Abstract: Various embodiments of methods and apparatus for image deblurring and sharpening using local patch self-similarity are disclosed. In some embodiments, an input blurred image is down-sampled to generate a downsized image. The downsized image is convolved with a blur kernel to obtain a smoothed image. For each of a plurality of patches of the input blurred image, a corresponding patch in the smoothed image is found. High frequency components between each of the plurality of corresponding patches in the smoothed image and corresponding patches of the downsized image are computed. The high frequency components are applied to the plurality of patches of the input blurred images to generate a deblurred version of the input blurred image.

Abstract: Efficient and accurate estimation of the position and size of an object is achieved. Convolution of an image with a smoothing filter is repeated to generate a plurality of smoothed images L(x, y, ?i) of different scales. Then, a differential image G(x, y, ?i) between each pair of the smoothed images L(x, y, ?i) of scales ?i and ?i×2 is generated. Then, a combined image AP is generated by combining the differential images G(x, y, ?i), and a position estimating unit estimates the position of the object based on the combined image AP.

Abstract: Systems, methods and computer program products are disclosed for resampling a digital image. According to an implementation, a source image can be presharpened and upsampled to a first upsampled image having a specified image size and a first level of presharpening. The source image is also presharpened and upsampled to a second upsampled image having the specified image size and second level of presharpening that is less than the first level of presharpening. The first and second upsampled images are deblurred. A binary edge mask image is generated from the deblurred, upsampled images. The binary edge mask image is dilated and blurred to generate a deep mask image. The first and second, deblurred upsampled images are blended together using the deep mask image.

Abstract: The present invention relates to an image processing method and to a device to process a moving image, which is shot by a predetermined video shooting device, in increments of access units. The imaging blur is represented by a corresponding low-pass filter showing the properties of imaging blur which occurs at the time that said moving image is shot with said video shooting device. According to the present method an inverse filter having inverse properties as to said low-pass filter is generated. A processing is performed to apply said inverse filter to said access unit to be processed. Hereby, the step of generating an inverse filter includes an iterative algorithm in the frequency domain based on the original blurred image and the point spread function.

Abstract: An image capturing apparatus includes: an image capturing unit capturing an image of an object; an image capture controller that causes the image capturing unit to capture first color component images having a first color component by multi-shot exposure, and causes the image capturing unit to capture second and third color component images, a displacement information acquiring unit that acquires displacement information; an image adding unit that aligns and adds the first color component images based on the displacement information to generate an added image; a calculator that calculates a first point spread function based on the displacement information; a first correcting unit that corrects the second and third color component images using the first point spread function; and a combining unit that combines the added image with the corrected second and third color component images.

Abstract: A novel phase-coded aperture, associated imaging system, and design method is disclosed. The optical imaging system includes a coded-aperture followed optically by a detector array and includes an image processor. A diffraction pattern in the form of a band-limited uniformly redundant array is formed on the detector array when focusable radiation from a point source in object space is modulation by the transmission function of the coded-aperture. Since diffraction effects cannot be ignored in the optical regime, an iterative phase retrieval method is used to calculate the phase-coded aperture transmission function. Correlation type processing can be applied for the image recovery.

Abstract: A method for selectively transforming a multi-dimensional input array comprising D dimensions includes, with a computing system, determining a D-dimensional convolution of the input array at only selected points in the array, each the convolution being a function of a product of D one-dimensional kernels; determining partial convolutions at each dimension iteratively, an iterative determination of one of the partial convolutions being determined, in part, from a previous iterative determination; and collecting transformed values from the convolutions into the input array to form a transformed input array.

Abstract: The present invention relates to the parallel calculation of convoluted data. In particular, the invention relates to Gaussian pyramid construction and parallel processing of image data, such as parallel calculation of repeatedly convoluted data for use in a SIFT algorithm.

