37results about How to "Achieve high-precision reconstruction" patented technology

The invention provides a space-spectrum joint sparse prior based satellitic hyperspectral compressed sensing reconstruction method. The method includes hyperspectral data-block diagonal random measurement of an onboard encoding end and compressed sensing construction of a ground decoding end. The onboard encoding-end adopts block-diagonal hyperspectral-data random measurement matrixes to perform independent random sampling on each spectral band so as to acquire measurement data and then sends the measurement data to the ground decoding end through a data link, the ground decoding end decomposes the data into low-rank components and sparse components, low-rand prior between the hyperspectral-data spectral bands and sparse prior in the spectral bands are jointed to establish a convex optimization reconstruction model, iteration solution is performed to obtain low-rank components and sparse components of reconstructed hyperspectral data, and the low-rank components and the sparse components are merged to obtain the reconstructed hyperspectral data. By the method, precision and efficiency of satellitic hyperspectral-image compressed sensing reconstruction are improved.

The invention belongs to the technical field of optical precision testing, and relates to a dual-wavelength phase-shift interference method based on a partial compensation method and an implementationdevice. The method comprises the steps of building a partial compensation method dual-wavelength phase-shift interferometer, and acquiring measured wavefront wrapped phases of two single wavelengths;modeling a partial compensation method dual-wavelength ideal interferometer, and acquiring residual wavefronts and wrapped phases of the two single wavelengths; eliminating known and unknown wavefront variations in the measured wavefront wrapped phases by adopting an error separation and elimination algorithm, and finally optimizing and reconstructing surface-shape error of the measured asphericsurface by adopting reverse iteration. The device comprises a first laser, a second laser, a first slit, a second slit, a first plane mirror, a second plane mirror, a first beam splitter, a second beam splitter, a beam expander, a collimating mirror, a standard plane mirror, a phase shifter, a partial compensating mirror, a measured aspheric surface, an imaging lens and an interferogram acquisition assembly containing a sparse array sensor. The method and device provided by the invention are particularly applicable to processing quality measurement for gradient aspheric surfaces with a small surface-shape error, molded aspheric surfaces with a great surface-shape error and free curved surfaces.

A three-dimensional temperature reconstruction combination method based on a flame light field refocus image is disclosed. The invention relates to the three-dimensional temperature reconstruction combination method based on the flame light field refocusing image. By using an existing nearest neighbor method, the physical reconstruction precision of the outermost layer and secondary outer layer offlames is low, and when a Lucy-Richardson deconvolution method is used in intermediate layer flame physical reconstruction, the precision is low. In the invention, the above problems are solved. Themethod has the following steps of 1, using a light field camera to shoot the flame and record the three-dimensional light field image of the flame; 2, acquiring the light field refocusing image of theflame; 3, acquiring the point spread function of the light field camera; 4, acquiring a flame chromatography image; and 5, acquiring a fitting relationship between a black body plane surface radiation force and a gray scale, and reconstructing the radiation force for the gray scale of the flame chromatography image based on the fitting relationship, and according to the radiation force, acquiringa three-dimensional flame temperature. The method is used for the flame image processing field during a high temperature flame temperature reconstruction process.

The invention discloses a hyperspectral image compressed sensing method based on the heavy weighting laplacian sparse prior. The hyperspectral image compressed sensing method based on the heavy weighting laplacian sparse prior is used for solving the technical problem that an existing hyperspectral image compressed sensing method is low in reconstruction accuracy. According to the technical scheme, a little linear observation of each pixel spectrum is collected randomly as compressed data, a compressed sensing model based on the heavy weighting laplacian sparse prior and a sparse regulated regression model is established, and solving on the established models is conducted. According to the hyperspectral image compressed sensing method based on the heavy weighting laplacian sparse prior, and a little linear observation of each pixel spectrum is collected randomly as compressed data, so that resource consumption in the image collecting process is reduced; the strong sparsity of the hyperspectral image is depicted accurately through the heavy weighting laplacian sparse prior, inhomogeneous constraint on the nonzero element of the traditional laplacian sparse prior is overcome, and the reconstruction accuracy of the hyperspectral image is improved. It is tested that when a sampling rate is 0.15 and the compressed data consist strong noise with 10 db signal-to-noise ratio, and the peak signal-to-noise ratio promotes over 4 db relative to a background technology method.

