92 results about "Noise map" patented technology

Frame-by-frame estimation of signal-dependent noise variance involves morphological deletion of pixel values of noise image corresponding with edges of the initial frame. A noise map is computed as a pixel-by-pixel estimation of noise variance of the initial digital image. The noise reduction procedure involves three successive stages: frame-by-frame estimation of signal-dependent noise variance; motion estimation and compensation, flicker-effect estimation and compensation; recursive averaging providing motion compensation artifacts correction.

The invention relates to the technical field of transient electromagnetic signal processing in geophysical exploration, and discloses a transient electromagnetic signal-to-noise separation method based on a variational mode decomposition principle, which is suitable for electromagnetic data filtering in a time domain. The method comprises the following steps: decomposing noisy transient electromagnetic data s(t) into K intrinsic mode functions uk(t); performing Hilbert transform on each intrinsic mode function to obtain a unilateral frequency spectrum; demodulating the frequency spectrum of each intrinsic mode function uk(t) to a corresponding base band according to the unilateral frequency spectrum, calculating a demodulated L2 norm, and calculating the bandwidth of the decomposed intrinsic mode function to obtain a constraint variation model; introducing a secondary penalty factor alpha and a Lagrange multiplication operator lambda, and converting the constraint variation model intoa non-constraint variation form; and calculating an optimal solution in a non-constraint variation form to obtain a series of optimized intrinsic mode components after VMD decomposition. The method solves the problem that transient electromagnetic data collected in geological interpretation is interfered by noise, is simple and convenient, has excellent noise separation and amplitude maintaining performance, can extract noise images, and provides conditions for subsequent noise analysis.

The invention relates to a volume cloud rendering method and device, electronic equipment and a storage medium. The method comprises the steps: drawing at least one layer of grid model outwards from an original grid model of volume cloud according to a vertex normal vector; screening pixel points of the grid model based on the noise threshold corresponding to each layer of the grid model to obtaina drawing model; calculating illumination information corresponding to the drawing model according to the illumination parameters; and rendering the drawing model according to the illumination information to obtain a to-be-displayed volume cloud. According to the technical scheme, the shape of the volume cloud is determined based on the grid model instead of the shape of the noise map, and if theshape of the volume cloud needs to be changed, only the number of additionally drawn layers needs to be set and the noise threshold value of the pixel points needs to be screened, and the specific noise map does not need to be selected in advance; therefore, the number of times of adopting the noise map is reduced, the performance overhead of generating the volume cloud is further reduced, and the volume cloud can smoothly run on the mobile terminal equipment.

Provided in the invention is a method and system for diagnosing a circuit fault based on acoustical measurement. With the method and system, phenomenon of virtual connection and loosening of the circuit can be determined accurately. A noise source is determined based on association and matching of a noise frequency and an electrical signal frequency and a circuit connection unit or computer corresponding to the noise source is further determined; and a noise map of the noise source is monitored to monitor an abnormal loosening situation of the corresponding circuit connection unit. On the basis of the technical scheme, whole-course monitoring of the circuit can be realized based on noise signal measurement; and the cost is low. Besides, a database with abundant resources can be formed based on computer collection and comparison and can be used for determining circuit faults accurate in future. And a self-learning capability of the database is also provided, so that the determination time is substantially saved and the system performance is optimized.

The invention relates to a noise map construction method based on a building surface noise model (BSNM). The noise map construction method comprises the steps of acquiring road streetscape data, evaluating noises of spatial points of buildings on two sides of a road, processing the road streetscape data, rendering a road streetscape data map and constructing a noise map. By means of the acquired road noise, noises of the buildings on two sides of the road are evaluated, the BSNM is constructed, a building noise multi-spectrum map is embedded into the surfaces of the corresponding buildings in the noise map, and a visual BSNM is presented. The noise map construction method has the advantages that by means of the noise map construction method, the map capable of representing urban street noises faithfully is constructed for users, and users can experience street noise pollution really.

