Showing papers on "Dark-frame subtraction published in 2002"

A multi-band spectral subtraction method for enhancing speech corrupted by colored noise

Abstract: The spectral subtraction method is a well-known noise reduction technique. Most implementations and variations of the basic technique advocate subtraction of the noise spectrum estimate over the entire speech spectrum. However, real world noise is mostly colored and does not affect the speech signal uniformly over the entire spectrum. In this paper, we propose a multi-band spectral subtraction approach which takes into account the fact that colored noise affects the speech spectrum differently at various frequencies. This method outperforms the standard power spectral subtraction method resulting in superior speech quality and largely reduced musical noise.

Fixed pattern noise removal in CMOS imagers across various operational conditions

Abstract: CMOS imagers can possess higher levels of imager noise than their predecessors, CCDs. This noise can be of the form of temporal variation and fixed pattern. The fixed pattern component of this noise can be removed, which is known already in the art. The invention in this disclosure is that proper correction can be developed for all imager conditions (imager integration time and imager temperature) using a single FPN (fixed pattern noise) dark map, a single FPN PRNU (pixel response nonuniformity) map, imager integration time and imager temperature. Without this invention, a dark frame capture and a flat field capture (integrating sphere), are required before every image capture, a practical impossibility in typical picture taking. Further, the estimates of both FPN maps (dark and PRNU) in this invention are improved estimates relative to such captured directly preceding image capture since such have be formed with multiple frame averaging at calibration time, thus removing any temporal noise from these map estimates. These dark FPN and PRNU FPN maps are modified by a scaling and biasing functional with the measured values of integration time and of imager temperature. A second approach is to make the biasing and scaling functions dependant only on mean dark response taken from the imager's dark pixels, at time of capture.

Method and apparatus for enhancing digital images utilizing non-image data

Abstract: A method and apparatus for enhancing a digital image captured by a digital camera: provides one or more data selected from the group consisting of camera model type, image sensor type, type of light source, type of image compression, previous noise reduction processing history, previous spatial sharpening processing history, and camera noise magnitude; employs the one or more data to generate one or more noise processing parameters; and employs the one or more noise processing parameters to enhance the spatial detail of the digital image.

Method and system for modifying a digital image taking into account its noise

Abstract: The invention concerns a system and a method for calculating a transformed image (I-Transf) from a digital image (INUM) and formatted data related to defects of a set of image capture and/or scanning appliances (P3). The invention comprises automatically determining data characterising noise from formatted data and/or from said digital image. The transformed image can thus be corrected so as not to exhibit visible or intruding defects, in particular noise-related defects, for subsequent use. The invention is applicable to photographic or video image processing, in optical devices, industrial controls, robotics, metrology and the like.

Efficient dark current subtraction in an image sensor

Abstract: A method for dark current subtraction which enables a dark frame to be reused for dark current subtraction for multiple image frames. The dark frame is reused by scaling it according to changes in the dark current levels associated with the dark frame and the image frames.

Computationally efficient noise reduction filter for enhancement of ultrasound images

Abstract: In ultrasound imaging, acquired images are corrupted by slowly varying multiplicative non-uniformity. When the image is corrected for non-uniformity alone, noise in the dark regions of the original image becomes multiplicatively enhanced, thereby providing an unnatural look to the image. A pre-filtering technique is used to reduce noise in ultrasound pixel images by shrinking initial image data and processing the shrunken image with known segmentation-based filtering techniques that identify and differentially process structures within the image. The segmentation is based on both gradient threshold and the distance from the near field of the ultrasound image. This modification selectively suppresses near-field artifacts. After processing, the shrunken image is enlarged to the dimensions of the initial data and then blended with the initial image to form the final image. During blending, a small predetermined fraction of intensity-dependent, uniform random noise is added to the non-structure region pixels whose intensities are above a pre-specified intensity threshold, to mitigate ultrasound speckles while leaving non-echogenic regions undisturbed.

Noise reduction techniques for Bayer-matrix images

Abstract: In this paper, some arrangements to apply Noise Reduction (NR) techniques for images captured by a single sensor digital camera are studied. Usually, the NR filter processes full three-color component image data. This requires that raw Bayer-matrix image data, available from the image sensor, is first interpolated by using Color Filter Array Interpolation (CFAI) method. Another choice is that the raw Bayer-matrix image data is processed directly. The advantages and disadvantages of both processing orders, before (pre-) CFAI and after (post-) CFAI, are studied with linear, multi-stage median, multistage median hybrid and median-rational filters .The comparison is based on the quality of the output image, the processing power requirements and the amount of memory needed. Also the solution, which improves preservation of details in the NR filtering before the CFAI, is proposed.

