Showing papers on "Pixel published in 1997"

Recovering high dynamic range radiance maps from photographs

Abstract: We present a method of recovering high dynamic range radiance maps from photographs taken with conventional imaging equipment. In our method, multiple photographs of the scene are taken with different amounts of exposure. Our algorithm uses these differently exposed photographs to recover the response function of the imaging process, up to factor of scale, using the assumption of reciprocity. With the known response function, the algorithm can fuse the multiple photographs into a single, high dynamic range radiance map whose pixel values are proportional to the true radiance values in the scene. We demonstrate our method on images acquired with both photochemical and digital imaging processes. We discuss how this work is applicable in many areas of computer graphics involving digitized photographs, including image-based modeling, image compositing, and image processing. Lastly, we demonstrate a few applications of having high dynamic range radiance maps, such as synthesizing realistic motion blur and simulating the response of the human visual system.

CMOS image sensors: electronic camera-on-a-chip

Abstract: CMOS active pixel sensors (APS) have performance competitive with charge-coupled device (CCD) technology, and offer advantages in on-chip functionality, system power reduction, cost, and miniaturization. This paper discusses the requirements for CMOS image sensors and their historical development, CMOS devices and circuits for pixels, analog signal chain, and on-chip analog-to-digital conversion are reviewed and discussed.

Image segmentation in video sequences: a probabilistic approach

Abstract: "Background subtraction" is an old technique for finding moving objects in a video sequence--for example, cars driving on a freeway. The idea is that subtracting the current image from a time-averaged background image will leave only nonstationary objects. It is, however, a crude approximation to the task of classifying each pixel of the current image; it fails with slow-moving objects and does not distinguish shadows from moving objects. The basic idea of this paper is that we can classify each pixel using a model of how that pixel looks when it is part of different classes. We learn a mixture-of-Gaussians classification model for each pixel using an unsupervised technique--an efficient, incremental version of EM. Unlike the standard image-averaging approach, this automatically updates the mixture component for each class according to likelihood of membership; hence slow-moving objects are handled perfectly. Our approach also identifies and eliminates shadows much more effectively than other techniques such as thresholding. Application of this method as part of the Roadwatch traffic surveillance project is expected to result in significant improvements in vehicle identification and tracking.

Comparing images using color coherence vectors

Abstract: Color histograms are used to compare images in many applications. Their advantages are efficiency, and insensitivity to small changes in camera viewpoint. However, color histograms lack spatial information, so images with very different appearances can have similar histograms. For example, a picture of fall foliage might contain a large number of scattered red pixels; this could have a similar color histogram to a picture with a single large red object. We describe a histogram-based method for comparing images that incorporates spatial information. We classify each pixel in a given color bucket as either coherent or incoherent, based on whether or not it is part of a large similarly-colored region. A color coherence vector (CCV) stores the number of coherent versus incoherent pixels with each color. By separating coherent pixels from incoherent pixels, CCV’s provide finer distinctions than color histograms. CCV’s can be computed at over 5 images per second on a standard workstation. A database with 15,000 images can be queried for the images with the most similar CCV’s in under 2 seconds. We show that CCV’s can give superior results to color his∗To whom correspondence should be addressed tograms for image retrieval.

The representation and recognition of human movement using temporal templates

Abstract: A new view-based approach to the representation and recognition of action is presented. The basis of the representation is a temporal template-a static vector-image where the vector value at each point is a function of the motion properties at the corresponding spatial location in an image sequence. Using 18 aerobics exercises as a test domain, we explore the representational power of a simple, two component version of the templates: the first value is a binary value indicating the presence of motion, and the second value is a function of the recency of motion in a sequence. We then develop a recognition method which matches these temporal templates against stored instances of views of known actions. The method automatically performs temporal segmentation, is invariant to linear changes in speed, and runs in real-time on a standard platform. We recently incorporated this technique into the KIDSROOM: an interactive, narrative play-space for children.

