181 results about "Reference noise" patented technology

A new technique for identifying from images a camera, or other imaging device such as a scanner, is based on the device's reference noise pattern, a unique stochastic characteristic of all common digital imaging sensors, including CCD, CMOS (Foveon™ X3), and JFET. First, one determines from images the sensor's reference pattern noise, which uniquely identifies each sensor. To identify the sensor from a given image, the presence or absence of the reference pattern noise in the image under examination is established using a correlation detector or other means.

In the frequency domain, the method comprises the following steps:

• • a) calculating a first noise reference by analyzing the spatial coherence of the signals picked up;
  • b) calculating a second noise reference by analyzing the directions of incidence of the signals picked up;
  • c) estimating a main direction of incidence of the signals picked up;
  • d) selecting as a referent noise signal one or the other of the noise references as a function of the estimated main direction;
  • e) combining the signals picked up into a noisy combined signal;
  • f) calculating a probability that speech is absent in the noisy combined signal on the basis of the respective spectral energy levels of the noisy combined signal and of the referent noise signal; and
  • g) selectively reducing the noise by applying variable gain that is specific to each frequency band and to each time frame.

The invention provides a voice noise reduction method and a voice collection device. According to the voice noise reduction method, frequency-domain transform and filtering are carried out through a filter assembly on a noise-cancelled voice signal formed through filtering and a reference noise signal formed through filtering, an original noise signal corresponding to the reference noise signal and a noise-containing voice signal corresponding to the noise-cancelled voice signal are output in a filtered manner, and a first crosstalk-resistant filter unit forming the reference noise signal and a second crosstalk-resistant filter unit forming the noise-cancelled voice signal update filter coefficients under control. The technical scheme has the following advantages: by employing only one pair of adaptive filters and using a post-filter to control the updating of the filter coefficients of the one pair of adaptive filters, the structure is simplified; and by adjusting the filter coefficient updating weights of the voice signal and the noise signal in a targeted manner, excellent crosstalk-resistant and noise reduction effects are achieved.

The invention discloses an active noise reduction system and a method. The active noise reduction system comprises a sound acquisition device, a motion signal acquisition device, a controller and a sound playing device, wherein the sound acquisition device is used for acquiring error noise signals in the environment; the motion signal acquisition device is used for acquiring motion signals at a specific position on a device generating primary noise signals and related to the primary noise signals; the controller is used for using a support vector machine to generate reference noise signals according to the motion signals, adjust the reference noise signals according to the error noise signals and then generate secondary noise signals; and the sound playing device is used for playing the secondary noise signals in the environment. The error noise signals are generated by overlapping the primary noise signals and the secondary noise signals. In the above active noise reduction system and the method, the support vector machine is used for predicting noise in the environment in real time. The generated reference noise signals have real-time performance and accuracy, and the noise reduction effects are improved significantly.

A micropower neural amplifier with adaptive power biasing for use in multi-electrode arrays is provided. The micropower neural amplifier includes a low noise gain stage. The low noise gain stage is implemented using an amplifier and pseudoresistor elements.

A new technique for identifying whether images are derived from a common imager, e.g., a camera, or other imaging device such as a scanner, based on the device's measured or inferred reference noise pattern, a unique stochastic characteristic of all common digital imaging sensors, including CCD, CMOS (Foveon™ X3), and JFET. The measured or inferred noise pattern of two images may be extracted and then cross correlated, with a high correlation being consistent with a common imager. Various preprocessing techniques may be used to improve tolerance to various types of image transform. It is also possible to perform the analysis without explicit separation of inferred image and inferred noise.

Acquired video signals are forwarded to pre-processing block (103) for sampling, gain amplification and A/D conversion, and then forwarded to image buffer (104). Video signals in image buffer (104) are forwarded to noise estimation block (106). On the basis of taking conditions, video signals and a reference noise model, noise estimation block (106) works out the quantity of noise per ISO sensitivity, and per color signal. The calculated noise quantity is forwarded to noise reduction block (105). On the basis of the noise quantity estimated at the noise estimation block (106), the noise reduction block (105) applies noise reduction processing to video signals in image buffer (104), and video signals after noise reduction processing are forwarded to signal processing block (108).

This invention provides a method for determining, in a speech dialogue system issuing speech prompts, a score value as an indicator for the presence of a wanted signal component in an input signal stemming from a microphone, comprising the steps of: using a first likelihood function to determine a first likelihood value for the presence of the wanted signal component in the input signal, using a second likelihood function to determine a second likelihood value for the presence of a noise signal component in the input signal, and determining a score value based on the first and the second likelihood values, wherein the first likelihood function is based on a predetermined reference wanted signal, and the second likelihood function is based on a predetermined reference noise signal.

A music piece reproducing apparatus has: a sound collecting unit collecting noise data of an ambient environment; an analyzing unit performing a frequency analysis to the noise data collected by the sound collecting unit and extracting a feature amount showing a feature of the noise data; a noise parameter database classifying the noise data every environment; and a reproducing unit reproducing music piece data which has been recorded in a recording medium and classified every category, wherein the analyzing unit classifies the noise data with reference to the noise parameter database on the basis of the extracted feature amount of the noise data, selects the previously-corresponded category of music pieces in accordance with a classification result of the noise data, and selects predetermined music piece data from music piece data classified into the selected category.

