Showing papers on "Dark-frame subtraction published in 2000"

System and method for dual microphone signal noise reduction using spectral subtraction

Abstract: Speech enhancement is provided in dual microphone noise reduction systems by including spectral subtraction algorithms using linear convolution, causal filtering and/or spectrum dependent exponential averaging of the spectral subtraction gain function. According to exemplary embodiments, when a far-mouth microphone is used in conjunction with a near-mouth microphone, it is possible to handle non-stationary background noise as long as the noise spectrum can continuously be estimated from a single block of input samples. The far-mouth microphone, in addition to picking up the background noise, also picks up the speaker's voice, albeit at a lower level than the near-mouth microphone. To enhance the noise estimate, a spectral subtraction stage is used to suppress the speech in the far-mouth microphone signal. To be able to enhance the noise estimate, a rough speech estimate is formed with another spectral subtraction stage from the near-mouth signal. Finally, a third spectral subtraction function is used to enhance the near-mouth signal by suppressing the background noise using the enhanced background noise estimate.

Non-linear image filter for filtering noise

Abstract: A system and method for reducing noise using recursive noise level estimation. The system and method for noise reduction substitute a target pixel in a processing window with a weighted average of a plurality of neighboring pixels according to the degree of similarity between the target pixel and the neighboring pixels. The similarity is based on the noise level affecting the image and the local brightness of the processing window. The filter is based on fuzzy logic and filters out noise without smoothing the image's fine details. The filter uses a human visual system (HVS) response to adjust brightness.

An algorithm for integrated noise reduction and sharpness enhancement

Abstract: Noise reduction and image sharpening are techniques to improve video image quality. However, noise filters tend to blur image detail, while filters for image sharpening tend to increase noise. So, cascading the two filters does not always give the best performance. We present an integrated filter that reduces noise and sharpens details in a noisy video signal depending on local image statistics. This allows both features to be maximally exploited.

Bit-depth extension of digital displays using noise

Abstract: A method for extending bit-depth of display systems. The method includes the steps of measuring the static display noise of a display device (14), using the display noise to create pseudo-random noise (12) and subtracting the pseudo-random noise (12) from a contone image (10). After the noise-compensated image data is quantized and displayed, the noise in the display device (14) will substantially convert the noise-compensated image data back to contone image data with few or no contouring artifacts. Other embodiments include using the inherent noise of the human visual system (22) instead of the static display noise, or both. Specific adjustments can be made to the noise of the human visual system (22) for color displays.

Noise reduction method, apparatus, and program for digital image processing

Abstract: A method of processing a digital image channel to remove noise, includes the steps of: identifying a pixel of interest; calculating a noise reduced pixel value from a single weighted average of the pixels in a sparsely sampled local region including the pixel of interest; replacing the original value of the pixel of interest with the noise reduced pixel value; and repeating these operations for all of the pixels in the digital image channel.

Method for enhancing a digital image with noise-dependent control of texture

Abstract: A method is described for enhancing a digital image channel, e.g., a channel comprising a texture signal, by providing a predetermined estimate of the noise expected for the digital channel based on a predetermined relationship between the image intensity values and the expected noise for given intensities. After a local estimate of signal activity is generated for the digital image channel, a gain adjustment is generated from the predetermined estimate of noise and the local estimate of signal activity; the gain adjustment is applied to the image pixels in the digital channel in order to generate a digital channel with enhanced image values.

Electronic still camera capable of removing noise component in image data and signal processing method thereof

Abstract: Upon depression of a shutter key of a key input unit, a sensed image (DATA 1 ) is captured by exposing a CCD for exposure time T 1 with a mechanical shutter opened, and a dark frame image (DATA 2 ) is captured by exposing the CCD for exposure time Ta with the mechanical shutter closed. A correction value corresponding to exposure time T 1 is then determined by looking up a correction data table stored in a storage area of a data memory, and DATA 2 is corrected using the determined correction value. By subtracting the corrected DATA 2 from DATA 1 , a dark output component contained in DATA 1 is removed.