Spline-Based Image Registration

Abstract: The problem of image registration subsumes a number of problems and techniques in multiframe image analysis, including the computation of optic flow (general pixel-based motion), stereo correspondence, structure from motion, and feature tracking. We present a new registration algorithm based on spline representations of the displacement field which can be specialized to solve all of the above mentioned problems. In particular, we show how to compute local flow, global (parametric) flow, rigid flow resulting from camera egomotion, and multiframe versions of the above problems. Using a spline-based description of the flow removes the need for overlapping correlation windows, and produces an explicit measure of the correlation between adjacent flow estimates. We demonstrate our algorithm on multiframe image registration and the recovery of 3D projective scene geometry. We also provide results on a number of standard motion sequences.

The pixel: A snare and a delusion

Abstract: The pixel is an explicit feature of remotely-sensed imagery, and a primary concept of the raster GIS (Geographical Information System) which is the usual vehicle for integration. This Letter addresses the underlying spatial conceptualization of the pixel, which is the parallel of the grid cell in spatial analysis, and the regular grid in sampling. It is argued that integration of remote sensing and GIS can only possibly advance if we develop methods to address the conceptual short-comings of the pixel as a spatial entity, and stop pretending that it is a true geographical object. Three major strands of research which address this issue are highlighted, including mixture modelling, geostatistics and fuzzy classification.

Protecting images with an image watermark

Abstract: A robust means of watermarking a digitized image with a highly random sequence of pixel brightness multipliers is presented. The random sequence is formed from ‘robust-watermarking-parameters’ selected and known only by the marker and/or the marking entity. A watermarking plane is generated having an element array with one-to-one element positional correspondence with the pixels of the digitized image being marked. Each element of the watermarking plane is assigned a random value dependent upon a robust random sequence and a specified brightness modulation strength. The so generated watermarking plane is imparted onto the digitized image by multiplying the brightness value or values of each pixel by its corresponding element value in the watermarking plane. The resulting modified brightness values impart the random and relatively invisible watermark onto the digitized image. Brightness modulation is the essence of watermark imparting. Detection of an imparted watermark requires knowing the watermarking plane with which the watermark was imparted. Regeneration of the watermarking plane requires knowledge of the robust-marking-parameters used in its formulation. This is generally only known to the marker and/or marking entity. Once regenerated, the watermarking plane is used together with a verifying image located in a ‘visualizer’ to demonstrate the existence of the watermark. The process of watermark detection is enhanced by application of a blurring filter to the marked image before detection is attempted.

Active matrix type liquid crystal display device

Abstract: A display device with a plurality of gate signal lines extended in the first direction; a plurality of drain signal lines extended in the second direction; a plurality of pixel regions in a state that each pixel region includes a switching device, a pixel electrode which is connected with the drain signal line and a counter electrode. The counter electrode is formed with respect to the pixel electrode by way of an insulation film and is formed every pixel. The pixels include first pixels and second pixels, wherein only the first pixels have a counter voltage signal line which extends in the first direction, and the counter voltage signal line is connected with the counter electrodes in the first pixels. Further, the counter electrodes of the first pixels and the counter electrodes of the second pixels are electrically connected with each other using conductive stride over the gate signal line.

Digital signature of color images using amplitude modulation

Abstract: Watermarking techniques, also referred to as digital signature, sign images by introducing changes that are imperceptible to the human eye but easily recoverable by a computer program. Generally, the signature is a number which identifies the owner of the image. The locations in the image where the signature is embedded are determined by a secret key. Doing so prevents possible pirates from easily removing the signature. Furthermore, it should be possible to retrieve the signature from an altered image. Possible alternations of signed images include blurring, compression and geometrical transformations such as rotation and translation. These alterations are referred to as attacks. A new method based on amplitude modulation is presented. Single signature bits are multiply embedded by modifying pixel values in the blue channel. These modifications are either additive or subtractive, depending on the value of the bit, and proportional to the luminance. This new method has shown to be resistant to both classical attacks, such as filtering, and geometrical attacks. Moreover, the signature can be extracted without the original image.