Incorporating pixel replacement for negative values arising in dark frame subtraction

Abstract: An apparatus and method of processing images in a digital camera and improving the quality of images produced by camera processing software. In low-light digital camera photography, a dark frame image is subtracted from an original picture image to form a resulting image. The resulting image formed from the dark frame subtraction is evaluated for negative and invalid pixel values. Each negative and invalid pixel value is replaced using a pixel replacement algorithm with appropriate positive values.

Method for sharpening a digital image with signal to noise estimation

Abstract: A method of sharpening a digital image having image pixels according to its noise content, includes the steps of providing an image sharpener having a variable parameter of sharpening; generating a noisy pixel belief map corresponding spatially to the image pixels having belief values indicating the likelihood that the modulation about respective pixels are due to system noise; and using the noisy pixel belief map to vary the parameter of the image sharpener.

Calculating noise estimates of a digital image using gradient analysis

Abstract: A method of estimating the noise in a digital image for use in subsequent image processing, includes receiving a source digital image having a plurality of pixels; selecting a pixel of interest; providing a plurality of orientations for the pixel of interest; using gradient analysis on the source digital image and the plurality of orientations to select the most suitable orientation for estimating noise for the pixel of interest; using the selected orientation in the pixel of interest to determine a noise-free pixel estimate for the pixel of interest; and repeating for other pixels of interest and using the noise-free pixel estimates to calculate a noise characteristic value representing the noise estimate for the source digital image.

Bit-depth extension with models of equivalent input visual noise

Abstract: A method for extending bit-depth of display systems. The method includes creating pseudo-random noise from human visual system noise. When applied to the image data, the noise causes spatiotemporal dithering. The pseudo-random noise is combined with image data, producing noise-compensated image data. The noise-compensated image data is them quantized.

Calculating noise form multiple digital images having a common noise source

Abstract: A method for estimating a noise characteristic value for a plurality of digital images that are affected by a common noise source includes receiving a plurality of source digital images that are affected by a common noise source, each source digital image including a plurality of pixels; calculating a total number of pixels included in the source digital images; and receiving a predetermined target number of noise estimates to be calculated for the source digital images. The method also includes using the total number of pixels and the predetermined target number of noise estimates to calculate one or more pixel sampling parameters for the source digital images; using the source digital images and the one or more pixel sampling parameters to calculate a predetermined number of noise estimates; and using the noise estimates to calculate a noise characteristic value for the source digital images.

User selected background noise removal for scanned document images

Abstract: A system and method of removing background noise from a digital image of a scanned document is described. The system and method is a reversible background noise removal technique that allows a user to select whether background noise is removed or not. In addition, since the present invention divides the background noise removal operations into a two phase process, the overall efficiency of the system and method are significantly improved.

Reduction of line noise appearance in large area image sensors

Abstract: A method of minimizing line correlated noise in an imaging system is described. In the described embodiment, gate lines are used to access pixels in an array of pixels, and the output of the pixels are read out from output lines. By randomizing the connection or positioning of the lines such that each gate or read out is coupled to pixels in different columns or rows, the line correlation of noise is reduced.

Imaging device and method for removing noise

Abstract: An imaging device in which noise due to a dark signal from a solid state image sensor can be removed. An image signal (A2) obtained through normal imaging is stored in an image memory, and an image signal (B2) obtained by closing a shutter immediately thereafter is stored in the image memory. Subsequently, the image signals (A2) and (B2) are delivered to a preprocessing circuit (6a) through a memory controller (6c). An operating circuit (61) subtracts the image signal (B2) from the image signal (A2) thus removing the fixed pattern noise and the offset component. A bright point sensing circuit (62) determines the generating position of bright point noise from the image signal (B2) and outputs a timing pulse (T1) in response to the determination. Synchronously with the timing pulse (T1), a dark point correction circuit (63) detects a pixel possibly generating dark point noise in an image signal (C1) outputted from the operating circuit (61) and corrects that pixel using the data on a nearby pixel.