Abstract: This invention provides a system and method captures a moving image of a scene that can be more readily de-blurred as compared to images captured through the above-referenced and other known methods operating on an equivalent exposure-time interval. Rather than stopping and starting the integration of light measurement during the exposure-time interval, photo-generated current is switched between multiple charge storage sites in accordance with a temporal switching pattern that optimizes the conditioning of the solution to the inverse blur transform. By switching the image intensity signal between storage sites all of the light energy available during the exposure-time interval is transduced to electronic charge and captured to form a temporally decomposed representation of the moving image.

Abstract: A method for imaging includes receiving an input image that includes a matrix of pixels (32) having input pixel values. The input pixel values are processed so as to detect a region of saturation in the input image. A first image filtering operation is applied to the input pixel values of at least the pixels that are outside the region of saturation, so as to generate first filtered pixel values. A second image filtering operation is applied to the input pixel values of at least the pixels that are within the region of saturation, so as to generate second filtered pixel value. An enhanced output image is generated by stitching together the first and second filtered pixel values.

Abstract: A system and method are provided for approximating a diffusion profile utilizing gathered lighting information associated with an occluded portion of an object. In use, the present technique gathers information associated with an occluded portion of an object that is illuminated with a two-dimensional pattern of light including an edge which defines an illuminated portion and the occluded portion of the object. To this end, a diffusion profile of the object is approximated, utilizing such information.

Abstract: Methods and apparatus, including computer program products, implementing and using techniques for document authentication. An electronic document is presented to a user. The electronic document has data representing a signed state and a current state. A disallowed difference between the signed state and the current state is detected, based on one or more rules that are associated with the electronic document. A digital signature associated with the electronic document is invalidated in response to the detecting.

Abstract: A method of navigating a space vehicle. An image of a planet surface is received. The received image is processed to identify edge pixels and angle data. The edge pixels and angle data are used to identify planetary features by shape, size, and spacing relative to other planetary features. At least some of the planetary features are compared with a predefined planet surface description including sizes and locations of planet landmarks. One or more matches are determined between the planetary feature(s) and the planet surface description. Based on the match(es), a location of the space vehicle relative to the planet is determined.

Abstract: A method for generating a multi scale representation of an input image, the method comprising the procedures of: estimating a scale factor corresponding to said input image; determining a set of Gaussian difference kernels according to said estimated scale factor, and according to a predetermined set of Gaussian kernels; and generating a multi-scale representation of said input image by applying each of said set of Gaussian difference kernels on said input image.

Abstract: An embodiment of a method for resolving features on a probe array is described that, comprises acquiring a plurality of micro-shifted images of a region of a probe array; reconstructing an image of the probe array using the micro-shifted images; and deriving intensity values for one or more probe features disposed on the probe array from the reconstructed image.

Abstract: An autofocus method includes acquiring multiple images each having a camera lens focused at a different focus distance. A sharpest image is determined among the multiple images. Horizontal, vertical and/or diagonal integral projection (IP) vectors are computed for each of the multiple images. One or more IP vectors of the sharpest image is/are convoluted with multiple filters of different lengths to generate one or more filtered IP vectors for the sharpest image. Differences are computed between the one or more filtered IP vectors of the sharpest image and one or more IP vectors of at least one of the other images of the multiple images. At least one blur width is estimated between the sharpest image and the at least one of the other images of the multiple images as a minimum value among the computed differences over a selected range. The steps are repeated one or more times to obtain a sequence of estimated blur width values. A focus position is adjusted based on the sequence of estimated blur width values.

Abstract: An image processing apparatus includes: a controller that causes an image forming unit to form an image so that a plurality of unit images are arranged in a first direction with a constant distance between adjacent unit images; an acquisition unit that acquires read image data indicating a position and a grey level value of each pixel constituting the image formed under control of the controller; and a calculation unit that calculates a misregistration with respect to the first direction of the image formed by the image forming unit, based on a convolution operation performed on the read image data acquired by the acquisition unit, the convolution operation using a reference periodic function whose variable is a position in the first direction.

Abstract: A system and method of image processing employ mathematical deconvolution to estimate the magnitude and location of a target object within an image. Both the nature of internal reflections and the convolution process by which each internal reflection contributes to blurring of the acquired image data may be measured and modeled. In accordance with mathematical deconvolution techniques, the combined effects of these internal reflections may be reduced to the extent that respective contributions of the target object and each individual reflection may be distinguished and quantified.