Noise suppressing processor and recording medium

Abstract: PROBLEM TO BE SOLVED: To provide a noise suppressing processor for suppressing noise components contained in a radiographic image, capable of effectively suppressing noises, irrespective of the radiation dose. SOLUTION: The noise suppressing processor 100 comprises a band-limiting image signal generating part 1 for generating band limiting image signals Bk for supporting an image having different frequency bands, an index value acquiring part 2 for finding a picture element vector for a target pixel for the band limiting image and detecting an edge direction as noise characteristics using the pixel vector, a noise suppressing processing part 3 for changing the characteristics of a smoothing filter so as to render smoothing processing along the detected edge direction and smoothing the individual band-limiting image singles Bk using the smoothing filter, and an image reforming part 4 for reforming processed image signals Sproc for supporting the image having noises suppressed, according to the band-limiting image signals fBk having noise components suppressed.

Noise measurement in video images

Abstract: Many image enhancement algorithms easily fail in the presence of noise. Noise adaptive algorithms are much more robust, and a good and accurate enough algorithm for noise measurement is a key component. The proposed algorithms combine good performance with low complexity.

Device and method for processing image signal

Abstract: PROBLEM TO BE SOLVED: To provide a device and method for processing an image signal that can appropriately reduce noise according to change in shooting conditions and levels for achieving an improved image. SOLUTION: A digital camera 10 allows the noise of each pixel of image data 32a that is supplied at a signal-processing part 34 to be detected, generates a threshold to image data 44 whose noise is to be detected at a threshold generation part, obtains the average value of the pixel of a periphery to the target pixel at a noise reduction processing part 34d, relatively shows the difference between the target pixel and the periphery as level difference, judges the level difference with the threshold as a reference, considers that many noise constituents are contained and use the average value when the level difference is small, judges that signal constituents are larger than the noise ones when the level difference is at least at the threshold for changing according to judgment using the pixel level of the object, and substitutes pixel data for inhibiting the noise of the image.

Image error diffusion device with noise superposition

Abstract: Disadvantages of conventional image processing devices using a vector error diffusion method include the generation of periodic texture noise of specific selected colors, resulting in marked reduction of graininess. A noise overlay unit is provided as a front stage of an error adding unit for adding error to input image data. If noise is superimposed on the L*a*b* color space or L*C*h color space, quantitative image evaluation matching human perception is possible, and overlay noise can be effectively optimized. If the total sum of the absolute amount of overlay noise is zero, the color tone of the entire image can be maintained.

Dual-Microphone Spectral Subtraction

Abstract: Mobile phones are constantly decreasing in size, thereby complicating the acoustical functionality. Signal processing methods can be used to partially mitigate this problem. In this paper we suggest a method which uses multiple spectral subtraction functions and two microphones, introducing only a short signal delay. The idea is to use spectral subtraction methods to extract the noise as well as the speech during a single time-frame. The environment background noise may not be stationary, thereby limiting the method to only employ short estimates of the background noise signal. Results are presented for experiments in various environments, showing a reduced noise level in the processed signal compared with the un-processed signal, and with preserved speech quality.

Method and device for eliminating noise

Abstract: PROBLEM TO BE SOLVED: To provide a noise elimination method and a noise elimination device, which can aggressively eliminate noise components dominating in low frequency band, make the value of S/N positive in dB and extract a relatively high level received signal embedded in the noises, by paying attention to macroscopic colored noise. SOLUTION: A signal, with its time base, amplitude, and phase are specified, or a zero point signal is inserted into a transmitted signal at an insertion part 1. Noise components within a received signal are interpolated and predicted by a noise elimination part 2 by using the specified signal or the zero point signal. The transmitted original signal is reproduced by subtracting the noise components from the received signal.

