73 results about "Phase gradient" patented technology

The invention relates to the technical field of detection, in particular to a complex shell detection method and system based on structured light scanning. The method comprises the following steps: calculating a wrapped phase value and a phase modulation amplitude of each pixel point under each frequency, calculating discrete probability distribution containing candidate stripe series and a corresponding probability for each pixel point based on a multi-frequency wrapped phase difference, generating an initial pixel confidence map, and constructing an energy function; the penalty weight of the smooth item is determined according to the phase modulation amplitude and the wrapped phase gradient local variance, obtaining an initial fringe order distribution diagram, updating discrete probability distribution by using an initial pixel confidence map, re-minimizing an energy function to obtain a corrected fringe order distribution diagram, and reconstructing a final three-dimensional shape. And iteratively optimizing a projector principal point and a radial distortion coefficient in system calibration parameters. According to the scheme, the geometric details of the surface can be reserved, and the processing effect and the measurement precision of the complex curved surface and the high-curvature area are improved.

The invention relates to the technical field of geological disaster monitoring, and discloses a large-gradient deformation area phase optimization method and device, a medium and equipment, and the method comprises the steps: obtaining SAR data of a to-be-monitored area, and processing the SAR data to generate a differential interference image set; phase gradients of the differential interference image set in multiple directions are calculated to obtain a direction gradient image set, time sequence stacking and multi-direction fusion are carried out to obtain a phase gradient rate diagram, and a binary boundary diagram is determined through threshold segmentation; traversing the differential interferogram by using a sliding window, and performing structural consistency screening on neighborhood pixels in the window according to the binary boundary diagram to obtain a candidate pixel set; and performing amplitude consistency check, determining a homogeneous pixel set homogeneous with the central pixel, constructing a complex covariance matrix, and decomposing characteristic values to obtain an optimized phase of the central pixel. According to the method, the monitoring stability and precision are enhanced, the reliability and interpretability of a point selection result are improved, and an accurate phase reconstruction method is provided for a large-gradient deformation region.

The invention provides a composite material defect intelligent quantitative detection method based on multi-sensor phase signals, which comprises the following steps: calculating a phase gradient field according to continuous phase distribution, extracting a gradient peak region to indicate a wave reflection position, and fusing background interference correction to obtain defect boundary indication; if the phase gradient peak value exceeds a preset threshold value, gradient field features are analyzed through a convolutional neural network, defect phase change and background interference components are separated, and defect offset is judged; acquiring an offset mode through simulation comparison of the defect offset and the propagation path, classifying defect types corresponding to the offset mode by applying a support vector machine, and determining a type identification result; constructing a mapping function according to the type identification result and the offset, and associating phase change to geometric parameters, such as dimension depth orientation, to obtain a quantization parameter set; and verifying the matching degree with a phase gradient field through a quantization parameter set, and iteratively adjusting a mapping function to strengthen a stability relationship, so as to obtain a quantization model of defect parameters and phase changes.

The invention discloses a harmonic generation and beam control method and system based on an amplitude-phase regulation metasurface array. The method comprises the following steps of: constructing an amplitude-phase regulation and control metasurface array by adopting a digital control multi-channel radio frequency chip; based on a Fourier series and far field superposition model, calculating a time-varying amplitude phase sequence of harmonic waves and a phase gradient of beam pointing; the sequence is converted into a control signal through an FPGA, and each channel is modulated in real time; after an array receives a dot frequency signal, a target harmonic wave beam with a preset characteristic is radiated to a specified direction through periodic amplitude-phase modulation and space phase gradient processing. The system comprises an amplitude-phase regulation metasurface array construction module, a harmonic wave and beam control model construction module, a digital control signal generation loading module and a target harmonic wave generation radiation module, and is used for realizing the harmonic wave generation and beam control method. Accurate generation of any order of harmonic waves and flexible deflection of wave beams are achieved, the system integration degree, the regulation and control precision and the real-time performance are improved, and the application field is wide.

