color images with significant improvement in readability.

Journal ArticleDOI
TL;DR: A new and effective image indexing technique that employs local uni-color and bicolor distributions and local directional distribution of intensity gradient and introduces the histogram of directional changes in intensity gradient.

Journal Article
HE Sai-xian1
TL;DR: An adaptive Canny algorithm of edge-detection method is proposed that not only keeps the Canny's excellent performance in good localization, only one response to a single edge and good detection, but also improves the performance in the detail edge- Detection andGood detection.
Abstract: This paper is based on Canny algorithm.An adaptive Canny algorithm of edge-detection method is proposed.This algorithm not only keeps the Canny's excellent performance in good localization,only one response to a single edge and good detection,but also improves the performance in the detail edge-detection and good detection. Canny adaptive algorithm divides image into sub-images and detects them with adaptive threshold value according to the whole image edge information, that improves the automaticity of edge-detection.With the mathematic analysis and test result,it is demonstrated that the adaptive edge-detection method is an efficient improving approach on edge-detection.

Journal ArticleDOI
TL;DR: Experimental results show that the recognition ability of the system can be drastically enhanced after integrating these two image features, which are highly complementary to provide an image retrieval system.

Journal ArticleDOI
TL;DR: An innovative edge detection algorithm, using both the gradients and the zero crossings to locate the edge positions, is presented, and experimental results indicate that the proposed edge detector is near equal to the Canny in the performance and is fast in the speed.

Proceedings ArticleDOI
23 Aug 2004
TL;DR: In this paper, a method for the registration of range and color images, based on the range intensity image that is simultaneously acquired with a range image, is proposed. But the method is limited to the detection of correspondences between images.
Abstract: This paper proposes a method for the registration of range and color (or intensity) images, based on the range intensity image that is simultaneously acquired with a range image. The gradient constraint between the range intensity image and the color image is introduced, and a linear equation for the registration parameters is derived, which combines displacement estimations for extrinsic and intrinsic parameters. By using the equation, precise automatic registration without explicit detection of correspondences is achieved. Experimental results illustrate the proposed method.

Patent
22 Nov 2004
TL;DR: In this article, the original image data is compared to the captured image data of a corresponding region of interest to determine the color difference of the captured images at the region-of-interest.
Abstract: Systems and methods for detecting image quality defects in images rendered by a rendering device. Original image data is rendered by an image rendering device, and the rendered image is captured by an image capturing device. Regions of interest may be identified to provide information indicating where image quality defects of the rendering device, may be identified. At each region of interest, the original image data may be compared to the captured image data of a corresponding region of interest to determine color difference of the captured image at the region of interest. The color difference may be subsequently converted from a device independent color space to a device dependent color space. Based on the converted color difference and input intensity of the original image data at the region of interest, a colorant error may be determined for the region of interest, and/or a scan line and a row line including the region of interest.

Patent
28 Jul 2004
TL;DR: In this article, a multi-view image generation unit was proposed for generating a multiview image on basis of an input image, which consists of edge detection means, depth map generation means, and rendering means.
Abstract: A multi-view image generation unit (100) for generating a multi-view image on basis of an input image is disclosed. The generation unit (100) comprises: edge detection means (102) for detecting an edge in the input image; depth map generation means (104) for generating a depth map for the input image on basis of the edge, a first group of elements of the depth map corresponding to the edge having a first depth value, related to a viewer of the multi-view image, and a second group of elements of the depth map corresponding to a region of the input image, being located adjacent to the edge, having a second depth value, related to the viewer of the multi-view image, the first value being less than the second value; and rendering means (106) for rendering the multi-view image on basis of the input image and the depth map.

Patent
27 Aug 2004
TL;DR: In this paper, an image compression unit converts a spectral image input via an image input unit into R, G, and B data, and stores these data in an input image storage unit.
Abstract: In a system for converting an input image signal input from an image input device into an output image signal to be output by an image output device, an image compression unit converts a spectral image input via an image input unit into R, G, and B data, obtains principal component data by making principal component analysis of the spectral image, and stores these data in an input image storage unit. When the principal component data and R, G, and B data are loaded and stored in an output image storage unit, a spectral reflectance reconstruction unit reconstructs the spectral reflectance of each pixel using these data. A printer model determines the dot quantities of inks used to record each pixel in an image output device on the basis of the calculated spectral reflectance, and generates an output image signal for the image output device. In this way, image data which allows to estimate the spectral reflectance characteristics of an input image is provided, and faithful color reproduction can be realized.