Abstract: A method of generating output image data comprises obtaining derivative data relating to a reference image; obtaining a constraint for output image data; and generating the output image data from the derivative data relating to the reference image in dependence on the constraint. This method can be used to recover a robust output image from the derivative of an input image.

Abstract: Object motion during camera exposure often leads to noticeable blurring artifacts. Proper elimination of this blur is challenging because the blur kernel is unknown, varies over the image as a function of object velocity, and destroys high frequencies. In the case of motions along a 1D direction (e.g. horizontal), applicants show that these challenges can be addressed using a camera that moves during the exposure. Through the analysis of motion blur as space-time integration, applicants show that a parabolic integration (corresponding to constant sensor acceleration) leads to motion blur that is not only invariant to object velocity, but preserves image frequency content nearly optimally. That is, static objects are degraded relative to their image from a static camera, but all moving objects within a given range of motions reconstruct well.

Abstract: Disclosed are methods and devices for processing a raster image so that the white color filled in the raster image during building the raster image can be differentiated from that contained in the figure objects presented in the page, which will be inserted into the raster image later. The method comprises building the raster image and filling the raster image with a white color; and transferring the white color in a figure object described in a page and inserting the figure object into the raster image. In this way, the white color filled in the raster image during building the raster image can be differentiated from that contained in the figure objects presented in the page, which will be inserted into the raster image later.

Abstract: Embodiments of the present invention provide for improved timing control in 2-D image processing to maintain a constant rate of memory fetches and pixel outputs even when the processing operations transition to a new line or frame of pixels. A one-to-one relationship between incoming pixel rate and outgoing pixel rate is maintained without additional clock cycles or memory bandwidth as an improved timing control according to the present invention takes advantage of idle memory bandwidth by pre-fetching a new column of pixel data in a first pixel block of a next line or frame while a new column of an edge pixel block on a current line is duplicated or zeroed out. As the edge pixel block(s) on the current line are processed, the data in the first pixel block of the next line or frame become ready for computation without extra clock cycles or extra memory bandwidth.

Abstract: A projection system comprises an image input element and an optical imaging element. The image input element is configured to input an original image or a processed image. The optical imaging element, an optical system with axisymmetrical structure and specific spherical aberration, is configured to generate an image with extended depth of field on an image projection surface in accordance with the original image or the processed image.

Abstract: A method is provided for averaging a sequence of image frames. A noise-reducing filter is applied to the image frames to generate filtered frames. A deconvolution filter is applied to the filtered frames to generate corresponding deconvolved frames. The filtered frames are transformed by an affine transformation to align them, generating aligned frames. The aligned frames are motion corrected by non-linear transformation based on intensity rank matching, generating a sequence of motion-corrected frames. The motion-corrected frames are averaged to generate a resultant frame.

Abstract: A method for selectively transforming a multi-dimensional input array comprising D dimensions includes segmenting the input array into a number of sub-arrays with a computing system; determining a D-dimensional convolution of the input array at only selected points in each the sub-array, the convolution being a function of a product of D one-dimensional kernels; determining partial convolutions at each dimension iteratively, an iterative determination of one of the partial convolutions being determined, in part, from a previous iterative determination; collecting transformed sub-array values to form a transformed input array; and storing the transformed input array.

Abstract: An image processor is provided with the high-frequency component detecting unit that calculates a convolution for each of regions of interest in part of an image region of input image data, to output the high-frequency component detection result for each pixel of interest; the smoothing unit that smoothes the image data, to output the smoothed image data; and the image data processing unit that combines based on the high-frequency component detection result the image data with the smoothed image data, to output the combined image data; wherein the image data processing unit varies the combination rate for the smoothed image data, based on the high-frequency component detection result. Thereby, even when a region in the image data contains such high frequency components as a checkered pattern has, an image resistant to moiré that would be generated in magnifying or reducing of the image on a region basis can be displayed without deterioration of image quality.

