Showing papers on "Dark-frame subtraction published in 2006"

Digital camera identification from sensor pattern noise

Abstract: In this paper, we propose a new method for the problem of digital camera identification from its images based on the sensor's pattern noise. For each camera under investigation, we first determine its reference pattern noise, which serves as a unique identification fingerprint. This is achieved by averaging the noise obtained from multiple images using a denoising filter. To identify the camera from a given image, we consider the reference pattern noise as a spread-spectrum watermark, whose presence in the image is established by using a correlation detector. Experiments on approximately 320 images taken with nine consumer digital cameras are used to estimate false alarm rates and false rejection rates. Additionally, we study how the error rates change with common image processing, such as JPEG compression or gamma correction.

CCD noise removal in digital images

Abstract: In this work, we propose a denoising scheme to restore images degraded by CCD noise. The CCD noise model, measured in the space of incident light values (light space), is a combination of signal-independent and signal-dependent noise terms. This model becomes more complex in image brightness space (normal camera output) due to the nonlinearity of the camera response function that transforms incoming data from light space to image space. We develop two adaptive restoration techniques, both accounting for this nonlinearity. One operates in light space, where the relationship between the incident light and light space values is linear, while the second method uses the transformed noise model to operate in image space. Both techniques apply multiple adaptive filters and merge their outputs to give the final restored image. Experimental results suggest that light space denoising is more efficient, since it enables the design of a simpler filter implementation. Results are given for real images with synthetic noise added, and for images with real noise

Efficient removal of impulse noise from digital images

Abstract: A new impulse noise removal technique is presented to restore digital images corrupted by impulse noise. The algorithm is based on fuzzy impulse detection technique, which can remove impulse noise efficiently from highly corrupted images while preserving image details. Extensive experimental results show that the proposed technique performs significantly better than many existing state-of-the-art algorithms. Due to its low complexity, the proposed algorithm is very suitable for hardware implementation. Therefore, it can be used to remove impulse noise in many consumer electronics products such as digital cameras and digital television (DTV) for its performance and simplicity.

Random Telegraph Signal in CMOS Image Sensor Pixels

Abstract: In this work, the 1/f noise of the Source Follower (SF) in pinned-photodiode CMOS pixels is characterized. It is found that the 1/f noise in these pixels is actually due to a very limited number of traps and results in a Random Telegraph Signal (RTS). It is pointed out how the correlated-double sampling (CDS) reacts on this RTS. The temperature dependency of the imager read noise revealed two mechanisms of RTS during CDS.

Impact of Random Telegraph Signal in CMOS Image Sensors for Low-Light Levels

Abstract: This paper presents the investigation of fluctuating pixels resulting from the Random Telegraph Signal (R.T.S) of the source-follower transistor. The work takes into account the impact of the source-follower noise power spectral density (P.S.D) dispersion through correlated double sampling (C.D.S) readout circuit used in CMOS active pixel image sensors. The results allow the determination of the output r.m.s noise versus R.T.S noise characteristics. The distinctiveness of the observed flickering pixels is discussed in detail and the proposed mechanisms behind the phenomena are viewed in light of the collected data. We show that R.T.S noise main parameters dispersion of the source-follower transistor is a major factor influencing the circuit output r.m.s noise value distribution in a pixel array. Results are compared with experimental data.

Characterization of noise in digital photographs for image processing

Abstract: Many conventional image processing algorithms such as noise filtering, sharpening and deblurring, assume a noise model of Additive White Gaussian Noise (AWGN) with constant standard deviation throughout the image. However, this noise model does not hold for images captured from typical imaging devices such as digital cameras, scanners and camera-phones. The raw data from the image sensor goes through several image processing steps such as demosaicing, color correction, gamma correction and JPEG compression, and thus, the noise characteristics in the final JPEG image deviates significantly from the widely-used AWGN noise model. Thus, when the image processing algorithms are applied to the digital photographs, they may not provide optimal image quality after the image processing due to the inaccurate noise model. In this paper, we propose a noise model that better fits the images captured from typical imaging devices and describe a simple method to extract necessary parameters directly from the images without any prior knowledge of imaging pipeline algorithms implemented in the imaging devices. We show experimental results of the noise parameters extracted from the raw and processed digital images.

