23 results about "Iterative reconstruction" patented technology

The present disclosure provides a cone beam computed tomography system for proton therapy comprising an x-ray source configured to generate x-rays and positioned to rotate around a patient at a treatment isocenter, a detector assembly positioned opposite the x-ray source and configured to receive x-rays that have passed through the patient, a rotation mechanism configured to rotate the x-ray source and the detector assembly around the patient while maintaining their relative positions, a translation mechanism configured to translate at least one of the x-ray source, the detector assembly, or a patient couch during rotation to implement a wobbled scan orbit, and a scatter rejection grid attached to the detector assembly and focused on the x-ray source to reduce scattered x-rays. The translation mechanism distributes missing projection data across multiple radii and angles, enabling improved reconstruction quality through iterative reconstruction methods.

The application provides a prestressed steel bar nondestructive detection method and system based on ultrasonic imaging, relates to the technical field of prestressed concrete structure nondestructive detection, and is characterized in that a quasi-distributed ultrasonic transmitting-receiving array is arranged along both sides of the prestressed steel bar, each probe is sequentially excited and a multi-path echo sequence signal is synchronously collected, Hilbert-Huang transformation is performed on each path signal, a marginal spectrum is obtained through empirical mode decomposition and Hilbert spectrum integration, a barycentric frequency is extracted as a characteristic value which is highly sensitive to interface stiffness degradation, the barycentric frequencies of all paths are taken as projection data, an algebraic iterative reconstruction algorithm is used in combination with the spatial coordinates of the probes, a dispersion characteristic tomographic image of the measured section is reconstructed, a barycentric frequency variation curve is extracted from the image along the direction of the steel bar, and the area where a continuous shift occurs and the shift amount exceeds a threshold value is determined as an interface stiffness abnormal area, so that two-dimensional imaging diagnosis and spatial positioning of the interface state of the prestressed steel bar are realized.

This invention relates to the field of medical image reconstruction technology, specifically to a method for reconstructing low-dose CT images of patients with coronary artery disease based on image enhancement. The method includes: acquiring multi-angle raw projection data of low-dose CT images from patients with coronary artery disease; performing statistical noise modeling and signal recovery on the projection data to obtain denoised projection data; generating an initial coronary CT image sequence through an iterative reconstruction algorithm; extracting cardiac structures and completing region segmentation to obtain labeled images; performing region-specific differential image enhancement based on the labeled images; enhancing vessel wall details and extracting calcified plaque features; combining edge-preserving noise suppression with post-optimization; and outputting high-quality low-dose CT reconstructed images. This method effectively improves the noise artifact problem caused by low-dose scanning, preserves key anatomical details and lesion features of the coronary arteries and myocardium, and enhances the clinical diagnostic practicality of low-dose CT images for coronary artery disease.

The invention relates to an emission tomography image reconstruction method. The method comprises the steps that an emission tomography data set is acquired, iterative reconstruction is performed on the emission tomography data set to obtain a target reconstruction image, and in the iterative reconstruction process, the following steps are performed in at least one iteration: an intermediate iteration image before the execution of the current iteration is acquired; the intermediate iteration image is obtained through initial reconstruction or multiple reconstruction of an emission tomography data set; segmenting the intermediate iteration image to obtain a plurality of sub-regions; each sub-region corresponds to different anatomical structures or regions of interest; inputting the at least one sub-region into a trained neural network, and obtaining a pixel prediction value of the sub-region; updating at least one pixel value of the intermediate iteration image based on the pixel prediction value to obtain a corrected intermediate iteration image; and performing activity reconstruction based on the corrected intermediate iteration image to obtain an intermediate iteration image of the current iteration. The method can improve the convergence speed.

This specification provides a method and computer device for estimating the point spread function of an image, relating to the field of medical device technology. The method includes: selecting at least one feature point in the reconstructed image space; establishing an orthographic projection model based on the geometric parameters of the CT system and the current scanning parameters to simulate the diffusion distribution of the feature point as an ideal point source in the projection data domain, thereby obtaining the orthographic projection point spread function; performing a virtual backprojection operation on the feature point based on the current reconstruction method and analyzing its response distribution to obtain the backprojection point spread function; and synthesizing the orthographic projection point spread function and the backprojection point spread function to obtain the overall point spread function for iterative reconstruction or post-processing of CT images. Thus, the point spread function corresponding to the CT system can be adaptively and quickly estimated based on the scanning and reconstruction parameters, facilitating improved image quality in iterative reconstruction or post-processing of CT images.