Abstract: Convolutions are frequently used in signal processing. A method for performing an ordinal convolution is disclosed. In an embodiment of the disclosed subject matter, an ordinal mask may be obtained. The ordinal mask may describe a property of a signal. A representation of a signal may be received. A processor may convert the representation of the signal to an ordinal representation of the signal. The ordinal mask may be applied to the ordinal representation of the signal. Based upon the application of the ordinal mask to the ordinal representation of the signal, it may be determined that the property is present in the signal. The ordinal convolution method described herein may be applied to any type of signal processing method that relies on a transform or convolution.

Abstract: Systems and methods for processing images in a structured light system which may be used to determine the correspondences in a camera-projector system. Those correspondences can later be used to construct a 3D model, to calibrate a projector or for other purposes. The optical and geometric characteristics of the system are initially determined. The capability of establishing correspondences is affected by the limitations of the system and the properties of the surfaces. Once images of the patterns projected on the surface are acquired, they are iteratively segmented and deconvolved using the known characteristics of the system. The result is a set of correspondences with a reduction of artifacts introduced by the limitations of the system.

Abstract: Embodiments related to the removal of blur from an image are disclosed. One disclosed embodiment provides a method of performing an iterative non-blind deconvolution of a blurred image to form an updated image. The method comprises downsampling the blurred image to form a blurred image pyramid comprising images of two or more different resolution scales, downsampling a blur kernel to form a blur kernel pyramid comprising kernels of two or more different sizes, and deconvoluting a selected image in the blurred image pyramid according to a Richardson-Lucy deconvolution process in which a bilateral range/spatial filter is employed.

Abstract: Among other disclosed subject matter, a computer-implemented method for pattern matching includes receiving a pattern image, a mask image and a search image, the mask image having an arbitrary shape and identifying a portion of the pattern image. The method includes evaluating a normalized cross-correlation equation based on the pattern image, the mask image and the search image, including at least a convolution of the mask image and the search image. The method includes outputting a result of evaluating the normalized cross-correlation equation, the result indicating whether the search image matches the portion of the pattern image.

Abstract: In an ultrasonic contrast imaging method in which signals specific to a contrast agent are extracted to form an image, imaging is performed with the following three well-balanced properties: frame rate, spatial resolution, and contrast-agent selectivity. A first chirp signal whose frequency increases with time and a second chirp signal obtained by inverting the first chirp signal about the time axis are used, and they are individually transmitted and received. A decoder having decode filters each adapted for a coded sequence, which is obtained when the chirp signal is decomposed into a pulse signal and a coded sequence, is provided. The decode filters are adapted for echo signals received in response to two chirp signals, respectively, and outputs of the two decode filters are input to a subtractor, whereby the difference therebetween is determined. Thus, contrast-agent selectivity is balanced with spatial resolution.

Abstract: An image processing device according to an example of the invention comprises a differentiating section which differentiates input image data to generate gradient image data, a gradient domain deconvolution processing section which applies a deconvolution to the gradient image data, the deconvolution performing deblurring corresponding to a prior distribution of an image gradient of an image to generate deconvolved data, and an integrating section which integrates the deconvolved data to generate deblurred image data.

Abstract: Methods and a computer program product for deriving a super-resolution image of a physical object by fusing cameras of multiple resolutions (spatial, temporal, or spectral), the super-resolution image characterized by a resolution exceeding a “camera imaging resolution” associated with each of a sequence of lower-resolution images of the physical object. The sequence of images of the physical object is obtained at a plurality of relative displacements with respect to the object by a hybrid camera system comprising at least two imaging systems. The imaging systems are characterized by respective temporal and spatial resolution and by spectral sensitivity, and may be distinct from one another in one or more of the foregoing dimensions. The imaging systems are either fixed, or subject to know motion, relative to each other. Image sequences derived by each imaging system are coregistered and deconvolved to solve for a resultant sequence of images.