Multispectral video fusion

Abstract: We present a technique for enhancing underexposed visible video footage by fusing it with simultaneously captured video from nonvisible sensors, such as short wave IR or near IR. Although IR sensors can capture video in low-light for night-vision applications, they lack the color and the relative luminances of visible spectrum sensors. RGB sensors capture color and correct relative luminances, but are underexposed, noisy, and lack fine features due to the short exposure times necessary for video. Our novel fusion technique removes noise from the RGB source (via IR/RGB mutual information) and then introduces IR texture details. By enhancing the RGB instead of IR, our results contain proper relative luminances and colors.

Efficient impulsive noise suppression based on statistical confidence limits

Abstract: Probably, the most well-known vector filter is the vector median filter (VMF) which is based on the theory of robust statistics and performs good noise suppression in color images. However, the VMF is designed to perform a fixed amount of smoothing. This may lead to too much unnecessary substitutions in the input image and, as a result, blurring and loss of image details. In order to avoid this drawback when dealing with impulsive noise, the switching schemes aim at selecting a set of pixels of the input image to be filtered leaving the rest of the pixels unchanged. In this paper, two switching filters which base the selection of the noisy pixels to be filtered on statistical tests are proposed. The proposed filters present good noise suppression while preserving fine image details appropriately. Comparisons to classical and recently introduced impulsive noise multichannel filters are provided. Moreover, the noisy pixel selection techniques are computationally simple, and the filters significantly reduce the computational complexity of the VMF.

Characterization of digital image noise properties based on RAW data

Abstract: In this paper, image noise characterization based on digital camera RAW data is studied and three different imaging technologies are compared. The three digital cameras used are Canon EOS D30 with CMOS sensor, Nikon D70 with CCD sensor and Sigma SD10 with Foveon X3 Pro 10M CMOS sensor. Due to different imaging sensor constructions, these cameras have rather different noise characteristics. The applicability of different analysis methods to these different sensor types is also studied. Digital image has several different noise sources. Separating these from each other, if possible, helps to improve image quality by reducing or even eliminating some noise components.

Method and apparatus for image processing capable of effectively reducing an image noise

Abstract: A noise eliminating device may include a noise eliminating mechanism that eliminates an isolated noise in an image being photographed. The noise eliminating mechanism may include a noise pixel detection mechanism and a pixel correction mechanism. The noise pixel detection mechanism may detect a noise pixel by scanning the image. The pixel correction mechanism may correct a level of a detected noise pixel based on a level of a pixel located in a predetermined area from the noise pixel. A noise eliminating method for eliminating an isolated noise in an image being photographed may include detecting a noise pixel by scanning the image and correcting a level of the detected noise pixel based on a level of a pixel located in a predetermined area away from the noise pixel.

Noise and zero point drift in 1.7μm cutoff detectors for SNAP

Abstract: We present the results of a detailed study of the noise performance of candidate NIR detectors for the proposed Super-Nova Acceleration Probe. Effects of Fowler sampling depth and frequency, temperature, exposure time, detector material, detector reverse-bias and multiplexer type are quantified. We discuss several tools for determining which sources of low frequency noise are primarily responsible for the sub-optimal noise improvement when multiple sampling, and the selection of optimum fowler sampling depth. The effectiveness of reference pixel subtraction to mitigate zero point drifts is demonstrated, and the circumstances under which reference pixel subtraction should or should not be applied are examined. Spatial and temporal noise measurements are compared, and a simple method for quantifying the effect of hot pixels and RTS noise on spatial noise is described.

A filter for adaptive noise reduction and sharpness enhancement for electronically displayed pictures

Abstract: Electronic images that are degraded by noise and data reduction, such as MPEG encoding, display artefacts in the reproduced image, such as ringing ('ripples') and blocks ('huge pixels'), and noise in the image may be apparent as grainyness. By performing image analysis, both on a frame-by-frame and pixel-by-pixel basis it is possible to identify and separate edges in the image, ringing artefacts and the boundaries between block transistions. By applying noise reduction according to the analysis, followed by sharpness enhancement, it is possible to clean up the image for further utilization.