The invention relates to the technical field of medical instrument volume detection, and discloses a medical instrument internal volume detection system. A continuous wave and a pulse modulation signal are simultaneously transmitted through an annular phased array microwave antenna array, three-dimensional dielectric constant distribution is reconstructed by adopting a compressed sensing algorithm, and Doppler time-frequency domain characteristics are extracted through short-time Fourier transform; and inputting the phase gradient field and Doppler velocity field features into a convolutional neural network to generate a four-dimensional space-time volume evolution model, separating a macroscopic volume trend and a microcosmic pulsation component by adopting an adaptive Kalman filtering algorithm, and predicting an abnormal hemodynamic event through a long-short-term memory network. The technical problem of real-time precise detection of the multi-dimensional volume parameters of the blood circulation medical instrument under the complex pulsating flow condition is solved, and accurate evaluation of the instantaneous volume value, the pulsation amount and the pulsating frequency spectrum and early warning of abnormal events are achieved.

The invention relates to the technical field of electric energy distribution and signal energy distribution control, and discloses a product quality detection method based on machine vision, which comprises the following steps: acquiring signal energy distribution data representing product surface energy density, and mapping the signal energy distribution data to a complex frequency domain space to analyze a multi-scale phase component; constructing a local structure tensor by using a phase gradient, determining a background texture direction consistency index, establishing a phase feedback compensation mechanism, and adjusting a phase consistency weighting coefficient; according to the method, the problem of phase dispersion caused by non-uniform reflection is solved, the energy signal-to-noise ratio of the weak features is improved by dynamically adjusting the gain loop and energy distribution, and the signal-to-noise ratio of the weak features is improved. And precise quality identification under a high-reflection working condition is realized.

This invention discloses a phase compensation method for ship targets using ISAR based on block phase gradient autofocus. The steps include: pulse compression; range cell migration correction; range-oriented block segmentation; block phase compensation; calculation of the offset between adjacent image blocks; and image block alignment and stitching. This invention performs range-oriented block segmentation of the target echo, overcoming the shortcomings of existing technologies that ignore the spatial variation of rotational phase error, leading to defocusing in ship imaging, thus improving the accuracy of phase compensation and achieving good image focusing effect. Because this invention uses a phase gradient autofocus algorithm to compensate for the phase error within each block, and then uses a cross-correlation method to align and stitch the images of each block region, it overcomes the shortcomings of parameter estimation methods, such as complexity, time consumption, and poor real-time performance, and has the advantages of simple implementation, high efficiency, and strong real-time performance.

The invention relates to a dual-polarization 2bit transmission type metasurface unit working in a Ka wave band, which can realize four-state phase regulation and control of horizontal polarization and vertical polarization on a specified frequency band. And an implementation scheme with a better system integration advantage is provided for the high-frequency multi-polarization reconfigurable antenna system. The unit is composed of two groups of linearly polarized 2bit reconfigurable units which are orthogonally arranged, and each unit is composed of a pair of active receiving antennas at the bottom and a pair of active transmitting antennas at the top; a pair of PIN diodes are respectively integrated in a receiving antenna and a transmitting antenna, and four discrete phase states are realized through combination of switch states, so that 2-bit phase regulation and control are completed. The invention has the following advantages: 1, low transmission amplitude loss and precise phase gradient characteristic are shown in the dual-polarization Ka wave band, and high-precision wave beam regulation and control can be realized; 2, due to the fact that dual-line polarization adopts a synchronous regulation and control mode, the unit can complete phase control only through two feeder lines, and compared with an independent polarization control scheme, the complexity of a feed and bias network is remarkably reduced, so that interference of a bias circuit on electromagnetic radiation characteristics is effectively reduced; and 3, the emergent wave beam of the transmission type metasurface is located on the same side with the transmission type metasurface, so that the shielding of the feed source can be reduced, the interference of a reflection path is avoided, the aperture efficiency is improved, and the sidelobe is reduced.