This invention relates to an autonomous driving scene reconstruction method based on MCMC and scale-aware time-varying scheduling, comprising: initially modeling the scene as a set of anisotropic 3D Gaussian spheres, each Gaussian sphere being parameterized by learnable parameters to construct a learnable parameter space; iteratively reconstructing the autonomous driving scene based on the learnable parameter space, wherein in each iteration, a scale-based semantic masking mechanism is used to automatically distinguish scene regions using the statistical distribution characteristics of the Gaussian sphere scale; wherein, during the iteration process, based on the stochastic gradient Langevin dynamics update mechanism, an adaptive time-varying noise scheduling strategy is introduced to model the noise injection process in Markov chain Monte Carlo MCMC sampling as a control process that dynamically changes with the number of training iterations. Compared with existing technologies, this invention has the advantages of stable exploration capability and robust geometric constraints.

This invention discloses a method and apparatus for parallel matrix factorization tensor rank reduction reconstruction based on real trace coordinates. The method includes: converting non-uniformly sampled pre-stack 5D seismic data in the time-space domain to the frequency-space domain via Fourier transform; determining weight values ​​for mapping coordinates from real trace coordinates to regular grid points using the IDW interpolation operator; reconstructing the spatial four-dimensional data corresponding to each frequency slice: mapping the spatial four-dimensional data corresponding to the frequency slice to regular grid points using weight values; iteratively reconstructing the spatial fourth-order tensor at the regular grid points using a projection gradient descent algorithm based on BB step size and a parallel matrix factorization rank reduction algorithm; and performing a one-dimensional fast Fourier inverse transform on the reconstructed frequency-space domain seismic data with respect to the frequency variable to obtain the reconstructed time-space seismic data. This invention avoids introducing amplitude and phase errors and can efficiently and accurately reconstruct pre-stack 5D non-uniformly sampled seismic data without the need for surface elementization.

This application discloses an infrared image super-resolution reconstruction method, apparatus, device, and medium, relating to the field of image processing technology. It improves reconstruction stability and image quality by actively controlling half-pixel-level subpixel displacement to acquire multiple frames of infrared images and combining subpixel registration, a forward degradation model, and a dual-constraint objective function for iterative reconstruction. The method includes: controlling the relative position between the infrared imaging system and the target scene to generate half-pixel-level subpixel displacement, obtaining multiple frames of original infrared images; performing subpixel registration on the multiple frames of original infrared images to obtain the subpixel displacement amount of each frame of original infrared images relative to a reference frame of infrared images; establishing a forward degradation model based on the subpixel displacement amount, using the forward degradation model to characterize the mapping relationship between each frame of original infrared images and the high-resolution image to be reconstructed; constructing an objective function based on the forward degradation model; and solving for the high-resolution image using an iterative optimization algorithm based on the objective function.

The present application relates to a kind of image iterative reconstruction methods suitable for a large number of not-through area scene, belong to computer tomography imaging technical field, solve the poor stability of the reconstruction effect in prior art when reconstructing for different to-be-reconstructed workpieces.The iterative reconstruction method includes: segmentation is determined to initial reconstruction image for artifact region, using preset attenuation coefficient replaces the attenuation coefficient value of other pixel points except artifact region in initial reconstruction image to obtain compensation reconstruction image;Determine penetration threshold according to original sinogram;Compensation weight sinogram is generated using penetration threshold, original sinogram and initial reconstruction image, and iteration termination condition is determined based on penetration threshold and original sinogram;Obtain final reconstruction image by several iterations.Improve the stability of the reconstruction image effect and reconstruction process for different to-be-reconstructed workpieces.

This invention discloses a sprinkler irrigation control method based on communication link attenuation inversion, belonging to the fields of agricultural automatic sprinkler irrigation technology and wireless communication engineering technology. The method includes: deploying multiple fixed communication nodes in farmland to form a mesh link; each node sequentially sending ranging signals and recording the actual received power of each link; calculating the theoretical power using a free-space loss model to obtain the additional attenuation; discretizing the farmland into a three-dimensional voxel grid and establishing a system of linear equations relating the attenuation to the voxel attenuation coefficient; solving the equations using an algebraic iterative reconstruction algorithm to obtain the soil moisture distribution matrix within the farmland; identifying regions based on the moisture distribution and generating differentiated sprinkler irrigation commands; and controlling the electromagnetic valves of each zone via a wireless network to perform variable-volume sprinkler irrigation. This invention eliminates the need for buried soil moisture sensors, utilizes communication link tomography technology to obtain a true three-dimensional moisture field, is unaffected by weather, achieves precise closed-loop sprinkler irrigation, and significantly improves water-saving efficiency and crop uniformity.

The application belongs to the technical field of power system operation control and optimal scheduling, and provides a power transmission and distribution reactive power reserve planning method and system considering power distribution network power flow security domain, generates a power flow security domain sample set considering network reconstruction based on an improved radial iterative reconstruction method, uses a physical information graph neural network (PI-GNN) to simulate and solve the security domain in consideration of new energy uncertainty, defines an equivalent four-dimensional security domain cost model of the power distribution network on the basis of considering optimal carbon emission power flow, and deduces and forms a planning scheme of the power transmission and distribution system reactive power reserve considering the operation state under a typical scene. The implementability and economy of the planning scheme under multiple working conditions throughout the year are improved.