System and method for sharpening vector-valued digital images

Abstract: Sharpening multi-spectral digital images without increasing noise is accomplished by filtering vector values rather than independent scalar values. A low-pass filter is performed on image A to obtain a blurred image B1 with noise and signal suppressed. The resulting blurred image B1 is subtracted from the original image A to produce a high frequency band C1 that contains noise and signal. Vector difference mean filtering is performed on the original image A to produce a filtered image B2 with noise suppressed. The filtered image B2 is subtracted from the original image A to produce a noise band C2 that contains noise with very little signal. The noise band C2 is subtracted from the high frequency band C1 to produce a signal band D that contains the signal. The signal band D is then added to the filtered image B2 to further enhance detail in the noise filtered band.

Fixed-Pattern Noise Elimination Apparatus, Solid-State Image Sensing Apparatus, Electronic Appliance, and Fixed-Pattern Noise Elimination Program

Abstract: A fixed-pattern noise elimination apparatus 4 eliminates a fixed-pattern noise component resulting from dark current out of an image signal outputted from an effective pixel of a solid-state image sensor 2 during image sensing, by using: a first dark current that is calculated based on a previously acquired signal outputted from a light-shielded pixel of the solid-state image sensor 2 and that is stored in a compensation data memory 6 ; a second dark current that is calculated based on a previously acquired signal outputted from the effective pixel of the solid-state image sensor 2 with no incident light and that is stored in the compensation data memory 6; and a third dark current that is calculated based on a signal outputted from the light-shielded pixel of the solid-state image sensor 2 during image sensing. With this configuration, a fixed-pattern noise component resulting from dark current can be effectively eliminated irrespective of the temperature of the solid-state image sensor.

Method for filtering image noise using pattern information

Abstract: Disclosed is a method for removing image noise using pattern information, which filters noise caught by a sensor during preprocessing of a compression codec, so as to increase a compression efficiency, and noise caused by the codec during post-processing of the codec, so as to obtain high quality images. The method includes the steps of: (a) carrying out region dispersion with respect to input image signals so that the image signals are dispersed with a predetermined pixel size; (b) calculating mean brightness of the input image signals and carrying out noise dispersion with respect to the input image signals; (c) switching a low frequency and a high frequency based on image signals which are subjected to the region dispersion and the noise dispersion; (d) removing noise based on a statistic after obtaining the region average with respect to the image signals having the low frequency; and (e) removing noise based on a similarity of pixels after analyzing patterns with relation to the image signals having the high frequency.

Image processing method and image processing circuit

Abstract: A method and circuit for suppressing the generation of unnatural vertical streaks in output image data. A detection processing circuit generates a first noise correction value based on first and second noise detection signals from an OB region. A correction processing circuit performs an offset process on a first noise correction value to generate a second noise correction value and performs an FIR filter process on the second noise correction value to generate a noise correction signal NC. The correction processing circuit then corrects the effective image signal from the effective image region using the noise correction signal and performs a horizontal LPF process on the corrected effective image signal to generate output image data.

Image processing device, image processing program, image processing method, and electronic camera for controlling degree of noise reduction in image data

Abstract: An image processing device changes degree of noise reduction for image data in accordance with tone correction to be performed according to the image data and includes a change rate acquisition part and a noise reduction part. The change rate acquisition part obtains, at a plurality of portions in the image data, a change rate of a signal level of the image data before and after tone correction. The noise reduction part controls a degree of noise reduction for each portion in the image data according to the change rate.