The invention discloses a hyperspectral image compressive sensing method based on nonseparable sparse prior. The hyperspectral image compressive sensing method based on nonseparable sparse prior is used for solving the technical problem that existing hyperspectral image compressive sensing methods are low in reconstruction precision. According to the technical scheme, a few of linear observed values of each pixel spectrum are collected and serve as compressed data, and the resource demand in the image collection process is reduced while substantial data compression is achieved. In the reconstruction process, empirical Bayesian reasoning is utilized to construct nonseparable sparse prior of sparse signals, potential correlation among nonzero elements in the sparse signals is taken into full consideration, and high-precision reconstruction of hyperspectral images is achieved. Because a wavelet orthogonal basis serves as a dictionary according to the method, dependency on end members is eliminated. In addition, through reasoning based on a Bayesian framework, full-automatic estimation of all unknown parameters is achieved, human adjustment is not needed, and adaptability is wide. Experiments show that when the sampling rate is 0.1, the peak signal to noise ratio obtained according to the hyperspectral image compressive sensing method is increased by above 4 db compared with that obtained according to a background technology compressive sensing method.

The invention discloses a high spectral image compression sensing method based on manifold structuring sparse prior and solves a technical problem of low precision existing in a high spectral image compression sensing method in the prior art. The method is characterized in that a few linear observation values of each pixel spectrum are sampled randomly and are taken as compression data, through the manifold structuring sparse prior, sparsity of a high spectral image after sparsification in the spectrum dimension and manifold structure of the high spectral image in the space dimension are etched, through a hidden variable Bayes model, signal reconstruction is carried out, and sparse prior learning and noise estimation are unified to one regularization regression model for optimization solution. The sparse prior acquired through learning can not only fully describe the three-dimensional structure of the high spectral image, but also has relatively strong noise robustness. The sparse prior is utilized to realize high precision reconstruction of the high spectral image. Based on tests, Gauss white noise is added to the compression data to make the signal to noise ratio of the compression data to be 15db, the sampling rate is 0.09, and thereby the 23db peak value signal to noise ratio is acquired.

The invention discloses a high-precision high-order time rearrangement synchronous extrusion transformation time-frequency analysis method. The method comprises the steps: converting a signal x (t) into a frequency domain signal X (omega); selecting an order number N and a frequency window function G (omega), calculating short-time Fourier transform of the X (omega) under omega<k>G (omega), constructing square matrixes alpha N (t, omega) and beta N (t, omega), and calculating N-order group delay estimation; superposing a time-frequency value to a GD estimation position along a time direction,and calculating N-order time rearrangement synchronous extrusion transform to obtain a time-frequency spectrum; and reconstructing the signal. According to the invention, a novel GD estimation methodis provided, the estimation precision of the rapidly changing GD is improved; and the time-frequency spectrum energy is rearranged in the time direction to be gathered near the real GD, so that the energy diffusion is effectively inhibited, and the time-frequency readability is improved. The invention further provides a reconstruction method, and the reconstruction signal is high in reduction degree.

The invention discloses a revolution surface sound field reconstruction method based on a near-field acoustic holography technology. Sound pressure information measured on the cylindrical holographicsurface is utilized; a statistical optimal cylindrical surface near-field acoustic holography technology is applied to reduce the reconstruction distance, so that the reconstruction precision is improved; the limitation that only a regular cylindrical sound field can be reconstructed by adopting a statistical optimal method is overcome by applying a spherical wave superposition near-field acousticholography technology; the reconstruction of the radiation sound field of the shell structure with any revolution surface appearance based on the cylindrical measurement surface is realized; a plurality of spherical wave sources are placed to replace a single spherical wave source, the limitation of the method on the sound field appearance is overcome; negative effects of errors in high-order spherical waves on reconstruction precision and stability are effectively inhibited; the effectiveness of the method is fully displayed through numerical simulation, the radiation sound field of any point of the revolution surface appearance shell structure in the three-dimensional space can be reconstructed, visual display is achieved, and therefore the sound radiation characteristics and the soundfield bright spot positions of different areas of the sound field outside the structure can be visually obtained.

The invention discloses an automobile morphology active vision detection system and method based on an unconstrained homocentric beam family, and aims at solving the problem of automobile morphology active vision detection based on the unconstrained homocentric beam family. The automobile morphology active vision detection system based on the unconstrained homocentric beam family is mainly composed of a camera (1), a three-dimensional target (2), a laser line projector (3), a two-dimensional target (4), a two-dimensional target base (5), a connecting piece (6) and a tripod (7). The automobilemorphology active vision detection method based on the unconstrained homocentric beam family comprises the four steps of image acquisition, calibration of a position relation between the two-dimensional target (4) and the laser line projector (3), laser line reconstruction in a coordinate system of the camera (1), and automobile characteristic point reconstruction in the coordinate system of the camera (1). The automobile morphology active vision detection system and method based on the unconstrained homocentric beam family, which is simple in structure and reliable in performance is provided.