Proceedings ArticleDOI
27 Jun 2004
TL;DR: This work aims at producing a high-resolution color image directly from raw "color mosaic" images obtained by a single CCD equipped with a color filter array based on a generalized formulation of super-resolution that simultaneously performs both resolution enhancement and demosaicing.
Abstract: Super-resolution has been applied mainly to grayscale images, but producing a high-resolution color image using a single CCD sensor has not been investigated in detail until recently. This work aims at producing a high-resolution color image directly from raw "color mosaic" images obtained by a single CCD equipped with a color filter array. This method is based on a generalized formulation of super-resolution that simultaneously performs both resolution enhancement and demosaicing. The other key factor of our topic is a precise sub-pixel registration of multiple raw images. We examined direct registration of raw images based on an imaging model, which shows precise estimation of motion among severely aliased raw images. The proposed method is verified through experiments using synthetic and real images.

Proceedings ArticleDOI
23 Mar 2004
TL;DR: A shape representation method, the mountain-climbing sequence (MCS), that is invariant to translation, rotation, and scale problems and shows a superior matching ratio even in the presence of a modest level of deformation.
Abstract: Content-based image retrieval (CBIR) work includes feature selection, object representation, and matching. If a shape is used as feature, edge detection might be the first step to extract that feature. Invariance to translation, rotation, and scale is required by a good shape representation. Sustaining deformation contour matching is an important issue at the matching process. An efficient and robust shape-based image retrieval system is proposed. We use the Prompt edge detection method [H.J. Lin et al., (2001)] to detect edge points, which is compared with the Sobel edge detection method. We also introduce a shape representation method, the mountain-climbing sequence (MCS), that is invariant to translation, rotation, and scale problems. The results of our proposed method show a superior matching ratio even in the presence of a modest level of deformation.

Proceedings ArticleDOI
18 Dec 2004
TL;DR: A novel fuzzy principle component analysis (PCA) based color transfer method that accomplishes the transformation based on a set of corresponding user-selected regions in images along with a blending ratio parameter set by the user.
Abstract: Each image has its own color content that greatly influences the perception of human observer. Being able to transfer the color content of an image into another image, while preserving other features, (like texture), opens a new horizon in human-perception-based image processing. In this paper, after a brief review on the few efficient works performed in the field, a novel fuzzy principle component analysis (PCA) based color transfer method is proposed. The proposed method accomplishes the transformation based on a set of corresponding user-selected regions in images along with a blending ratio parameter set by the user. Results show more robustness and higher speed when comparing our proposed method with other available approaches.

Patent
Akihiko Iketani1, Noboru Nakajima1, Tomokazu Sato1, Sei Ikeda1, Masayuki Kanbara1, Naokazu Yokoya1 
07 Sep 2004
TL;DR: In this paper, a data processing device calculates an image conversion parameter corresponding to each partial image, and a partial image which has been imaged according to the current partial image is geometrically converted.
Abstract: A data processing device calculates an image conversion parameter corresponding to each partial image. According to the image conversion parameter calculated, a partial image which has been imaged according to the current partial image is geometrically converted. The images are combined to generate a mosaic image. Moreover, the current partial image and at least a part of the mosaic image generated are displayed on a display unit.

Patent
Hiroaki Fukuda1
06 Aug 2004
TL;DR: In this paper, a character/line drawing detection unit determines whether a part of the binary image data corresponds to a character or line drawing portion that consists of characters or line drawings.
Abstract: An image processing apparatus can obtain a copy image having excellent reproducibility with a relatively simple structure of a processing unit that processes image data. The image processing apparatus generates image formation data by processing image data. A gradation processing unit binarizes the image data according to an error diffusion process so as to convert the image data into binary image data. A character/line drawing detection unit determines whether a part of the binary image data corresponds to a character/line drawing portion that consists of characters or line drawings so as to detect the character/line drawing portion. An image processing unit applies an image processing to the part of the binary image data by using an image processing method according to a result of determination of the character/line drawing detection unit.