Suppression of noise amplification during colour correction

Abstract: This paper relates to correcting colours in digital photography. Three low pass filters are used to prevent noise amplification during colour correction. It is well known that a straightforward application of a low pass filter reduces the sharpness of an image. The filtering is applied in a non-conventional way which naturally compensates for the blur introduced by filtering. This compensation is achieved by combining data from all the colour channels, thus preserving the original sharpness of the image. This method is particularly useful, but not limited to, CMOS image sensor based digital cameras.

Method and apparatus for correcting a retained image artifact

Abstract: A technique for compensating for a retained image includes employing bimodal readout of alternating light and dark images. The bimodal readout technique results from reading either light or dark frames more rapidly, allowing additional time to be allocated to the X-ray exposures occurring prior to the light frames or to the other reading operation. The bimodal readout may be accomplished by a binning procedure by which scan lines are binned and read, typically during dark frame readout. The images acquired from reading the dark frames may then be used to compensate for a retained image artifacts present in the image derived from light frames.

Multi-pass dark frame subtraction

Abstract: A method and system for improving the quality of an image obtained by an electronic imaging system is disclosed. The method and system comprise capturing an image frame, capturing a partial dark frame and subtracting the partial dark frame from a corresponding section of the image frame. The steps of capturing a partial dark frame and subtracting the partial dark frame from a corresponding section of the image frame are repeated for additional partial dark frames.

A motion region estimation technique for Web camera application

Abstract: This paper presents the steps involved in processing the output image obtained from image subtraction. A Web camera or Web cam is used as a video clip capture device and MATLAB version 6.01 with Image Processing Toolbox is used as the analysis software. Image frames extracted from the video clips undergo image subtraction for motion detection purposes. The continuous steps presented in this paper are thresholding, noise removal and motion region estimation. Thresholding determines the areas of output images (from subtraction) consisting of pixels with values lying within the threshold value range. Threshold value also indicates the sensitivity of motion to detection. The image still contains a small amount of noise after thresholding. Noise is removed using a median filtering method. Motion region estimation is done by executing an AND or OR operation on two subtracted images from three successive frames. Both the output image from AND and OR operations will estimate the motion region in a different time frame.

Noise reduction apparatus, noise reduction method, program and medium

Abstract: A noise reduction apparatus with simultaneous processing for allowing Simultaneous processing of image signals of a portion containing a noise processing object pixel to be inputted, wherein the image signals are existing at Spatially different locations or the image signals are contained in fields or frames at different times. A filtering is performed by low pass filter. A noise component extractor extracts noise components based upon outputs of the filter operation and noise attenuation means of attenuating noise using outputs of the noise components extraction means with respect to the noise processing object pixel within the outputs of the simultaneous processing means.

Automatic classification for noise of infrared images into processes by means of the principal component analysis

Abstract: Noise characterization and classification is an important task to evaluate the performance of an infrared imaging system. The focal plane array infrared cameras present several types of noises: fixed pattern noise, 1/f noise, pure temporal noise, etc. The existence of bad pixels showing a singular behavior must be included in the noise description. In this paper we show how the principal component analysis is able to classify the noise of a set of frames into different subsets. The classification method is integrated into a software package that performs the classification of the obtained eigenimages into processes. This method is specially adapted to the analysis of noise in a set of frames because it produces a corresponding set of images characterizing the noise. A result of the analysis provided with this method is the extraction of the fixed pattern noise, the bad pixel identification, the 1/f nosie components and analysis, the pure temporal noise, and some other processes having intermediate time scales.

Autofocusing method that is insensitive to scene illumination level

Abstract: A method, system and program product for providing automatic focus adjustment for an image device, comprising the steps of: differentiating an image along some axis to obtain a difference image; computing a variance of the difference image; determining a noise contribution to the variance; subtracting the noise contribution from the variance; using the adjusted noise variance as a factor in making the automatic focus adjustment. In a preferred embodiment, the variance is normalized, and the noise contribution is determined by determining the shot noise and the read noise.

Removal of background noise of image data by an image processing device or image reading apparatus

Abstract: An image processing device includes a background noise detecting device configured to detect from image data of an image, background noise data of the image data. An image memory stores the image data, and an inputting/outputting device inputs and outputs the image data into and from the image memory. A background noise data removing device removes the background noise data from the image data outputted from the image memory by the inputting/outputting device. An on/off device turns on and off a background noise data removing operation of the background noise data removing device according to an external control signal.