Noise eliminating device, noise eliminating method, noise eliminating program and recording medium

Abstract: PROBLEM TO BE SOLVED: To provide an imaging apparatus and a noise eliminating method capable of noise elimination corresponding to photographing conditions by detecting noise from an image being photographed, eliminating the noise by performing pixel interpolation using pixels around the detected noise, and setting an optimal noise elimination parameter based on the photographing conditions. SOLUTION: One embodiment of the present invention relates to a noise eliminating device comprising a noise eliminating means for eliminating an isolated noise in an image being photographed. The noise eliminating means may include a noise pixel detection mechanism and a pixel correction mechanism. The noise pixel detection mechanism may detect a noise pixel by scanning the image. The pixel correction mechanism may correct a level of a detected noise pixel only based on a level of a pixel located in a predetermined area from the noise pixel. COPYRIGHT: (C)2007,JPO&INPIT

Methods and systems for enhanced accuracy image noise addition

Abstract: Methods and apparatus for reconstructing a simulated image of an object that includes a predetermined amount of noise are provided. A method includes receiving a base image, determining an amount of noise in the base image, generating a noise field image based on the determined amount of noise in the base image, combining the noise field image and the base image to generate a simulated image that includes the predetermined amount of noise, and displaying the simulated image.

Line noise removal device, line noise removal method and line noise removal program

Abstract: PROBLEM TO BE SOLVED: To provide a line noise removal device or the like not degrading image quality of an area not having line noise, and capable of removing even line noise not having periodicity. SOLUTION: This line noise removal device has: an image binarization means binarizing an input image to generate a binary image; a line noise certainty factor calculation means generating a rotation image wherein the binary image is rotated about each of a plurality of rotation angles, calculating an edge characteristic amount about each of areas wherein black pixels inside each rotation image continue, and calculating a line noise certainty factor; a line noise area determination means selecting a rotation angle candidate from rotation angles, and determining a line noise area about the rotation image corresponding to each rotation angle candidate on the basis of the line noise certainty factor; a density conversion means applying local image enhancement to an area corresponding to the line noise area of the input image to generate a density conversion image; and an image composition means composing the density conversion images when there are a plurality of rotation angle candidates to generate a composite image. COPYRIGHT: (C)2008,JPO&INPIT

Reduction of position dependent noise in a digital image

Abstract: Noise is reduced in a digital image generated by an imaging device utilizing information which specifies noise level as a function of position for the digital image. A noise reducing algorithm is applied to the digital image while one or more parameters of the noise reducing algorithm are varied. The one or more parameters are varied as a function of field position in the digital image based on the obtained noise level as a function of field position. In this way, noise is substantially reduced and is spatially equalized in the digital image.

Fuzzy metrics and peer groups for impulsive noise reduction in color images

Abstract: A new method for removing impulsive noise pixels in color images is presented. The proposed method applies the peer group concept defined by means of fuzzy metrics in a novel way to detect the noisy pixels. Then, a switching filter between the identity operation and the Arithmetic Mean Filter (AMF) is defined to perform a computationally efficient filtering operation over the noisy pixels. Comparisons in front of classical and recent vector filters are provided to show that the presented approach reaches a very good relation between noise suppression and detail preserving.

Image processing method, apparatus, and program

Abstract: An image processing apparatus capable of eliminating noise from a digital image with edges maintained intact. The apparatus includes: an input means for accepting a digital image; a noise amount determination unit for determining the amount of noise in the digital image, and setting noise elimination power; a noise suppression unit for eliminating the noise from the digital image based on the noise elimination power; and an output means for outputting the noise suppressed digital image. The noise amount determination unit includes: a chrominance component separation unit for separating chrominance components from the digital image; a frequency transformation unit for transforming each of the separated chrominance components into the frequency domain; and a noise elimination power setting unit for calculating the amount of noise in each chrominance component transformed into the frequency domain, and setting the noise elimination power according the calculated amounts of noise.

Impulse noise detector using mathematical morphology

Abstract: Switching schemes have been studied for removing impulse noise. As a switching scheme, pixel-wise median of the absolute deviations from the median (PWMAD) detector was proposed. Since PWMAD detector uses only a single parameter, it is easy to optimize a parameter. However, PWMAD detector can not detect noisy pixels accurately when noisy pixels exist in neighborhood of edge pixels. In this paper, we propose an impulse noise detector using mathematical morphology. We use mathematical morphology in order to improve noise detection in neighborhood of edge pixels. Moreover, the proposed method requires only a single parameter by using mathematical morphology. Carrying out the simulation, we will illustrate the noise detection ratio of the proposed method, and show that it is more accurately to detect noisy pixels than PWMAD detector without increasing parameters.