Pixel independence: measuring spatial interactions on a CRT display

Abstract: The standard working assumption of careful CRT imaging is that each pixel is imaged independently, through a point nonlinearity (the monitor's gamma function, relating screen luminance to input voltage), and then blurred by the point-spread function of the beam spot on the phosphor. Unfortunately most monitors have inadequate video bandwidth, DC restoration, and high-voltage regulation to live up to this ideal model. Two tests are recommended for assessing a CRT's deviation from the pixel-independence model.

Region-based image querying

Abstract: Retrieving images from large and varied collections using image content as a key is a challenging and important problem In this paper, we present a new image representation which provides a transformation from the raw pixel data to a small set of localized coherent regions in color and texture space This so-called lblobworldr representation is based on segmentation using the expectation-maximization algorithm on combined color and texture features The texture features we use for the segmentation arise from a new approach to texture description and scale selection We describe a system that uses the blobworld representation to retrieve images An important and unique aspect of the system is that, in the context of similarity-based querying, the user is allowed to view the internal representation of the submitted image and the query results Similar systems do not offer the user this view into the workings of the system; consequently, the outcome of many queries on these systems can be quite inexplicable, despite the availability of knobs for adjusting the similarity metric

Wide intrascene dynamic range CMOS APS using dual sampling

Abstract: A CMOS active pixel sensor (APS) that achieves wide intrascene dynamic range using dual sampling is reported. A 64/spl times/64 element prototype sensor with dual output architecture was fabricated using a 1.2 /spl mu/m n-well CMOS process with 20.4 /spl mu/m pitch photodiode-type active pixels. The sensor achieves an intrascene dynamic range of 109 dB without nonlinear companding.

Adaptive document binarization

Abstract: A new method is presented for adaptive document image binarization, where the page is considered as a collection of subcomponents such as text, background and picture. The problems caused by noise, illumination and many source type related degradations are addressed. The algorithm uses document characteristics to determine (surface) attributes, often used in document segmentation. Using characteristic analysis, two new algorithms are applied to determine a local threshold for each pixel. An algorithm based on soft decision control is used for thresholding the background and picture regions. An approach utilizing local mean and variance of gray values is applied to textual regions. Tests were performed with images including different types of document components and degradations. The results show that the method adapts and performs well in each case.

Video orbits of the projective group a simple approach to featureless estimation of parameters

Abstract: We present direct featureless methods for estimating the eight parameters of an "exact" projective (homographic) coordinate transformation to register pairs of images, together with the application of seamlessly combining a plurality of images of the same scene, resulting in a single image (or new image sequence) of greater resolution or spatial extent. The approach is "exact" for two cases of static scenes: (1) images taken from the same location of an arbitrary three-dimensional (3-D) scene, with a camera that is free to pan, tilt, rotate about its optical axis, and zoom, or (2) images of a flat scene taken from arbitrary locations. The featureless projective approach generalizes interframe camera motion estimation methods that have previously used a camera model (which lacks the degrees of freedom to "exactly" characterize such phenomena as camera pan and tilt) and/or which have relied upon finding points of correspondence between the image frames. The featureless projective approach, which operates directly on the image pixels, is shown to be superior in accuracy and the ability to enhance the resolution. The proposed methods work well on image data collected from both good-quality and poor-quality video under a wide variety of conditions (sunny, cloudy, day, night). These new fully automatic methods are also shown to be robust to deviations from the assumptions of static scene and no parallax.

Apparatus and method for combining a plurality of images

Abstract: An image combine apparatus for combining a plurality of images to generate a panoramic image. The image combine apparatus identifies an overlapping region of two inputted images and determines a boundary of the two images. The image combine apparatus then sets a tone correction area having a predetermined width such that the boundary of the two images is the center of the area, and performs tone correction within the area. The image combine apparatus performs linear tone correction in accordance with a distance between a pixel and the boundary. In the neighbor of the boundary of the two images within the tone correction area, density of the image gradually changes, thus a combined image whose boundary of the two images is inconspicuous is obtained.