Proceedings ArticleDOI
23 Aug 2004
TL;DR: An appearance-based nude image detection system that is novel in that shape information is used to classify the nude images, and detect small nude images in a large background image.
Abstract: We propose an appearance-based nude image detection system. The proposed system is novel in that shape information is used to classify the nude images, and detect small nude images in a large background image. First, the proposed system finds skin regions using texture characteristics of the human skin, which then generates the skin likelihood image. Since the skin likelihood image contains shape information as well as skin color information, we used the skin likelihood image as a high level feature to classify the nude images. The image feature vector (resized skin likelihood image) is used as an input to a nonlinear-SVM. Experimental results show that the proposed system can achieve an excellent classification performance. Additionally, the proposed system can detect small nude images in a large image.

Proceedings ArticleDOI
17 May 2004
TL;DR: This paper proposes to derive a general criterion based on the probability density function using the notion of shape gradient, which is then applied to criteria based on information theory, such as the entropy or the conditional entropy for the segmentation of sequences of images.
Abstract: The paper deals with video and image segmentation using region based active contours. We consider the problem of segmentation through the minimization of a new criterion based on information theory. We first propose to derive a general criterion based on the probability density function using the notion of shape gradient. This general derivation is then applied to criteria based on information theory, such as the entropy or the conditional entropy for the segmentation of sequences of images. We present experimental results on grayscale images and color videos showing the accuracy of the proposed method.

Proceedings ArticleDOI
14 Mar 2004
TL;DR: This work uses edge direction and magnitude, called an edge map, as the main component of the speed function, and results are significantly superior to those obtained using edge magnitude alone.
Abstract: Extracting object boundaries in thermal images is a challenging task because of the amorphous nature of the images and the lack of sharp boundaries. Classical edge-based segmentation methods have the drawback of not connecting edge segments to form a distinct and meaningful boundary. Many level set approaches, which can deal with changes of topology and the presence of corners, have been developed to extract object boundaries. Previous researchers have used image gradient, edge strength, area minimization and region intensity to define the speed function. Our approach uses edge direction and magnitude, called an edge map, as the main component of the speed function. The edge map points toward the nearest boundary; its magnitude represents the total gradient energy in the half plane. The experimental results are significantly superior to those obtained using edge magnitude alone.

Patent
07 Dec 2004
TL;DR: In this paper, a method of determining location of a target image within a two-dimensional input image by employing a three-dimensional reference image comprising determining object edge points in the input image was proposed.
Abstract: Computer software for and a method of determining location of a target image within a two-dimensional input image by employing a three-dimensional reference image comprising determining object edge points in the input image; given estimated seeker line-of-sight orientation and range data, computing a two-dimensional reference image from the three-dimensional reference image; applying a weighted-direction-cross-product Hough Transform to the object edge points and points of the two-dimensional reference image; classifying the input image as containing a target image or not; if the image contains a target image, identifying a location of the target image within the two-dimensional input image; computing confidence measure as to the selected location of the target image; and in the case of sequential decision, also performing confidence accumulation through multi-frames or multi-looks.

Patent
23 Jan 2004
TL;DR: In this article, a method and apparatus for interpolating color image information are provided, where one or more image data values for a portion of a digital image in a vicinity of a target pixel are received and stored in a local array.
Abstract: A method and apparatus for interpolating color image information are provided One or more image data values for a portion of a digital image in a vicinity of a target pixel are received and stored in a local array A processor determines whether there is an edge in the vicinity of the target pixel based on the data values in the local array If there is not an edge in the vicinity of the target pixel, then long scale interpolation is performed on the image data values in the local array, in order to result in interpolating color information that is missing from the image If there is an edge in the vicinity of the target pixel, then short scale interpolation is performed using image data values in a subset of the local array in a closer vicinity of the target pixel As a result, accurate color rendering of a digital image is achieved, even in the presence of an edge portion that exhibits great contrast between regions of the image