The invention provides an InSAR orbit phase fine modeling method combining a frequency domain and a space domain, and the method comprises the steps: 1, obtaining a synthetic aperture radar image data set of a target region, carrying out the preprocessing, and carrying out the pairwise interference, and obtaining an initial interferogram; 2, a flat ground winding phase and a terrain winding phase in the interferogram are reserved, and an orbit phase influenced by a baseline is obtained; thirdly, the azimuth fringe frequency and the distance fringe frequency of the interferogram are obtained through Fourier transform and linear frequency modulation transform, and the azimuth phase gradient and the distance phase gradient are calculated; fourthly, an orbit fine phase model is established according to the phase gradient and the space polynomial model; according to the method, the track accurate phase model combining the frequency domain and the space domain is established, linear and non-linear track phases caused by inaccurate base lines are removed at the same time under the condition that InSAR interference diagram unwrapping is not needed, and the precision of track phase error removal is improved.

The application relates to a radio frequency coil design method for transmit array spatial encoding imaging, and belongs to the technical field of magnetic resonance imaging. In view of the problems of low single-dimension coding efficiency, strong coupling interference between coils and nonlinear phase gradient of a traditional TRASE technology, axial and radial coding coils are designed based on a target field method and a magnetic dipole method, two-dimensional radio frequency phase coding is realized through reverse current solenoid decoupling combination, and excitation and signal receiving are realized in combination with a saddle coil. Through optimization of the coil structure and decoupling design, the technical scheme achieves multi-dimensional efficient coding, high-precision phase gradient, strong anti-interference and system light weight, and is suitable for the portability and dynamic monitoring scene of low-field magnetic resonance.

The invention discloses a preemption and perception cross spoofing attack detection method based on systematic residual check. The method comprises the following steps: acquiring a damaged signal subjected to a PSCS attack; performing discrete Fourier transform on the damaged signal to obtain a plurality of resource elements, and extracting a phase gradient and an amplitude deviation for each resource element; defining a non-attack state constraint function and an attack state constraint function by using the phase gradient and the amplitude deviation corresponding to each resource element, so as to express the construction of test statistics as a minimum and maximum optimization problem; a minimum and maximum optimization problem is converted into a semi-definite programming problem through a kernel tensor completion technology, and test statistics are obtained by solving the semi-definite programming problem; based on the test statistics, calculating a detection threshold according to a false alarm probability; and calculating the current test statistic of the to-be-detected signal in the current time slot, and if the current test statistic is greater than the detection threshold, judging that the PSCS attack exists, thereby realizing the effective seizing and sensing cross spoofing attack detection method.

This disclosure provides a method for generating an image of an object using an MRI system, the method comprising providing a readout-coded imaging shearing (SIREN) gradient pulse in a phase gradient signal waveform. The phase gradient signal waveform is applied to the phase gradient coil of the MRI system. The application of the SIREN gradient pulse provides the object's SIREN k-space, which has SIREN k-space lines with shear angles. MR image spatial data from this SIREN k-space is then obtained by applying reconstruction techniques. Finally, the image of the object is generated by transforming the SIREN MR image spatial data into regular image spatial data based on the shear angle using a decoding algorithm.

The invention relates to the technical field of thermal ablation, in particular to a thyroid nodule ultramicro-invasive ablation precise puncture positioning method, which comprises the following steps of: acquiring continuous gray frames and extracting tissue gray to complete normalization and phase demodulation to generate phase time data; detecting a brightness peak value according to phase time data to obtain a tip echo, calculating a phase gradient in a neighborhood by using a fixed window to generate a stable domain data set, calculating a spatial range according to a stable domain, extracting a phase change signal along with time, and completing frequency domain dominant frequency extraction and period analysis to generate a period consistent region; according to the method, the continuous image phases are extracted, the time sequence is combined, the needle point recognition stability is enhanced, the echo region is restrained through the brightness peak value and the phase gradient, and the accuracy of positioning is improved. Position judgment is enhanced through spatial range and period analysis, and the guiding precision is improved by combining a three-dimensional distance and a symbol sequence.