Software-based encoder for a software-implemented end-to-end scalable video delivery system

Abstract: A software-based encoder is provided for an end-to-end scalable video delivery system that operates over heterogeneous networks. The encoder utilizes a scalable video compression algorithm based on a Laplacian pyramid decomposition to generate an embedded information stream. The decoder decimates a highest resolution original image, e.g., 640×480 pixels, to produce an intermediate 320×240 pixel image that is decimated to produce an intermediate 160×120 pixel image that is compressed to form an encodable base layer 160×120 pixel image. This base layer image is decompressed to form an image that is up-sampled by interpolation to produce an up-sampled 320×240 pixel image. This up-sampled image is subtracted from the intermediate 320×240 pixel image to form an error image that is compressed and encoded as a first enhancement 640×480 pixel layer. The decompressed base layer image is also up-sampled at step to produce an up-sampled 640×480 pixel image that is subtracted from the original 640×480 pixel image 200 to yield an error image that is compressed to yield a second enhancement 320×240 pixel layer. Collectively, the base and enhancement layers comprise the transmitted embedded bit stream. At the receiving end, the decoder extracts from the embedded stream different streams at different spatial and temporal resolutions. Because decoding requires only additions and look-ups from a small stored table, decoding occurs in real-time.

Projection-type display apparatus

Abstract: A projection-type video display in which the entire input video signal is first digitally corrected through digital gamma correction. Then, a subset of the corrected video signal is corrected again through analogue correction techniques. In particular this portion of the digitally corrected video signal corresponds to the subject where slopes of applied voltage-transmissivity characteristic curve (V-T curve) of liquid crystal changes relatively rapidly. The result is a more accurate gamma correction which is also less expensive than conventional, purely digital gamma correction systems. Another feature is the inclusion of three liquid crystal light valves, each having have a plurality of pixels arranged in the matrix and identical shift registers in such a way that pixels of one scanning line are driven selectively and sequentially from the same horizontal direction according to a given set of video signals, a memory where at least video signals for pixels of one scanning line are written and stored, a first drive controller for outputting video signals of one scanning line from memory in the order in which the video signals were written and supplying the video signals to pixels of a scanning line in at least one of light valves, and a second drive controller for outputting video signals of one scanning line from memory in a reverse order in which video signals were written thereto for supplying another light valve. Also disclosed is recognition unit for recognizing different types of input video signal format systems and a write controller for controlling the way of writing input video signals of each scanning line into pixels of horizontal scanning lines in the light valves based on different type of input video signal format system.

Modulation coding for pixel-matched holographic data storage.

Abstract: We describe a digital holographic storage system for the study of noise sources and the evaluation of modulation and error-correction codes. A precision zoom lens and Fourier transform optics provide pixel-to-pixel matching between any input spatial light modulator and output CCD array over magnifications from 0.8 to 3. Holograms are angle multiplexed in LiNbO3:Fe by use of the 90° geometry, and reconstructions are detected with a 60-frame/s CCD camera. Modulation codes developed on this platform permit image transmission down to signal levels of ∼2000 photons per on camera pixel, at raw bit-error rates (BER’s) of better than 10-5. Using an 8–12–pixel modulation code, we have stored and retrieved 1200 holograms (each with 45,600 user bits) without error, for a raw BER of <2×10-8.