Imaging Filter and Noise Analyse Based on Low Intense X-ray Image System

Abstract: New structure of a low intensity Xray image system is mainly made of plane plate mode Xray intensifier of single proximity focus and CCD data acquisition and processing system. By the system composition, the paper analyses the image noise source of low Xray imaging system, and points out that the random noise is white noise which is obeyed by Poisson distribution in the whole body, yet the positivenegative interfering impulse is excited in the some locality. Then the compound methods of the "multiframe mean + extremum median filter" is submitted which deals with the imaging noise. Firstly, some frame images is superimposed, then mean image is calculated from those images, which is under the principle of noise noncorrelation. By the method, the information is enhanced and the noise is compressed. Secondly, based on the standard median filtering algorithm, the extremum median filter is ordered as much as possible to preserving the detail of the image when the noise are removed. That is to say, all the pixels are separated into signal and noise pixels according to the decision criterion given in the following; then, noise pixels are replaced with the median value of their neighborhood in the input image. The decision criterion: if a pixels value is the extremum (max or min) of its neighborhood, it is a noise pixel; else it is a signal pixel. This decision criterion is under such an assumption: inherent relationships exist among neighbor pixels. If a pixels value is far higher or lower than the others' value of its neighborhood, we may consider that it had been contaminated with noise. Else, if it is similar to the others, we consider that it represents an effective signal. By the calculation of PSNP, the methods are supper to any single method greatly. And the effect of image filter is satisfied.

Noise reduction in multispectral images using the self-organizing map

Abstract: In this paper, a new group of noise reduction methods for multispectral images is presented. First, a 1-dimensional Self-Organizing Map (SOM) is taught using the pixel vectors of the noisy multispectral image. Then, a gray-level index image is formed containing the indexes of the SOM vectors. Several gray-level noise reduction methods are applied to the index image for three noise types: impulse, Gaussian, and coherent noise. Tests are made for three kinds of noise distrubutions: for all channels, for channels 30-50, and for 9 selected channels. Error measures imply that the obtained results are very good for coherent noise images, but rather poor for other noise categories, compared to the bandwise coherent filter.© (2002) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

Design of a particle image velocimetry objective maximizing the image signal-to-noise ratio

Abstract: Particle image velocimetry systems use a camera to take snapshots of particles carried by a fluid at some precise instants. Signal processing methods are then used to compute the flow velocity field. In this paper, the design of the camera objective (optics) is addressed. The optimization is performed in order to maximize the signal-to-noise ratio of in-focus particles. Four different kinds of noise are considered: photon shot noise, thermal and read noise, background glow shot noise and noise made by the other particles. A semi-empirical model for the lens aberrations of a two-doublet objective is first addressed, since it is shown that lens aberrations (low f-value f#) should be used instead of the Fraunhofer diffraction (high f-value) for the fitting of the particle image size with the pixel size. Other important conclusions of the paper include the expression of optimum values for the magnification M, for the exposure period τ and for the pixel size ξ.

Image Noise Analysis and Pre-processing for Pattern Match

Abstract: This paper explains an image pre-processing methodology when using a single-camera 3D vision system for a CMM positioning. Three types of noise, mainly caused by 1) electromagnetic interference 2) season, weather and time changes and 3) object position or view angle changes, are classified and analyzed. Based on the analysis pre-processing measures are proposed and employed to minimize the influences of noise. Theoretical analysis and experiments show that the methodology has the advantages of simplicity, efficiency and robustness.

Arbitrary background picture segmentation

Abstract: In this paper, we present a new method for real time arbitrary background picture segmentation. We consider the following picture conditions: color picture, noisy picture, scene light changes, and still image arbitrary background. They are typical for many applications, e.g. for video security system, videophone, videoconference, V-commerce, etc. A set-theoretic approach has been used for picture model creation, adaptive picture processing and noise reduction. The presented method main problem, namely noise immunity and resolving power, is investigated by using picture clustering. The both luminance and chrominance picture components use allows to avoid an influence of "undesirable" object light changes and noise. It is shown that for arbitrary background picture segmentation, the required signal-to-noise ratio must be from 40 dB to 44 dB, depending on the frames change rate.

Low light level image processing with the method of frame integral

Abstract: In the low light level image processing,the time and space domain noise in the signal not only limits the lowest illuminance of the system but also make the image show random glitter. The quality of the image and the operative distance of the system can be enhanced by the image processing. It testified that photon noise and particle noise can be well controlled by the method of frame integral.