This invention discloses a method and system for analyzing and evaluating the treatment effects of patients in cardiovascular rehabilitation using computer vision and image processing technologies. The method involves measuring the error distribution through virtual repeated scanning to obtain a threshold for generating significant evidence in the image domain. The predicted projection difference is obtained by forward projecting the candidate Region of Interest (ROI) change field, and its consistency with the actual projection difference is calculated to obtain a projection consistency score (PCS), forming supporting evidence in the projection domain. When the evidence is at the boundary or conflicting, iterative reconstruction and verification of the ROI difference is triggered. Under projection constraints and regularization conditions, the change map within the ROI is solved, and the residual reduction rate and morphological criteria are calculated as a third evidence domain. The evidence from the image domain, projection domain, and ROI reconstruction domain is fused for mutual verification and adjudication, outputting whether the change is real or undeterminable and the cause code, and generating a follow-up trend report and evidence summary package. This invention can reduce false positives and improve the stability, verifiability, and clinical safety of rehabilitation efficacy assessment.

The application discloses a multi-focus image fusion method and system based on a semi-smooth Newton method, and belongs to the technical field of image processing. The semi-smooth Newton algorithm is used for solving, a proximal operator and a generalized Jacobian matrix are introduced, and the optimization solving process is expanded into a deep network. An adaptive feature iterative reconstruction unit is designed, auxiliary variable updating, a learnable edge convolution, focus gradient perception and a multi-scale frequency enhancement structure are integrated, and iterative optimization is realized in combination with a feature reconstruction block, a multiplier updating module and a matrix updating module. The second-order optimization algorithm is introduced into multi-focus image fusion for the first time, has physical interpretability and strong feature learning capability, and has better fusion effect than existing methods, and has important application value in high-precision imaging fields such as microscopy and remote sensing.

The application relates to a dynamic reconstruction method and device. The method comprises the following steps: obtaining an initial dynamic image of a target object according to dynamic scanning data of the target object; performing iterative reconstruction on each frame image in the initial dynamic image, and adjusting the iterative parameter of the reconstruction of each frame image according to the image quality index quantization value of each frame image in each iterative reconstruction; and determining a dynamic reconstruction image of the target object based on the iterative parameter of each iterative reconstruction of each frame image. In the multiple iterative processes of the dynamic reconstruction, the iterative parameter of the reconstruction of each frame image is continuously optimized and adjusted according to the image quality index quantization value of each frame image, the image quality of each frame image is improved, and the image quality of the finally obtained dynamic reconstruction image is improved.

The present application relates to the technical field of industrial big data processing, in particular to a self-adaptive agile data warehouse construction and intelligent operation and maintenance method for industrial scenes, comprising two core processes of self-adaptive agile data warehouse construction and intelligent operation and maintenance; the construction method realizes the scene-based agile construction of the data warehouse through industrial metadata feature vector space construction, adaptive scene-based clustering, ETL link adaptive generation and incremental iterative reconstruction; the intelligent operation and maintenance method realizes the intelligent operation and maintenance of the data warehouse in the whole life cycle through multi-dimensional operation and maintenance feature modeling, adaptive anomaly detection, automatic disposal and full closed-loop feedback optimization; the present application greatly improves the construction efficiency, scene adaptation capability and operation stability of the industrial data warehouse, and can be widely applied to various intelligent manufacturing industrial scenes.

The application provides a method and related equipment for enhancing a while-drilling seismic signal. The method comprises: acquiring a while-drilling seismic signal; calculating a time-varying structure element according to a local variance of each trace of the while-drilling seismic signal; performing morphological dilation and corrosion operations at different scales for each time point and spatial trace of each while-drilling seismic signal to calculate a multi-scale morphological gradient value; constructing a weight matrix according to the multi-scale morphological gradient value; and performing weighted morphological filter iteration reconstruction on the while-drilling seismic signal based on the weight matrix to obtain an enhanced while-drilling seismic signal. The embodiment of the application realizes real-time enhancement of the while-drilling seismic signal by dynamic structure element generation and multi-scale morphological gradient weighting, and combines weighted morphological filter iteration reconstruction, thereby significantly improving the signal-to-noise ratio, retaining lithology interface reflection characteristics, suppressing noise and interference, and providing reliable technical support for underground coal mine seismic monitoring.

