104results about How to "Fast refactoring" patented technology

The invention discloses an edge structure information based block compression perception reconfiguration method which mainly overcomes the defects of low speed and obvious block effect of a reconstructed image in the prior art. The method comprises the following steps of: (1) dividing an image into 32*32 blocks; (2) carrying out wavelet transform on the image blocks, reserving low-frequency signals of the image blocks, and compressing and sampling high-frequency signals; (3) zeroing all high-frequency image block parts, carrying out reverse wavelet transform on the reserved low-frequency information and the zeroed parts to obtain a blurred image and carrying out edge detection on the image; (4) reconstructing image blocks without edges by adopting an orthogonal matching pursuit method, reconstructing image blocks with edges by adopting an MP (Monolithic Processor) method guided by edge structure information, and reconstructing non-edge parts by adopting a generalized inverse method; and (5) splicing reconstructed image blocks to obtain a reconstructed image of an original image. Compared with the prior art, the method has the advantages of high speed and good reconstruction effect and is suitable for reconstructing natural images and SAR (Synthetic Aperture Radar) images.

The invention discloses a method for fast establishing interactive tunnel and wall rock body three-dimensional models, which comprises the following steps: A. tunnel line position information base establishment: collecting tunnel line position information and establishing a line position information base; B. geological exploration base establishment: carrying out standardized processing on hole drilling data obtained through geological exploration; C. three-dimensional stratigraphical profile constraint information base establishment: collecting three-dimensional stratigraphical profile constraint information; D. three-dimensional stratigraphical boundary constraint information base establishment: collecting self defined model establishing boundary; E. real three-dimensional tunnel model establishment: precisely establishing a tunnel three-dimensional model according to the information of tunnel line positions, the tunnel hole body outer cross section and the inner cross section; F. real three-dimensional wall rock model establishment: establishing a wall rock model according to the hole drilling and stratigraphical information; and G. tunnel digging and profile real-time cutting display based on the wall rack model body unit. The model establishing efficiency and precision of the tunnel simulation and digitalized design can be improved, the real-time cutting display of the tunnel passing through the wall rock and any vertical profile can be realized, and the fast reestablishment of the model can be realized.

The invention discloses a distributed microphon array sound source positioning method based on space sparsity, comprising steps of performing signal characteristic extraction through two steps of discrete cosine transform (discrete cosine transform DCT), utilizing the characteristics to construct a sparsity positioning model in order to comprehensively utilize the short time or long time characteristics of the voice signal, reducing the dimension of the model, utilizing a dictionary learning method to perform correction on the model, adopting an approximating I0 norm reconstruction algorithm to perform reconstruction on a sparse signal to find a sound source. The invention has no special requirement on the array structure, requires less calculation, effectively improves the capability of resisting noise and reverberation and positioning accuracy, and is suitable for indoor sound positioning and widely applicable to the video conference system, the voice recognition system and the intelligent robots.

According to a ground feature contour extraction method based on a ground laser point cloud provided by the invention, surface-based space measurement breaking through a traditional single-point measurement mode is introduced into urban measurement application, and the urban measurement efficiency is improved. As the most important space entity of a city, the three-dimensional information of ground objects, especially buildings, is the important basic geographic information of the city and has important application value. The whole scanning process is high in data acquisition efficiency, ground control points do not need to be specially arranged, scanning measurement is slightly influenced by the external environment, and the measurement precision is high. Not only is automation degree ofbuilding space data acquisition improved, but also scanned point cloud can express real three-dimensional information and detail structure of a building in detail. Accurate contour information is acquired by means of the point cloud data, and a three-dimensional model of the building is reconstructed. The method has the advantages of high interaction performance, high intelligent degree, high expandability, high extraction speed, high contour extraction precision and the like.

The invention relates to an irregular surface-orientated self-adaptive projection system, and belongs to the technical field of computer image processing. The system comprises a self-adaptive projection device, adopts a projected image geometric correction method based on a geometrical shape with an irregular surface, and uses a novel colour structure optical encoding method, thus improving the interference-resistant capacity of an encoded image and increasing the reconfiguration speed of the irregular surface. By using the system, the rapid reconfiguration can be performed on the irregular surface, and the distortion correction can be performed on a projected image according to the reconfiguration result, so that an observer can observe the continuously consistent and undeformed projected image on the irregular surface, thus reducing the cost of the projection system, improving the intelligentization level of the projection system, and expanding the application range of the projection system.

The invention provides a data processing method based on a distributed storage system. The method comprises the following steps: defining a compressed configuration item that comprises at least a compression mode, off-line compression start time and off-line compression termination time; when to-be-processed data is written in, using a client or a server, or the client and the server to successively perform chain type compression processing on the to-be-processed data, and determining the time of the server intervening to perform the chain type compression processing only in the off-line compression mode according to the off-line compression start time and the off-line compression termination time; when the to-be-processed data is read, at least enabling the client to perform chain type decompression processing; after the chain type compression processing or the chain type decompression processing is performed, responding to an opposite end device through a network. After the data processing method based on the distributed storage system is used, the bandwidth utilization rate is increased, the memory space is saved, the data reconstruction speed is accelerated, and the data handling capacity of the distributed storage system is improved when a write operation or read operation is performed.