A high-frequency circuit includes a plurality of beam former ICs. The beam former IC includes a parameter acquisition unit and a set value acquisition unit. The parameter acquisition unit acquires a phase shift parameter related to phase shift of each of a plurality of antenna elements set corresponding to a direction of a beam-shaped high-frequency signal transmitted and received by each of the antenna elements. The phase shift parameter is acquired based on at least one phase gradient parameter related to a phase difference between the antenna elements adjacent to each other, a physical position of each of the plurality of antenna elements, and a density parameter related to resolution of a beam direction of the high-frequency signal. The set value acquisition unit acquires at least a set value of the phase shift of each of the antenna elements according to the phase shift parameter.

The invention relates to a three-dimensional reconstruction method based on DCT (Discrete Cosine Transform) decoding. The method comprises the following steps: projecting a four-step phase shift stripe of a specific code to an object and collecting a deformed stripe; calculating a wrapped phase; the key improvement lies in that after a least square unwrapping model is constructed, improved divergence calculation based on a cosine function is adopted, nonlinear transformation is carried out on a multi-direction wrapped phase gradient so as to suppress noise errors, and an improved divergence field is generated; secondly, performing two-dimensional DCT (Discrete Cosine Transform) forward transformation on the divergence field, and solving an absolute phase field through inverse transformation after operation with a transfer function in a frequency domain; and finally completing three-dimensional reconstruction. The system comprises a projection module, an acquisition module and a processing module. According to the method, non-linear divergence calculation is introduced to be combined with DCT frequency domain solving, the defect that a traditional method is sensitive to noise is effectively overcome, and the reconstruction speed and robustness are improved while precision is guaranteed.

The invention relates to the technical field of data processing, and further relates to a data extraction method and system for digital geological mineral exploration. Comprising the following steps: step 1, performing analog-to-digital conversion on multiple paths of original magnetic field response voltage signals synchronously acquired by multiple optical pump atomic magnetometer sensing units to obtain multiple paths of original magnetic field response digital sequences; 2, performing phase gradient field iteration reconstruction processing on the magnetic field in-phase response matrix and the magnetic field orthogonal response matrix to obtain a reconstructed phase matrix and a magnetic field amplitude matrix; and 3, calculating a magnetic field amplitude deviation matrix and a phase deviation matrix according to the magnetic field amplitude matrix and the reconstructed phase matrix, converting the mineral anomaly candidate position into a geological space coordinate, and outputting geological mineral anomaly feature data.

The invention relates to the technical field of image recognition and pattern recognition, and discloses a fusion tracking method based on AI large model video analysis and RFID positioning, which comprises the following steps: when a target enters a visual non-overlapping blind area, according to a phase gradient change trend of a radio frequency identification signal sequence acquired by a radio frequency signal acquisition unit, acquiring a target image according to a phase gradient change trend of the radio frequency identification signal sequence; generating a probability weight map representing the spatial position of the target; projecting the image to an image coordinate system, and generating an airspace attention weight adjustment mask; according to the method, the mapping relation between the radio-frequency signal phase gradient and the image neuron response weight is established, the spatial topology constraint is introduced in the feature extraction stage, structural contraction of the feature space is achieved, and the target handover is completed by utilizing the mask to modulate the depth feature map and extracting candidate target feature vectors according to the modulated depth feature map. And the target binding stability in the cross-monitoring-area tracking process is effectively enhanced.