Abstract: An imaging device able to simplify an optical system while maintaining a high frame rate, but without needing expensive blurring restoration processing hardware, able to reduce costs, and in addition capable of obtaining a restored image little influenced by noise, and a method of same, which bypasses blurring restoration processing of an image processing device 150 by a switching unit 140 to perform camera signal processing of a camera signal processing unit 160 at the time of a through image and performs blurring restoration processing at the image processing device 150, then performs camera signal processing at the camera signal processing unit 160 to display the image only at the time of capturing an image.

Abstract: Systems and methods are disclosed to recognize human action from one or more video frames by performing 3D convolutions to capture motion information encoded in multiple adjacent frames and extracting features from spatial and temporal dimensions therefrom; generating multiple channels of information from the video frames, combining information from all channels to obtain a feature representation for a 3D CNN model; and applying the 3D CNN model to recognize human actions.

Abstract: A system and method for smoothing digital images using a non-orthogonal convolution kernel involves steps for rearranging and remapping input image data and the convolution kernel image from hexagonal coordinate mapping to orthogonal mapping of a memory space thus enabling implementation of a general box filter convolution strategy to smooth the input image.

Abstract: According to the Embodiments, an Image Processing apparatus includes a pixel interpolation processing unit. The pixel interpolation processing unit generates a sensitivity level value through addition of a first frequency range component of an image signal for a lacking color component and a second frequency range component of a frequency band lower than the first frequency range component. The pixel interpolation processing unit adjusts a ratio of the first frequency range component to be added to the second frequency range component.

Abstract: An image processing apparatus that superimposes additional information on image data, the apparatus sets, in the image data, a main region in which the superimposing intensity for superimposing the additional information is constant, and a superimposing intensity change region in which the superimposing intensity for superimposing the additional information is not constant, sets, the superimposing intensity according to the target pixel in the superimposing intensity change region, and superimposes the additional information on image data using the set superimposing intensity.

Abstract: A gray level weighting centroid method for holographic data storage is disclosed. Firstly, a first gray level image having a plurality of blocks is received. Then, a convolution calculation is performed on a weight matrix and the image, so as to obtain a second gray level image having a plurality of blocks each having a gray level value. The gray level values are divided into a bright gray level value and a dark gray level value by a threshold value, so as to convert the second gray level image into a thresholding image and find the positions of the borders of the blocks corresponding to the bright gray level value. Next, the blocks surrounded by the positions of the borders and respectively used as a centroid block are found by making the positions of the borders correspond to the first gray level image. Finally, the centroid block is calculated to obtain a centroid point.

Abstract: A method of generating an enhanced perfusion image comprising the use of a blind deconvolution algorithm and the adiabatic approximation to the Johnson and Wilson model (aaJW) and generation of an image, wherein the blind deconvolution algorithm and the aaJW model are used in the generation of values of the following parameters: voxel specific arterial input function cp[t] and voxel specific tissue residue function r[t].

Abstract: A method and system for transforming a video image from a High Dynamic Range (HDR) image on an array of pixels to a Low Dynamic Range (LDR) image. An old luminance generated from a color space of the HDR image is scaled and segmented into stripes. Each stripe has at least one row of the array. A target zone surrounding a current pixel in each stripe is determined from a search strategy selected from a linear search strategy and a zone history-based search strategy. A convolution of the scaled luminance at a current pixel of each stripe is computed using a kernel specific to the target zone. The convolution is used to convert the stripes to tone-mapped luminance stripes which are collected to form a tone mapped luminance pixel array that is transformed to the color space to form the LDR image. The LDR image is stored and/or displayed.

Abstract: Methods and a computer program product for deriving a super-resolution image of a physical object, the super-resolution image characterized by a resolution exceeding a “camera imaging resolution” associated with each of a sequence of lower-resolution images of the physical object. The sequence of images of the physical object is obtained at a plurality of relative displacements with respect to the object. An offset is passively associated with each of the plurality of images to derive effective camera movement, allowing for calculation of a kinetic point spread function on the basis of the effective camera movement. The image sequence is deconvolved, using the kinetic point spread function, to solve for a high-resolution image. Various applications such as portable cameras and infrared imaging for energy conservation are described.