An apparatus and method for allocating a data rate in a multi-antenna system are provided, wherein a noise threshold corresponding to a modulation scheme, a coding rate, or a combination of both applied to a base layer from among a plurality of layers is set as a reference noise threshold, and a data rate is allocated to an enhanced layer among the plurality of layers, based on the reference noise threshold and a use or non-use of hierarchical transmission.

The invention discloses an automobile active noise reduction system and method. The automobile active noise reduction method comprises error microphones, a rotating speed acquisition device and a noise reduction controller. The error microphones are connected with the noise reduction controller and are used for collecting noise signals and outputting the noise signals to the noise reduction controller. The rotate speed acquisition device and the noise reduction controller are used for acquiring rotating speed signals of an engine and outputting the rotating speed signals to the noise reductioncontroller. The noise reduction controller is used for performing active noise reduction simulation according to the noise signals, the rotating speed signals of the engine, reference noise data of acorresponding vehicle type and an established transfer function, optimal control parameters are determined, and active noise reduction is performed based on the optimal control parameters. The automobile active noise reduction system and method can control the order noise of the engine, and can freely adjust a control area in a vehicle according to the number and position of installed error microphones. The automobile active noise reduction system and method can be conveniently and quickly additionally installed into the vehicle needing active noise reduction, the cost is low, the operation is simple, and drivers and passengers can obtain better driving experience.

The invention discloses a non-contact heart rate measurement method based on a face video image. The method comprises the following steps: acquiring multiple frames of RGB images; searching a face area of the single-frame RGB image and determining a cheek center area as an interested face area ROI1, and performing space pixel averaging on a green channel of the ROI1 to obtain an average value s1i;extracting a part of fixed background area of the single-frame RGB image as an interested background area ROI2, and performing space pixel averaging on a green channel of the ROI2 to obtain an average value s2i; integrating the plurality of s1i to obtain an effective signal flow s1, and integrating the plurality of s2i to obtain a reference noise signal flow s2; performing normalization processing on the effective signal flow s1 and the reference noise signal flow s2, and then performing noise reduction to obtain a reconstructed signal flow sign; and calculating a heart rate value according to the reconstructed signal flow sign. According to the invention, the problems of low face motion and signal-to-noise ratio and the like in heart rate monitoring based on the face video image can be effectively solved.

System and method for improving a digital PLL's performance by making fine grained adjustments to the loop gain. A preferred embodiment comprises a plurality of loop gain adjustors (such as loop gain adjustors 605, 606, 607, and 608) that can incrementally adjust the loop gain. The incrementally adjusted loop gains are sequentially brought on-line so that the loop gain of the digital PLL is slowly decreased. By slowly decreasing the loop gain, the digital PLL is less perturbed by smaller noise transients that would take time to settle. Hence, the digital PLL can quickly acquire a signal and then decrease its loop gain and hence its bandwidth when it only needs to track a signal. The reduced bandwidth also reduces the overall noise in the digital PLL that is due to the reference noise contribution.

The invention discloses an active noise control method and system based on a fuzzy neural network and an armored vehicle driver's helmet. The method includes collecting a reference noise signal in a reference area by a reference microphone, and outputting it to a fuzzy controller as a reference signal; the error microphone collects noise The residual noise signal in the control area is used as an error signal, and is output to a fuzzy controller; and the fuzzy controller is based on an adaptive FX-RBF network training algorithm of a fuzzy neural network, and analyzes the reference signal and the error signal , and output the anti-phase target sound signal to the speaker. The active noise control based on the fuzzy neural network of the present invention has obvious noise reduction effect on noise below 2000 Hz, and the noise reduction effect on low frequency noise below 1000 Hz is particularly significant.

A traffic measurement device comprises a noise power setting unit configured to set a reference noise power, which is a reference value used to measure a traffic volume at a base station, a noise power measurement unit configured to measure a noise power at the base station, and a calculation unit configured to calculate the traffic volume based on the reference noise power and the measured noise power. Another traffic measurement device comprises a transmission power setting unit configured to set a reference transmission power, which is a reference value used to measure the traffic volume at a base station, a transmission power measurement unit configured to measure a transmission power at the base station, and a calculation unit configured to calculate the traffic volume based on the reference transmission power and the measured transmission power.

The invention relates to a system and a method for reducing background noise in magnetoencephalogram detection. A magnetic sensor array layer is arranged at the outer side of a head-mounted flexible substrate; a brain magnetic field sensor array and an outer magnetic field sensor array are distributed on left and right hemispheres of the brain; a magnetic shielding layer is arranged outside the magnetic sensor array layer; and a neck ring of a wearable magnetic field sensor is arranged on the neck of a human body. reduction of the background noise in magnetoencephalogram detection is achieved by collecting and analyzing a brain magnetic field of the head and an outer magnetic field of the head, analyzing the brain magnetic field of the neck and the outer magnetic field of the neck, fusing brain magnetic field information, carrying out noise analysis on the outer magnetic field information, obtaining a reference noise background signal, fusing the brain magnetic field information and subtracting the reference noise background signal, and a magnetoencephalogram with a high signal-noise ratio is obtained. The system and the method have the characteristics of being simple, convenient to implement, high in sensitivity, low in implementation cost, high in reliability, small in space and high in flexibility.