The invention discloses a user scenario recognition method based on a mobile terminal noise map. The method includes: using a mobile smart phone to acquire sensor data, and performing coarse-grain partition on a scenario that a user acquires noise decibel according to the character that voice data is susceptible to friction and vibration; and acquiring a test data set through a semi-manual annotation manner, calculating a sensor scenario determination threshold according to the characters of different sensors, and recording the acquired decibel data in different scenarios and performing scenario classification through a mobile terminal scenario recognition algorithm. According to the method, the acquisition scenario that the user acquires the decibel data is calculated out and speculated, the quality of visual data of the noise map and a user incentive system can be improved through analysis on the data recording quality, and then more people are encouraged to take part in an environment monitoring task of a community and even a city.

The invention provides a noise automatic monitoring equipment point laying method based on a noise map. The method comprises the steps of determining the total number of road traffic monitoring pointsthrough urban scale and urban road network density, and calculating the number of noise monitoring points of each type of road according to the total length of each type of road in a city; then basedon the equivalent sound level and road length of each road in the noise map, and selecting a specific monitored road; then using a road cell in the noise map as a reference, calculating the equivalent sound level of the road cell according to a calculating result in the noise map, combining adjacent road cells with a sound level difference in a threshold; and selecting the central point of each cell after combination as an area monitoring point. The method is applied to laying of the noise automatic monitoring equipment points, and settles problems of automatic monitoring equipment point laying plan shortage, no consideration for current sound environment quality in an artificial monitoring point, over broad method in a specification, etc.

An electronic apparatus is disclosed. The electronic apparatus includes a memory storing at least one instruction, and a processor, electrically connected to the memory, configured to, by executing the instruction, obtain, from an input image, a noise map corresponding to the input image; provide the input image to an input layer of a learning network model including a plurality of layers, the learning network model being an artificial intelligence (AI) model that is obtained by learning, through an AI algorithm, a relationship between a plurality of sample images, a respective noise map of each of the plurality of sample images, and an original image corresponding to the plurality of sample images; provide the noise map to at least one intermediate layer among the plurality of layers; and obtain an output image based on a result from providing the input image and the noise map to the learning network model.

The present invention is applicable to the technical field of noise maps and provides a method for drawing a noise map and device, the method for drawing a noise map comprises: obtaining an environmental noise uploaded by a terminal, the environmental noise includes a noise value, a time information, and a position information; when the terminal uploads a picture corresponding to the environmental noise, detecting whether the picture is a valid picture or not; when the picture is a valid picture and the position where the picture is taken lies indoor, identifying the environmental noise as an indoor environmental noise; when the picture doesn't exist, identifying the environmental noise as an outdoor environmental noise, or when the picture is invalid, identifying the environmental noise as an outer door environmental noise, or when the picture is a valid picture and the position where the picture is taken lies outdoor, identifying the environmental noise as the outdoor environmental noise; drawing a noise map according to the indoor environmental noise and the outdoor environmental noise. The present invention can not only improve drawing accuracy of the noise map, but also increase reliability of the noise map.

The invention discloses a method for representing a virtual sound of a traffic noise map. The method is an effective supplement for the present traffic noise map display technology. The method specifically comprises the following steps: acquiring noise audio frequency samples for reference distances of different vehicle models at different driving speeds on a practical road in the manner of audio frequency collection; digitally processing the noise audio frequency samples and storing at a background server side of the noise map; acquiring the position information and the noise data of representing points of the virtual sound in a front-end noise map; judging road sections influencing noise values of the representing points of the virtual sound and confirming vehicle model ratio and driving speed information data of the road sections; utilizing a random principle to perform digital audio sound mixing on the noise samples required by the representing points on the background server and converting a digital audio signal into an audio sound mixing file; and calling for and analyzing the audio sound mixing file by the front-end noise map, thereby realizing the virtual representing of the traffic noise. The method used for representing the virtual sound of the traffic noise map can overcome the defect that the traffic noise of the traditional noise map cannot be synchronously displayed.