The invention discloses a compressive-sensing-based object reconstruction method. The main problem of incapability of detecting an object in image reconstruction in the prior art is solved. The method is implemented by the following steps of: 1) performing hybrid Gaussian modeling on the object to obtain the probability density of the object by using a mixed factor analysis model; 2) blocking the whole image to be reconstructed; 3) reconstructing each image block by utilizing the learnt probability density of the object as the priori knowledge of the object; and 4) splicing the reconstructed image blocks to recover an original integral image to obtain the whole reconstructed image. The object detection and image reconstruction are realized at the same time, so that the method has the advantages of resource saving and high efficiency, and can be used for object detection.

The invention discloses a method for achieving the parallel reconstruction of data on an RAID (redundant array of independent disk), particularly achieving the parallel reconstruction thereof in a sub-array distribution manner, wherein the sub-array comprises the blocks of different disks. The method comprises the following steps: respectively and simultaneously searching the other blocks constituting the same sub-array together with each block in a data disk; and reconstructing the data in the block according to the data in the other blocks. The method of the invention can shorten the process of data reconstruction by times, more particularly, all the operations of reconstruction, either reading or writing the related data, are parallel operations, and the requests from onstage users can be responded on a real-time basis in the reconstruction process, thus effectively reducing the average time for responding to the users, improving the reliability of the system and enhancing the performance thereof at the same time.

The invention discloses a rapid compression perception reconstruction method facing to a wearable device, and belongs to the signal processing field. The method comprises the steps of: performing compression measurement to an original electrocardiogram signal x through a constructed m*n dimension random sparse binary observation matrix phi, so as to get an observation vector y, the length of which is m; and based on the observation matrix phi and the observation vector y, reconstructing the original signal by the compression perception reconstruction method learned from the alternating direction multiplier method and the block sparse Bayes, so as to get an estimation vector of the original signal, which is as shown in the description. Compared with the conventional compression perception reconstruction method learned from the block sparse Bayes, the method provided by the invention is advantaged in reaching quicker reconstruction speed on the premise of ensuring the reconstruction precision.

The invention discloses a packet marking probability selecting method and a packet marking probability selecting device for inter-domain retrospect. A corresponding indicating bit of a data packet entering or leaving the current AS is checked at an interface of a BGP router of the current autonomous system AS, which is externally connected with an adjacent AS; when the indicating bit of the data packet shows that the data packet has not been marked by the upstream AS, a BGP routing table is checked in the BGP router according to a prefix longest matching principle, and an AS_PATH attribute value of a corresponding routing table entry is obtained, so the corresponding AS hop count is obtained; a marking probability is set according to the AS hop count to mark the corresponding data packet; and when the marking is determined to be carried out, the indicating bit of the corresponding data packet is set to be marked. The method and the device have the advantages of high speed of reconstructing attack paths, accuracy of reconstructing the attack paths and low router marking overhead.

The invention relates to a signal reconstruction method based on block compressed sensing, which can be used for comprehensively and systematically designing signal sampling and phase reconstruction technology based on block compressed sensing from sampling, block optimizing and reconstitution respectively, wherein the design for each part is simple and efficient in consideration of practicability of systematic design at the same time; the designed signal reconstruction method based on block compressed sensing has the following advantages: (1) the method can save the space of a system; the design reduces the storage space needed by a measurement matrix via adopting a method of block compression, simultaneously compresses the block signal at a sampling end when sampling and reduces the complexity of the sampling end; (2) a proposed reconstruction algorithm is fast in reconstruction speed and strong in robustness, besides, a system effectively overcomes the shortcoming of blocking effect in the block compressed sensing.

The invention relates to a rapid method and system for directly reconstructing a structured light illumination super-resolution image in an airspace, wherein the rapid method and system can be widelyapplied to researches in the fields of biology, medicine, material science, microelectronics and the like. Therefore, a problem that the super-resolution microscopic real-time dynamic observation is difficult due to complex operation and serious time consumption of a current structured light illumination super-resolution image reconstruction method is solved. Compared with a traditional frequency-domain SIM super-resolution image reconstruction method, the provided method has the advantages that the operation is completely carried out in the space domain and thus the processes of frequency spectrum separation, frequency spectrum movement, frequency spectrum fusion and the like are avoided and time-consuming multi-times forward / inverse Fourier transform operations are avoided, so that a problem of artifact noise caused by frequency spectrum operation is completely solved; a demodulation matrix for super-resolution reconstruction can be generated in advance, the related operations are only multiplication and addition operations and thus the process is simple and easy to implement, so that the reconstruction speed of the SIM super-resolution image is greatly increased.