This invention provides a method and system for tracking unmanned aerial vehicles (UAVs), relating to the field of data processing technology. The method includes: acquiring multiple consecutive video frames at acquisition times with annotation information of the target to be tracked; extracting a phase information structure descriptor of the target from each video frame using Zernike moments; constructing a phase flow field describing the motion characteristics of the target based on the phase information structure descriptor; determining whether the target is occluded by combining an occlusion factor based on the phase information structure descriptor; if so, determining multiple predicted positions of the target using the Lie group motion principle, combined with the phase flow field; acquiring multiple predicted video frames with each predicted position as the focus center of the UAV camera; capturing the target from each predicted video frame with the goal of minimizing the phase gradient cross product; otherwise, capturing the target based on the phase information structure descriptor; and tracking the target based on the capture results. This improves tracking anti-interference and endurance capabilities.

The present application relates to the technical field of thickness measurement, and discloses a linear dimensional measurement method for AR module thickness, comprising: obtaining a spectral interference signal and resampling to a wave number space to generate a wave number domain complex sequence; extracting a real part, an imaginary part and a first derivative term of the wave number domain complex sequence; establishing an orthogonal component set; extracting a transient phase slope by using differential mapping between the orthogonal components; determining a wave number domain group delay distribution; calling a dispersion calibration matrix to correct nonlinear deviation; converting nonlinear phase evolution into linear phase gradient; fitting the linear phase gradient to solve a physical thickness value; and bypassing arctangent calculation and phase unwrapping by establishing phase gradient mapping logic, suppressing phase jump caused by interference envelope aliasing from the dimension of physical energy distribution, improving thickness calculation stability of an extremely thin medium layer under environmental interference, and realizing physical decoupling of material dispersion characteristics and geometric dimensional parameters.

A method according to the invention for acquiring T2*-weighted measurement data of a test object using a magnetic resonance system comprises the steps: - irradiating a first RF excitation pulse to excite spins in a first layer in the test object, - switching gradients to generate at least one gradient echo signal, - after an echo time following the irradiation of the first RF excitation pulse, acquiring a first gradient echo signal from the first layer, - after the irradiation of the first RF excitation pulse and before the elapsed echo time following the first RF excitation pulse, irradiating at least one further RF excitation pulse to excite spins in a further layer,wherein, before the application of a further RF excitation pulse and after each RF excitation pulse preceding the further RF excitation pulse, a pre-rephasing gradient of a moment corresponding to the further RF excitation pulse following the pre-rephasing gradient is switched, - recording generated further gradient echo signals from further layers. By switching the pre-rephasing gradients according to the invention, gradient echo signals of the spins of different layers excited by the first RF excitation pulse and the at least one further RF excitation pulse can be separated. Thus, the echo time TE, which elapses until a first gradient echo signal is measured after the first RF excitation pulse, can be used to generate and measure further gradient echo signals in further layers after the first gradient echo signal, which increases the readout efficiency.

A beam steering system includes a differential segmented array (DSA) antenna comprising a plurality of pyramid structures and elements arranged in an array comprising a first and second set of direction elements, where each element is defined between opposing faces of two adjacent pyramid structures and a position of each element is located at a distance from a common origin of the elements of the array; phase gradient determination circuitry to determine a first and second phase gradient for the direction elements, where the phase gradients are based on a first and a second angle of a target with respect to the DSA antenna, and an operating frequency of the DSA antenna; and phase shift determination circuitry to determine a first and second phase shift, for each of the elements, and to determine a resultant phase shift, for each element, by summing the respective first and second phase shifts.

This invention relates to the field of image recognition technology, specifically a method for identifying transparent foreign fibers in cotton using polarization detection combined with image processing. The method includes the following steps: acquiring images using a polarization sensor and extracting temporal and angular information; calculating pixel phase delay; performing frequency domain analysis to determine the principal direction and generating a birefringence response; screening skeleton points based on grayscale and constructing fiber paths; combining phase gradients for extreme value analysis and intensity reconstruction to obtain global coordinates; and then weighted filtering based on directional differences to eliminate redundancy and obtain fiber position coordinates. In this invention, by combining image acquisition and feature expression paths, polarization response and texture information are fused to form multi-dimensional discriminative features, reducing noise and uneven illumination interference. Combining fiber continuity and optical response analysis enhances the ability to identify weak responses, making boundaries clearer and more stable. Furthermore, redundant points are compressed under directional classification and spatial constraints, improving positioning consistency and recognition accuracy, and achieving stable identification of transparent foreign fibers in complex scenarios.