Red-eye detection/retouch apparatus

Abstract: In a red-eye detection/retouch apparatus, a first area extraction unit receives an image including a red eye and a signal representing a small area including the red eye and designated by a user and extracts a first area including the red eye. A first pixel selection unit detects the reddest pixel from the first area and outputs coordinate values and color information of the pixel. A determination unit determines whether the other red eye is present on the basis of the signal representing the small area and the coordinate values of the reddest pixel. The determination unit outputs an instruction signal for extracting a second area where the other red eye is present upon determining that the other red eye is present. A second area extraction unit receives the instruction signal and extracts the second area from the first area. A second pixel selection unit detects the reddest pixel from the second area and outputs coordinate values and color information of the pixel. An area expansion unit receives the coordinate values and color information of the pixels and the first area, detects a red-eye area by area expansion, and outputs the red-eye area. An image synthesis unit receives the red-eye area and the image including the red eye and outputs an image whose red-eye area has been corrected to a natural pupil color.

Serial analog-to-digital converter using successive comparisons

Abstract: A method for simultaneously performing bit serial analog to digital conversion (ADC) for a potentially very large number of signals is described. The method is ideally suited for performing on chip ADC in area image sensors. In one embodiment, to achieve N-bit precision, the method employs a one-bit comparator per channel (or set of multiplexed channels) and an N-bit DAC. To achieve N bits of precision, 2 N -1 comparisons are sequentially performed. Each comparison is performed by first setting the DAC output to the desired value and then simultaneously comparing each of the pixel values to that value. If a pixel value is greater than the DAC output value, its comparator outputs a one, otherwise it outputs a zero. By appropriately choosing the sequence of comparison values, the pixel values are sequentially generated. In another embodiment, the DAC is omitted and a continuous ramp signal is generated for comparison with the analog input.

Image authentication patterning

Abstract: Apparatus for producing a self-authenticating visual image of a selected view, using a digital image forming means, such as a digital camera, together with a position determining system that provides position information, including location and/or angular orientation and/or time of observation at the time a digital image is formed and/or distance from the camera to a selected object. The digital image of the selected view is presented as an array of pixels with associated pixel values. The position information is incorporated in the digital image by altering selected pixel bit values for a selected authentication pattern P of pixels in a chosen subset CS of the pixel array; and this selected pattern may be determined using the position information. The position information may be encrypted, using an encryption key based on position information, and may be stored separate from, or as part of, the digital image. The authentication pattern is represented by an authentication key or indicium, a single number or an ordered sequence of numbers, that is transmitted with the modified digital image and incorporated position information. The authentication pattern P is preferably chosen so that any circle or other connected pixel region of diameter at least equal to a selected number D contains at least one pixel from the subset CS. The pixel array may be overlaid by a set of polygons, such as triangles, rectangles, parallelograms or hexagons, with at least one pixel from the subset CS in each polygon.

Display system having electrode modulation to alter a state of an electro-optic layer

Abstract: Methods and systems for operating a display system. An example of the display system includes a first substrate having a plurality of pixel electrodes, an electro-optic layer operatively coupled to the pixel electrodes and an electrode operatively coupled to said electro-optic layer. In one example of a method of the invention, a first plurality of pixel data values is applied to the plurality of pixel electrodes. A first control voltage is applied to the electrode to alter a state of the electro-optic layer such that the first pixel data represented by the first plurality of pixel data values is substantially not displayed. A second plurality of pixel data values, representing a second pixel data, is applied to the plurality of pixel electrodes, and a second control voltage is applied to the electrode to alter the state of the electro-optic layer such that the second pixel data is displayed. In an example of a first aspect of the invention, a voltage difference between the first control voltage and the second control voltage is reduced to reduce capacitive shifting of the second plurality of pixel data values on the plurality of pixel electrodes. In an example of a second aspect of the invention, at least one of the first control voltage, the second control voltage and a pixel data value of the second plurality of pixel data values is determined by an illumination color used in displaying said second pixel data. In an example of a third aspect of the invention, the electrode receives a composite signal over time, and a first parameter of at least one of the first control voltage and the second control voltage is selected to provide an offset, for a portion of the composite signal, from a DC balanced signal over time with respect to a particular voltage.