This invention discloses a PET imaging noise suppression method based on a deep neural network. To address the problem that low-count PET images are prone to noise amplification and artifacts during iterative reconstruction using a point spread function (PSF) model, leading to decreased signal-to-noise ratio and microstructure discernibility, and difficulty in balancing resolution and quantitative accuracy, this invention performs iterative reconstruction of the original PET data using a PSF model to obtain an initial PET image and determine the PSF model parameters. Based on these parameters, a spatially variable condition map is generated that is spatially registered with the initial PET image. This spatially variable condition map is then used to infer a conditional denoising deep neural network to output a noise residual map, a detail compensation map, and a fusion weight map. Finally, based on the fusion weights, the noise residual and detail compensation are weighted and fused at the voxel level, and then combined with the initial PET image at the voxel level to obtain a noise-suppressed PET image. This achieves the technical effect of significantly reducing noise and artifacts and improving image quality while maintaining detail and contrast.

This application discloses a sparse adaptive reconstruction method based on dynamic adjustment and binary search. This method obtains compressed observations by constructing a random Gaussian observation matrix and uses discrete wavelet transform to sparsely represent the original signal. During the iterative reconstruction process, this application introduces for the first time a dynamic adjustment mechanism for the cumulative energy ratio atoms guided by a dynamic scaling factor, and employs a binary search strategy to achieve efficient sparse estimation. This method successfully creates a power quality disturbance signal reconstruction scheme that simultaneously possesses sparse adaptiveness, fast convergence, and high robustness to noise. Compared with methods such as OMP and the original SAMP, this method can reduce the number of iterations by 60%–80%, improve the reconstruction signal-to-noise ratio by 8–12 dB, and reduce the root mean square error by 10%–25%. It exhibits higher robustness and real-time performance in power quality disturbance signal reconstruction and is suitable for scenarios such as compressed sampling, edge computing, and high-speed signal reconstruction in power quality monitoring systems.

This invention discloses a target-driven, combined norm adaptive super-resolution imaging method. First, a generalized combined regularized objective function is constructed by introducing a generalized sparse norm and a generalized total variation norm, integrating the advantages of multiple norms to enhance adaptability to different scenarios. Then, based on a generalized Gaussian distribution model, the sparse norm and total variation norm are modeled as functions of the target scattering coefficients. A hyperbolic fit to the inverse function of the generalized Gaussian ratio is used to achieve dynamic adaptive updating of the norm order with iterative reconstruction results. Finally, an iterative reweighting method is employed to transform the non-convex objective function into a weighted norm form, and a closed-form solution expression is derived. This method significantly reduces the workload of manual parameter tuning, effectively improves the algorithm's adaptability to different target scenarios and imaging quality, and solves the problem of poor scene adaptability caused by a fixed norm order in forward-looking radar super-resolution imaging.

A weather editing method and system based on a 3D Gaussian splash model is proposed. First, an initial 3D scene is constructed based on the original acquired 2D initial image sequence to obtain an initial 3D Gaussian splash model represented by the original 3D Gaussian primitives. Then, a 2D edited image corresponding to the target weather is generated or acquired. Next, parameter update residuals are obtained based on the difference between the 2D edited image and the image to be rendered generated by the initial 3D Gaussian splash model. Finally, the model parameters of the 3D Gaussian splash model corresponding to the target weather scene, obtained by modifying the model parameters of the initial 3D Gaussian splash model, are iteratively updated based on the parameter update residuals. By introducing a gradient-guided optimization paradigm into the weather editing task of the 3D Gaussian splash model, direct gradient-guided generalizable updates replace the traditional view-by-view iterative reconstruction, fundamentally improving robustness to cross-view inconsistencies in 2D diffusion priors, while significantly shortening the editing time.

The present application relates to a kind of image reconstruction methods based on resolution progression, belong to computer tomography imaging technical field, solve the problem of low efficiency when the prior art iterative reconstruction.The image reconstruction method includes: by the acquisition sinogram of the workpiece to be reconstructed of detector obtains acquisition sinogram, to acquisition sinogram is down sampled and obtains at least two resolution levels of original sinogram;According to the order from low to high of resolution level, in turn, each resolution level of original sinogram is iteratively reconstructed, and the reconstructed image of highest resolution level is regarded as final reconstructed image.It improves the reconstruction efficiency when iterative reconstruction.

The present disclosure provides a cone beam computed tomography system for proton therapy comprising an x-ray source configured to generate x-rays and positioned to rotate around a patient at a treatment isocenter, a detector assembly positioned opposite the x-ray source and configured to receive x-rays that have passed through the patient, a rotation mechanism configured to rotate the x-ray source and the detector assembly around the patient while maintaining their relative positions, a translation mechanism configured to translate at least one of the x-ray source, the detector assembly, or a patient couch during rotation to implement a wobbled scan orbit, and a scatter rejection grid attached to the detector assembly and focused on the x-ray source to reduce scattered x-rays. The translation mechanism distributes missing projection data across multiple radii and angles, enabling improved reconstruction quality through iterative reconstruction methods.