On the Development of Impulsive Noise Removal Schemes

Abstract: Noise Suppression from images is one of the most important concens in digital image porcessing. Impulsive noise is one such noise, which may corrupt images during their acquisitioni or transmission or storage etc.A variety of techniques are reported to remove this type of noise.It is observed that techniques which follow the two satage process of detection of noise and filtering of noisy pixels achieve better performance than others. In this thesis such schemes of impulsive noise detection and filtering thereof are proposed.

Noise reducing device, electronic camera, and image processing program for reducing image noise by using blackout image

Abstract: A noise reducing device captures image data obtained by capturing a field with an image capturing part and a plurality of blackout image data obtained by capturing the field with the image capturing part under a light shielded state. This device reduces non-correlative random noise in the plural blackout image data. With random noise reduced, fixed pattern noise appears more accurately in resultant as blackout image data B. This device reduces the fixed pattern noise in the image data by using this blackout image data B.

Imaging apparatus for vehicle

Abstract: PROBLEM TO BE SOLVED: To stably output a noise-removed far infrared image. SOLUTION: A feature point extracting unit 5 extracts pixels whose luminance differences from a neighboring pixels are larger than a predetermined value as feature points A from pixels constituting a heat image picked up by an imaging unit 2 and a histogram calculating unit 6 calculates a luminance histogram showing a distribution state of luminance values of the extracted feature points A. A noise deciding unit 7 decides an output state of noise in the heat image based upon the calculated histogram and a noise reducing unit 8 removes the noise in the heat image according to a decision result of the noise deciding unit 7. Consequently, failure in long-time noise removal is eliminated, so the noise-removed far infrared image can stably be output. COPYRIGHT: (C)2007,JPO&INPIT

Circuit noise effects on star sensor position accuracy

Abstract: A method to measure image noise caused by circuit module was introduced.Firstly,many background star maps were taken at laboratory and calculated statistically,and max gray difference of the same pixel was taken for image noise caused by circuit module.The statistical data show that circuit module noise was accorded with normal probability density distribution.Secondly,based on the energy distribution model of ideal star point,a gray matrix of an ideal star of visual magnitude ~ 6 was calculated.Then star point with circuit image noise was produced with adding random noise matrix of normal distribution to ideal star point.Finally,based on subpixel centroid arithmetic and simulation data,star position error was calculated.The results show that circuit module can bring image noise of normal distribution;For 6 Mv stars at laboratory,this noise can achieve the accuracy level of 1/200 pixel.Compared with the accuracy level 1/25 pixel of actual star position measurement,circuit noise can not be neglected.

Reduction of Noise in Speech Signals through Image Processing using the Spectrogram

Abstract: Speech recognition systems have come to be used widely. When any speech recognition system is disturbed by surrounding noises, considerable reduction in the recognition rate is inevitable. It is much desired to develop noise reduction methods so that any speech recognition system can be used in realistic environments. We propose a novel scheme especially effective for reducing the noise generated in the vehicles. The reduction is achieved through image processing techniques applied to the corresponding spectrograms. Experiments have been conducted on speech sounds in the vehicles. The performances have been evaluated in terms of the output signal-to-noise ratio (SNR). The proposed scheme has been compared with the conventional spectral subtraction method, and found to be promising especially for speeches corrupted with great amount of car noises.

Image processing apparatus that reduces noise, image processing method that reduces noise, electronic camera that reduces noise, and scanner that reduces noise

Abstract: In a noise reduction apparatus or method, at least one of a luminance noise estimate value estimating a luminance noise of a target pixel and a chrominance noise estimate value estimating a chrominance noise of the target pixel is estimated in accordance with an average of luminance signals allocated to a first set of pixels including the target pixel. A noise reduction is performed to a selected one of the luminance noise and the chrominance noise one target pixel by one target pixel based on a corresponding one of a first set of information including the average of luminance signals and the luminance noise estimate value and a second set of information including an average of chrominance signals and the chrominance noise estimate value, the average of chrominance signals being allocated to a second set of pixels including the target pixel.