Edge-detection based noise removal algorithm

Abstract: What is disclosed is a method for removing noise by distinguishing between edge and non-edge pixels and applying a first noise removal technique to pixels classified as non-edge pixels and a second noise removal technique to pixels classified as edge pixels. The methodology operates on images while in a Color Filter Array (CFA) domain prior to color interpolation, and uses techniques suited to the classification, whether edge or non-edge.

Video postfiltering with motion-compensated temporal filtering and/or spatial-adaptive filtering

Abstract: A postfiltering process for improving the appearance of a video image includes motion compensated temporal filtering and spatial adaptive filtering. For each target pixel being filtered, the temporal filtering uses multiple motion vectors and one or more pixel values for a prior frame to determine one of more reference values for the target filter. In one embodiment, a weighted average of multiple motion vectors for blocks near or containing the target pixel value provides a filter vector that points to a pixel value in the prior frame. This pixel value is a reference value for the target pixel value and is combined with the target pixel value in a filter operation. Alternatively, multiple motion vectors for blocks near or containing the target pixel value point to pixel values in the prior frame that are averaged to determine a reference value for the target pixel value. In each alternative, the weighting for the average is selected according to the position of the target pixel value. The spatial filtering determines a dynamic range of pixel values in a smaller block containing the target pixel value and a dynamic range of pixel values in a larger block containing the target pixel value. The two dynamic ranges suggest the image context of the target pixel, and an appropriate spatial filter for the target pixel is selected according to the suggested context.

Pixel-matched holographic data storage with megabit pages.

Abstract: Digital data-page holograms consisting of 1024×1024 arrays of binary pixels have been stored and subsequently retrieved with an optical exposure consistent with a data rate 1 Gbit/s. Each input pixel was precisely registered with a single detector pixel, and a raw bit-error rate as low as 2.4×10-6 was demonstrated with global-threshold detection. To our knowledge, this is the first demonstration of the often-cited goal of holographic data storage of megabit data pages and a gigabit-per-second data rate.

Method and apparatus for automatic discriminating and locating patterns such as finder patterns, or portions thereof, in machine-readable symbols

Abstract: A method and apparatus for locating geometric shapes or edges thereof in data collection symbols initially samples and stores an image of light reflected from the symbol. Thereafter, two routines are performed. A first routine performs low level vision processing by identifying linked points along edges, lines, curves or within shapes. A rapid pixel linking subroutine identifies edge pixels in symbol images lacking distortion. If the edge or shape suffers from distortions, then one or more distortion compensating subroutines locate sequential linked pixels despite such distortions. The resulting coordinates of linked points, which represent lines, curves and other geometric shapes, are then employed by a second routine which identifies patterns within the identified lines/curves. Based on these identified patterns, types of symbols from various symbologies can be identified and located within the stored image.

Optical detection system, device, and method utilizing optical matching

Abstract: An optical detection system and method detects movement of an optical pointing device in a data processing environment. The system works with any surface than can diffusely scatter a collimated beam from a coherent light source. Specifically, the system comprises a coherent light source and an optical sensing assembly. The optical sensing assembly comprises a plurality of photosensor arrays and a plurality of optical elements. Each photosensor array includes pixels of a particular size and shape. Each optical element has an artificially limited aperture and is associated, through optical matching, with a respective photosensor array. The coherent light source generates a collimated beam that is diffusely reflected off of the surface. The optical sensing assembly receives the diffusely reflected, or scattered, collimated beam and passes it through the artificially limited apertures of the optical elements to the associated corresponding photosensor array. Passing the scattered light through the optical elements generates speckle images that appear on the pixels of the photosensor arrays. Based on the pixel shape, a pixel value associated with the speckle image provides a speckle image data signal. When there is translation of the pointing device, a new set of speckle images, each reassembling to a translated version of the previous speckle image, arc generated and another speckle image data signal is generated. The new and the previous speckle image data signals are then used in a motion detection analysis to determine the points of the two data signals that give a displacement value.