Bit-depth extension in a digital display using quantization of noise-compensated image data

Abstract: A method for extending bit-depth of display systems. The method includes the steps of measuring the static display noise of a display device (14), using the display noise to create pseudo-random noise (12) and subtracting the pseudo-random noise (12) from a contone image (10). After the noise-compensated image data is quantized and displayed, the noise in the display device (14) will substantially convert the noise-compensated image data back to contone image data with few or no contouring artifacts. Other embodiments include using the inherent noise of the human visual system (22) instead of the static display noise, or both. Specific adjustments can be made to the noise of the human visual system (22) for color displays.

Simulation of the image quality of an a-Si flat x-ray detector system in low-dose fluoroscopic applications

Abstract: One of the issues in (alpha) -Si:H X-ray detectors is signal to noise ratio for low dose fluoroscopic applications. An optimized sensitivity of the X-ray detection system together with low and isotropic system noise characteristics are primary pre-conditions needed for maximum image quality. However, in spite of high DQE numbers of this Flat Detector technology in radiological and fluoroscopic application areas, a SNR for low dose fluoroscopy is found, which is inferior to that found with Image Intensifier-TV based systems. The problem area is a small dose range, producing gray levels just above absolute dark. Except for the dark level, these levels can (dependent on the application area) contain clinically relevant information. Since scatter affects the darker parts of the relevant image areas there will be noise in those areas, caused by X-ray quantum statistics and readout noise. The objective of the simulations is to investigate the influence of the various system noise components on the image quality. A level of system noise can be found where the subjective image quality is mainly determined by the X-ray quantum statistics and where the readout noise does not necessarily have to be invisible in totally dark parts. The simulation concerns a threshold contrast detail detectability (TCDD) observation test, where observers score discs of various diameter and absorption in an image sequence (being a fixed scene of the test object with (temporal) X-ray noise and system noise). The dynamic sequence is based upon total simulation, i.e. the test object as well as the X-ray noise and the system noise components were simulated. To verify the simulations also an image sequence was acquired on a Flat Detector system. The observations are done at various dose levels, with and without post processing to obtain noise reduction like it is used in clinical practice for this kind of system. The sequences are observed on a medical CRT display.

Noise correction method for digital image sensor

Abstract: The method involves determining specific noise values for each line and column of the image sensor pixel matrix. Noise correction values are provided for the individual pixel image signal values. The pixel image signal values for each line or column of the matrix are combined with at least two different noise correction values, with the number of corrected pixel signal values dependent on the noise value.

Bit-depth extension of digital displays by quantizing noise-compensated image data

Abstract: A method for extending bit-depth of display systems. The method includes the steps of measuring the static display noise of a display device (14), using the display noise to create pseudo-random noise (12) and subtracting the pseudo-random noise (12) from a contone image (10). After the noise-compensated image data is quantized and displayed, the noise in the display device (14) will substantially convert the noise-compensated image data back to contone image data with few or no contouring artifacts. Other embodiments include using the inherent noise of the human visual system (22) instead of the static display noise, or both. Specific adjustments can be made to the noise of the human visual system (22) for color displays.

Noise detection method, noise detector, image data processing unit, and recording medium

Abstract: PROBLEM TO BE SOLVED: To provide a noise detection method that can detect a noise produced part, such as a mosquito noise included in decoded image data resulting from decoding image data generated, through frequency conversion and irreversible compression coding. SOLUTION: A motion in a temporal direction by each block consisting of a prescribed number of pixels of input image data is detected. The dispersion in the motion by each block is detected on the basis of the temporal directional motion by each detected block. The noise produced part is detected, on the basis of the dispersion of the motion detected in the motion dispersion detection step.

Propagation of Noise Statistics in Digital Photofinishing Image Processing

Abstract: A highly influential factor in the performance of image processing algorithms is the amount of noise present in the digital image. A priori knowledge of the expected levels of noise in the image dramatically improves the performance and efficiency in image processing routines. In a digital photofinishing system, image noise is primarily attributed to film grain and scanner noise. Therefore, if the film and scanner sources are known, it is possible to deduce the expected noise level in a digital image. However, image processing applied to the image after scanning will affect the noise statistics. In order for the image processing algorithms to deliver optimum performance, the estimated noise statistics need to be modified according to each processing step applied. We consider image-processing operations as applying transformations to the image data, and corresponding ones to the image noise statistics. We will discuss analytic equations that approximate the propagation of image noise statistics through several basic image transformations, and their interaction with algorithms in a digital photofinishing image chain.