The invention discloses an image denoising method based on a deep convolutional network. The method comprises the following steps: establishing an image degradation model, and sampling an image into a noise image; forming a training image pair by using the noise map and the original image block; inputting training data into a deep convolutional network for learning, enabling the deep convolutional network to use a convolutional layer and a non-linear activation layer to expand channels and then pass through a plurality of residual blocks, and reducing the number of the channels to the number of output channels through the convolutional layer and the non-linear activation layer; training by using a training algorithm, and learning a mapping function from a noise image to a noise-free image from the training image pair through an optimization method; and optimizing by using a loss function until the parameters of the deep convolutional network converge. The image priori of the data set is mined through training, the recognition capability of the network on the noise intensity is enhanced by receiving the noise level information for multiple times, and the performance of joint demosaicing and denoising is greatly improved.

A noise estimation apparatus and method to prevent texture from being erroneously determined as noise, is provided. To this end, the noise estimation method includes generating an initial noise map for an input image signal, generating at least one noise map based on a temporal change of an image signal continuing from the image signal, calculating a correlation value for each region between the initial noise map and the noise maps, and determining noise of an image according to a temporal change based on the correlation value.

The invention provides a three-dimensional volume cloud rendering method and system based on a three-dimensional noise map, and the method comprises the steps: S1, constructing a three-dimensional noise texture space, and constructing a three-dimensional volume cloud in the three-dimensional noise texture space; s2, constructing a real three-dimensional space, and mapping the three-dimensional volume cloud to the real three-dimensional space; s3, obtaining the sight line direction of the pixel points, determining the overall density of the three-dimensional volume cloud in the sight line direction, and determining the transparency of the pixel point rendering cloud based on the overall density; and S4, obtaining the position, the illumination intensity and the preset color of the sun, anddetermining the color of the pixel point rendering cloud in combination with the overall density and the sight line direction. According to the three-dimensional volume cloud rendering method based onthe three-dimensional noise map, only two groups of three-dimensional noise texture maps are needed, the storage space is reduced, layer clouds, accumulated clouds and layer accumulated clouds can berendered, and the real lighting effect of the clouds is displayed after rendering.

The invention discloses a noise map drawing method based on mobile perception in the technical field of noise monitoring, and aims to solve the problem that noise data of each corner are difficult tomonitor due to fixed-point layout of noise monitoring stations in the prior art. And when the scattered sparse noise is rendered by using an interpolation method, the technical problem that the influence of the space environment on noise propagation is not considered is solved. The method comprises the following steps of collecting and obtaining noise data in a target monitoring area through mobile noise measurement equipment, the noise data comprise noise value data and longitude and latitude corresponding to the noise value data, and the mobile noise measurement equipment comprises a smart phone; and rendering the noise data by using an inverse distance weighted interpolation method to obtain a noise map of the target monitoring area.

The invention discloses an image noise estimation method based on a deep convolutional neural network. The method is used for solving the technical problem that an existing image noise estimation method is poor in practicability. According to the technical scheme, a loss function is constructed by combining a noise error and a noise level error to train a deep convolutional neural network; the trained deep convolutional neural network obtains a noise map with accurate numerical distribution and accurate statistical mean square error from the noise pollution image; a deep convolutional neural network is adopted; convolutional kernel expansion values are firstly multiplied layer by layer and then symmetrically decreased layer by layer in an equal proportion mode, the extraction capacity of the deep convolutional neural network for random noise is improved, the number of convolutional kernels of each convolutional layer is large, a distribution model and the level of noise used in training are completely random, and the generalization capacity of the deep convolutional neural network for random noise is improved. According to the method, the noise image is extracted from the noise pollution image, the flexibility of modeling and analyzing the noise in subsequent operation is improved, and the practicability is good.

The method includes acquisition of an original image; low-frequency filtering of the original image to obtain an estimated image; a noise image development as a difference between the original and estimated images by morphologic filtering elimination of noise image pixels corresponding to sharp changes in the original image; dividing an intensity range of the estimated image into intervals, wherein each pixel of the estimated image relates to an appropriate interval; accumulating some noise image pixels corresponding to estimated image pixels; calculating interval estimations of noise dispersion using accumulated noise image pixels; improving interval estimations by removing noise pixels according to σ3 criteria, approximating interval estimations of noise dispersion resulting in tabular function of noise vs. signal intensity; calculating the base of estimated image and obtaining tabular function of the dependence of noise on signal intensity the noise map as a pixel-by-pixel noise dispersion estimation of the digital original image.