The invention relates to an ultrasonic elastography system and method based on a local least square fitting operator. Comprising the following steps: acquiring a plurality of wave fields under a target frequency based on spatio-temporal data of ultrasonic shear wave propagation; obtaining a relative coordinate matrix based on the fitting matrix; obtaining a fitting operator matrix based on the relative coordinate matrix and a least square principle; for each pixel of the complex wave field, extracting a local complex data block taking the pixel as a center; acquiring relative phase vectors respectively corresponding to the plurality of local complex data blocks; on the basis of the fitting operator matrix and the relative phase vectors corresponding to the multiple local complex data blocks, fitting wave number components corresponding to the multiple local complex data blocks are obtained; and on the basis of the fitting wave number components corresponding to the plurality of local complex data blocks, shear wave velocities corresponding to the plurality of local complex data blocks are obtained, and then an elastic graph is obtained. According to the method, random noise can be effectively suppressed in a phase gradient calculation stage, and a fine edge structure of a tissue is reserved.

The invention discloses a method and system for suppressing non-pulse noise to accurately measure pulse arrival time, and the method comprises the steps: S1, generating pulsar simulation parameters which comprise a rotation period P, a flow density S1250 and a pulse arrival phase delay standard deviation sigma delta p; s2, constructing a pulsar standard contour Q (x), and generating based on a pulse contour type Ptype, a sub-pulse number and a sub-pulse contour shape; s3, simulating multi-epoch observation data, injecting noise, adding phase delay delta p, and generating an average contour q (x); s4, judging a pulse region: positioning a candidate pulse region through baseline identification, and screening a final pulse region according to a contour type; s5, calculating a phase difference delta x between the standard contour and the average contour by using a phase gradient method; and S6, comparing a difference value between the phase difference delta x and the analog phase delay delta p, and verifying whether higher timing accuracy is obtained by removing the noise in the non-pulse region or not. The pulsar timing precision is remarkably improved by reducing noise interference in a non-pulse area, and the pulsar timing method is suitable for a high-precision scientific target.

A magnetic resonance imaging system and method, a computer readable storage medium. The present invention provides a magnetic resonance imaging method, a magnetic resonance imaging system and a computer readable storage medium. The magnetic resonance imaging method comprises: acquiring a plurality of parts of k-space using a plurality of imaging sequences to obtain a plurality of k-space data sets, each imaging sequence comprising a pre-spread phase gradient pulse and a plurality of phase encoding gradient pulses applied after the pre-spread phase gradient pulse, wherein the pre-spread phase gradient pulses in the plurality of imaging sequences have a standard area difference in turn when sorted by area value, the standard area difference being 2 / N of the area of any phase encoding gradient, wherein N is the number of groups of the plurality of imaging sequences; reconstructing a magnetic resonance image from each of the plurality of k-space data sets, respectively; and processing the plurality of k-space data sets to obtain a magnetic resonance image.

The embodiment of the invention provides a near-field MIMO radar sidelobe suppression method based on phase gradient self-focusing and a space adaptive time-frequency window function, and is applied to the technical field of radar signal processing. The method comprises the following steps: acquiring an MIMO radar echo signal of a target scene, and generating a wavenumber domain image based on the MIMO radar echo signal; performing phase gradient self-focusing processing based on the wave number domain image to obtain a phase correction image; performing spatial adaptive division based on the phase correction image to obtain a plurality of sub-block images; window function matching is carried out on each sub-block image, a target window function corresponding to each sub-block image is determined, windowing is carried out on the sub-block images based on the target window functions, and windowed sub-block images are obtained; fusing the plurality of windowed sub-block images based on a preset fusion strategy to obtain a fused image; and performing secondary phase gradient self-focusing processing on the fused image to obtain a target imaging result. The technical effect of improving the definition of the imaging result is achieved.