Restoration of impulse noise corrupted images based on gradients

Abstract: A new filtering algorithm is presented which can remove impulse noise from corrupted images while preserving details. The algorithm is based on a new impulse detection technique that uses image gradients. The proposed impulse detector can effectively categorize all the pixels in an image into two classes -- noise pixels and noise-free pixels. The noise-free pixels are kept untouched while the noise pixels are filtered by a noise cancellor such as median filter. Experimental results show that the proposed algorithm provides significant improvement over many existing techniques in terms of both subjective and objective evaluations. It also has the advantage of computational simplicity over those algorithms.

Picture processing apparatus, picture processing method, learning apparatus, and learning method

Abstract: An input picture signal having noise added through a transmission path or the like is supplied. Noise is removed from the input picture by a class categorizing adaptive process in which predictive coefficients are pre-learnt and decided for each class. A class corresponding to a noise component contained in the input picture is decided. Predictive coefficients of the class and the values of pixels of frames containing a considered pixel of the input picture signal are linearly combined. Thus, predictive pixel values are generated, which are free of noise. When motion of a considered pixel is detected and pixels that are used to decide a class and pixels that are used for a predictive calculation are compensated corresponding to the motion, noise accurately corresponding to a noise component can be removed from the input picture signal.

Noise Detection-Based Color Filters for Impulse Noise Removal

Abstract: In this paper,new filters for impulse noise removal in color image are proposed based on noise detection, called as detail-preserving filters(DPF).In the new method, impulse noise in color image is first preperly detected,and the detected noisy pixels are filtered out with adaptive selection of filter window,while the noise-free pixels are kept without any change.Experimental results show that the proposed detail-preserving filters remove impulse noise and preserve edges and details very well,that perform better than the classical vector median filter (VMF),direction-directance filter(DDF), and distance-magnitude vector filter(DMVF).

Bit-depth extension of digital displays by using psendo-random noise

Abstract: A method for extending bit-depth of display systems. The method includes the steps of measuring the static display noise of a display device (14), using the display noise to create pseudo-random noise (12) and subtracting the pseudo-random noise (12) from a contone image (10). After the noise-compensated image data is quantized and displayed, the noise in the display device (14) will substantially convert the noise-compensated image data back to contone image data with few or no contouring artifacts. Other embodiments include using the inherent noise of the human visual system (22) instead of the static display noise, or both. Specific adjustments can be made to the noise of the human visual system (22) for color displays.

Bit-depth extension in a digital display using noise-compensated image data

Abstract: A method for extending bit-depth of display systems. The method includes the steps of measuring the static display noise of a display device (14), using the display noise to create pseudo-random noise (12) and subtracting the pseudo-random noise (12) from a contone image (10). After the noise-compensated image data is quantized and displayed, the noise in the display device (14) will substantially convert the noise-compensated image data back to contone image data with few or no contouring artifacts. Other embodiments include using the inherent noise of the human visual system (22) instead of the static display noise, or both. Specific adjustments can be made to the noise of the human visual system (22) for color displays.

Fusing 3D shapes at different view points to obtain more accurate shape

Abstract: We obtain an exact 3D model of a scene using multiple images from different camera positions. The recovery of a 3D shape using two images has problems such as being weak in noise. Therefore we present methods to improve the accuracy of 3D shapes with multiple images. The system is divided into three stages: (1) the recovery of 3D shapes from different viewpoints, (2) fusing the 3D shapes to obtain a para-ideal shape, and (3) removing the outlier shapes and feature points by an evaluation function, and fusing the rest of the shapes. We demonstrate a significant improvement in the recovered shape. Experimental results show that our system performs well in removing noise with robustness. The maximum noise reduction rate was 82% in real image experiments.