The invention discloses a shaft end grounding binocular reconstruction system and method for concentric quadratic curve epipolar geometry, and aims to solve the problem of shaft end grounding binocular detection for concentric quadratic curve epipolar geometry. The concentric quadratic curve epipolar geometry shaft end grounding binocular reconstruction system is mainly composed of a base (1), a supporting rod (2), a laser connecting rod (3), a left camera (4), a laser (5), a right camera (6), a camera connecting block (7), a target plate (8), a linear sliding table (9), a connecting rod (11),a supporting rod fixing block (12), a connecting rod fixing block (13) and a laser fixing block (14). The target plate (8) is a flat plate of which the surface is pasted with a concentric quadratic curve; the supporting rod fixing block (12) is a cuboid iron block, two circular through holes with the axes perpendicular to each other are machined in a surface of the supporting rod fixing block (12), and threaded holes are machined in two ends of the supporting rod fixing block (12) respectively. The concentric quadratic curve epipolar geometry shaft end grounding binocular reconstruction system and method are simple in structure and reliable in performance.

The invention discloses an image restoration method and system based on equality structure multiple regularization, and the method comprises the steps: introducing an equality structure, and constructing a measurement Gaussian likelihood function; selecting a plurality of sparse transformations based on the constructed measurement Gaussian likelihood function, and giving a plurality of mutually independent probability density functions based on sparse transformation coefficients to calculate the conditional prior probability density of the target image; based on the condition prior probability density of the target image, calculating a mean value estimation model of the target image according to a Bayesian rule; and finally, achieving fast reconstruction of the target image by combination of an expectation maximization method and a conjugate gradient method iteration so that the reversibility limitation of sparse representation introduced by a comprehensive method is overcome, sparse transformation of any number is allowed to be adopted, and the convergence speed of the algorithm can be effectively improved through the joint sparse domain.

The invention discloses a shaft end grounding monocular reconstruction system and method for concentric quadratic curve epipolar geometry, and aims to solve the problem of the shaft end grounding monocular detection for concentric quadratic curve epipolar geometry. The concentric quadratic curve epipolar geometry shaft end grounding monocular reconstruction system is mainly composed of a base (1),a supporting rod (2), a camera connecting rod (3), a camera (4), a laser (5), a laser fixing block (6), a supporting rod fixing block (7), a two-dimensional target plate (8), a camera fixing block (9), a linear sliding table (10) and a laser connecting rod (12). The two-dimensional target plate (8) is a flat plate with a concentric quadratic curve pasted on the surface, and the supporting rod fixing block (7) is a cuboid iron block provided with two circular through holes with the axes perpendicular to each other and provided with circular through holes in the two ends respectively. The shaftend grounding monocular reconstruction system and method for concentric quadratic curve epipolar geometry are simple in structure and reliable in performance.

The invention provides a high-precision step surface shape measurement method based on spectral confocal. The method comprises the following steps: irradiating a line spectrum by using a scanning probe; dividing the upper surface of the measurement object into a plurality of unit areas with corresponding widths according to the scanning width of the scanning line spectrum; enabling the line spectrum to move from the section of one end of the unit area to the section of the other end in the direction perpendicular to the scanning motion direction of the three-dimensional motion platform for primary scanning, rotating the scanning probe and the line spectrum by an angle theta and moved to the initial position, and enabling the scanning probe to move from the section of one end of the unit area to the section of the other end for secondary scanning; collecting measurement data of primary scanning and measurement data of secondary scanning; calibrating the measurement data of the primary scanning of the same unit area by using the measurement data of the secondary scanning of the same unit area; and reconstructing the three-dimensional shape of the upper surface of the measured object by using the calibrated measurement data subjected to preliminary scanning. The invention further provides a high-precision step surface shape measuring device based on spectral confocal.

The invention discloses a global computing imaging method based on a focusing stack monomer data subset architecture. The method comprises the following steps: 1, finding out a scene image with the maximum focusing measure from scene image focusing stack data; 2, dividing area images corresponding to all single areas in the scene image; 3, finding out a single body region on the single body regionimage of the scene image with the maximum focusing measure; 4, finding out the focusing stack data of the jth monomer by utilizing the monomer region; 5, selecting a representative region in the monomer region, and screening out V scene images from the I scene images of the focusing stack data of the jth monomer to obtain a vth image of the jth monomer and a focusing stack monomer data subset; 6,carrying out depth reconstruction and full-focus imaging on the single body; 7, optimizing local consistency; 8, carrying out global fusion. According to the method, the calculation efficiency of focus stack reconstruction can be improved, and high-precision reconstruction of the three-dimensional scene monomer is realized.