Abstract: A workflow method for temporal nodule review by registering a reference image R with a floating image F, convolving the reference image R and the floating image with the same window function Hw to generate Rw and Fw, generating a subtraction image by performing subtraction Rw?Fw (g(r)) wherein r represents a voxel (x, y, z) in reference image R, applying a pattern detector to said subtraction image to detect corresponding nodules in reference image R and floating image F and displaying corresponding nodules.

Abstract: Method to bring out a temporal difference between corresponding structures in a reference image R and a floating image F by convolving the reference image R and the floating image F with a window function Hw to generate Rw and Fw, applying a non-rigid transformation resulting in a transformation field g(rR) mapping every location rR to a corresponding location rF in the floating image F and generating a subtraction image by performing subtraction Rw(r)?Fw(g(r)) wherein r represents a voxel (x, y, z) in reference image R.

Abstract: Video processing method and means for enhancing a video stream, by computing transform coefficients using a spatio-temporal transform comprising a spatial subband transform and a causal time wavelet transform performing filterings with multiscale causal wavelets, modifying the transform coefficients with a nonlinear processing, and computing a processed video stream from the modified transform coefficients using a spatio-temporal reconstruction transform comprising an inverse subband transform and a short delay inverse time wavelet transform, where the short delay inverse time wavelet transform is implemented with wavelet filters modified with window functions to control the processing delay of the entire video processing method.

Abstract: An image point in a displayed reference image R is selected and a non-rigid transformation resulting in a transformation field g(rR) mapping every location rR to a corresponding location rF in a floating image F is applied, next a rigid body transformation is applied to floating image F such that rF coincides with the selected image point and the transformed floating image is displayed.

Abstract: An image processing apparatus and an image processing method. The image processing apparatus includes a first convolution unit, a weight generator, a second convolution unit, an arithmetic unit and an outputting unit. The first convolution unit performs a convolution to output edge strength according to an original image signal and a high pass filter mask. The weight generator chooses a weight coefficient according to the edge strength. The second convolution unit performs a convolution to output an unsharp image signal according to the original image signal and a low pass filter mask. The arithmetic unit outputs a first sharpening signal according to the original image signal, the unsharp image signal and the weight coefficient. The outputting unit outputs a processed image signal according to the original image signal and the first sharpening signal.

Abstract: A method of direct volume rendering is provided comprising combining ambient occlusion volumes terms from a plurality of different filtered volumes using filters with different radii. During an ambient occlusion computation phase, the method obtains ambient occlusion terms from a plurality of different filtered volumes and, combining the ambient occlusion terms from the plurality of different filtered volumes into a composite ambient volume occlusion. During a subsequent volume rendering phase, data acquired from the subject is processed using the composite ambient volume occlusion to obtain a shaded rendition of an image of the subject.

Abstract: A digital image-processing device with a Bayer sensor and an image memory is provided in which the image data of the sensor is written into an image memory, and from this image memory, image data in the Bayer format with a length L and a width B is written continuously into a data buffer, and in which the sample values are combined by means of a computational device with the help of adders, in each case symmetrically to a central point of one or more (2n+1)×(2n+1) neighborhoods, and one or more (n+1)×(n+1) matrices are derived by means of the computational device, and from this (n+1)×(n+1) matrix or these matrices, with the help of additional adders, at least one n×n matrix is formed, and a first color component is in each case calculated from this by means of an adder network.

Abstract: A method and system for selectively transforming a spatially varying optical characteristic (F) of an image in a pixel array. The pixel array is segmented into stripes of contiguous rows. A two-dimensional convolution C(x, y) of F is determined at only selected pixels (x, y). C(x, y) is a function of a product of a horizontal kernel h(x) and a vertical kernel v(y). Determining C(x, y) at each selected pixel (x, y) includes determining n vertical convolutions, wherein each vertical convolution is equal to a scalar product of F and v(y) in a kernel space surrounding (x,y), forming an array (V) from the n vertical convolutions, and computing C(x,y) as a scalar product of V and a constant horizontal vector (H) formed from h(x). The stripes are collected to form a transformed image which is stored and/or displayed. A cache facilitates selective reuse of vertical convolutions for determining C(x,y).