Provided in the present application are a phase weighting apparatus, a beamforming apparatus, a system and a method. The phase weighting apparatus comprises: a primary phase weighting network, used for performing first phase weighting on an optical wave signal, a first phase gradient obtained by means of the first phase weighting being determined on the basis of an input port via which the optical wave signal is fed into the primary phase weighting network; and a secondary phase weighting network, connected to the primary phase weighting network and used for performing second phase weighting on a plurality of optical wave signals, a second phase gradient obtained by means of the second phase weighting being determined on the basis of the wavelengths of the optical wave signals. In the embodiments of the present application, the secondary phase weighting network can perform phase weighting on the signals output by the primary phase weighting network again on the basis of the wavelengths. In this way, when optical wave signals of different wavelengths are input into the phase weighting apparatus via the same port, different phase weight values can be obtained, thereby supporting forming beams in multiple directions, and helping to increase the control dimensions of the beams and enhance the flexibility of beam steering.

The application discloses a kind of based on porous material's double-side phase gradient composite sound absorption metasurface structure and design method, it is related to acoustic metamaterial and sound absorption metasurface technical field, structure includes: rigid metal frame and multiple periodic arrangement sound absorption metasurface subunit, each subunit is divided into acoustically isolated first, second functional area along the thickness direction, respectively control transmission, reflection sound wave phase, functional area is equipped with rigid baffle and the sound channel of porous sound absorption material in it. By differentiating design subunit parameter, make it form 0~2π phase distribution at target frequency, form phase gradient on both sides, in combination with the period design of λ / d>2, convert high-order diffraction wave into surface wave, cooperate porous material to realize sound energy efficient dissipation.The total thickness of the application is less than 1 / 8 of target wavelength, predetermined frequency band sound absorption coefficient α≥0.8, the performance is stable from-50° to +50° incident angle, adapts to a variety of porous materials and target frequency, widely used in noise control scene.

The present application relates to the technical field of detection, and particularly relates to a complex shell detection method and system based on structured light scanning. The method comprises the following steps: calculating the wrapped phase value and phase modulation amplitude of each pixel point at each frequency, calculating the discrete probability distribution containing the candidate fringe order and the corresponding probability for each pixel point based on the multi-frequency wrapped phase difference, generating an initial pixel confidence map, constructing an energy function, determining the penalty weight of the smoothing term according to the phase modulation amplitude and the local variance of the wrapped phase gradient, obtaining a preliminary fringe order distribution map, updating the discrete probability distribution by using the initial pixel confidence map, re-minimizing the energy function to obtain a corrected fringe order distribution map, reconstructing the final three-dimensional topography, and iteratively optimizing the projector principal point and the radial distortion coefficient in the system calibration parameters. That is, the scheme of the present application can preserve the surface geometric details, improve the effect of processing the complex curved surface and the high-curvature area, and improve the measurement accuracy.

This invention discloses an intelligent identification method for creep-type landslide hazards that combines image processing and semantic understanding. First, high-precision geometric modeling and filtering techniques are used to suppress irrelevant phases in InSAR interferograms while preserving deformation information related to landslides. Next, signal features are enhanced through phase gradient, RGB channel mapping, and generative adversarial networks to improve signal representation capabilities. Subsequently, convolutional operations, self-attention mechanisms, and structural reparameterization are used to optimize the deep learning model, improving its learning ability for landslide hazards. In the feature encoding stage, a multimodal caption generation model is used to obtain visual and linguistic features, which are then mapped to the same embedding space for seamless integration. Finally, an autoregressive language generation model and a multilayer perceptron are used to predict landslide hazards. This approach, by integrating visual, linguistic, and deep learning technologies, significantly improves the efficiency and accuracy of landslide hazard monitoring, providing effective support for disaster response and management.