The invention discloses a self-adaptive regularized smoothed l<0> norm method. A regularized SL0 algorithm is improved; in a steepest ascent method in an inner loop, a signal residual item estimated value iterative for the first time and a sparse signal estimated deviation value before and after the iteration are used as the selection basis of current regularization parameters; therefore, the signal sparse degree and the weight value of an error tolerance item in an outer loop every time can be adjusted self-adaptively; the balance of the two is kept in an optimization process; therefore, the reconstruction error of sparse signals can be effectively reduced; the anti-noise interference capability of the algorithm is improved; large-scale matrix inversion operation projected in operation of a feasible solution set in an iterative process can be avoided by introducing a SVD method; and the reconstruction speed to the sparse signals in the method disclosed by the invention is effectively increased.

The invention provides a remote sensing image super resolution method based on dictionary learning, comprising the following steps: S1, performing dictionary learning of corresponding feature types based on an image library of different feature types; S2, identifying the feature type of an original image; and S3, carrying out an image super resolution reconstruction process using a corresponding feature type dictionary. The invention puts forward a super resolution method based on dictionary learning using the idea of feature classification. As the number of atoms in a dictionary used in the reconstruction process is reduced greatly, the dictionary scale is reduced greatly. The proportion of effective atoms in the classification dictionary used is increased significantly, so that the number of effective atoms in the dictionary used in the reconstruction process is almost the same as a general dictionary, and the speed of reconstruction is improved significantly while the quality of reconstruction is ensured. The process only involves the adjustment of parameters, the computation efficiency is high, and the reconstruction quality of remote sensing images is high.

The invention discloses a synthetic aperture radar (SAR) imaging method based on a graphics processing unit (GPU). Through the SAR imaging method, based on compressive sensing theory, SAR imaging of a sparse scene is realized. The method is characterized in that: the method comprises the following steps: establishing a sparse SAR imaging module and determining a reconstruction algorithm of the module; based on the size of the SAR imaging, determining sizes of grids and thread blocks required by parallel parts of the reconstruction algorithm and allocating storage space, which is used for storing data in the reconstruction algorithm, for the GPU; and reading the observation data from outside by the GPU, performing sparse reconstruction on the SAR imaging based on the reconstruction algorithm, and returning reconstruction results. The parallel performance of the GPU is utilized in the SAR imaging method, and thus rapid reconstruction of the SAR imaging which is based on the compressive sensing theory is realized.

The invention discloses an asymmetrical disk array caching dispatching method, belonging to the technical field of computer storage. According to the method, a disk not involved in reconstitution is selected as a sacrificed disk, cache size of the sacrificed disk is reduced, and the cache size of a disk involved in reconstitution is correspondingly increased, so that number of times of input / output (I / O) requests to invalid disks and disks involved in reconstitution are reduced, the reconstitution speed of the invalid disk is improved, I / O response time of users is shortened, the availability and reliability of a storage system are improved, and the performance of the system is improved.

A non-vision-field imaging device comprises a near-infrared pulse laser used for emitting infrared pulse laser, a double-telescope module used for emitting the infrared pulse laser to an intermediatewall after beam expansion and collimation, and receiving the infrared pulse laser scattered back through the intermediate wall, and the double-telescope module is used for emitting the infrared pulselaser to the intermediate wall before the infrared pulse laser is scattered back by the intermediate wall.Wherein the single-photon detector is used for detecting the number of photons of infrared pulse laser scattered back through the intermediate wall, and the computer is used for recording the number of photons and the propagation time of the infrared pulse laser, counting the distribution of the number of photons in time and achieving three-dimensional reconstruction of the target on the basis of the distribution of the number of photons in time. According to the non-vision-field imaging device, a double-telescope system is adopted, coaxial receiving and transmitting of infrared pulse laser can be achieved, the difficulty of a reconstruction algorithm is lowered, and real-time trackingand positioning of a non-vision-field detection target can be achieved.

Embodiments of the present application provide a metadata management and data reconstruction method and device, and a storage medium. In an embodiment of that present application, the RAID technologyis used to organize and manage the storage resources in the storage system, Aiming at the hard disk domain in the storage system, the metadata in the hard disk domain is organized and managed according to the RAID technology. When the hard disk domain is created, the ID of the hard disk domain and the ID of each physical hard disk under the hard disk domain are stored in the attribute descriptionarea of each hard disk. When the storage pool is built, the relationship between the storage pool and the hard disk domain and the mapping relationship between the storage pool and the chunking groupcontained in the storage pool are stored in the attribute description area of each hard disk. Based on the metadata organization structure, when the physical hard disk in the hard disk domain has a hard disk failure, other blocks belonging to the same block group can be quickly indexed according to the information stored in the attribute description area, which is helpful to improve the reconstruction speed of the data stored in the physical hard disk.

The invention discloses a collaborative coding hyperspectral imaging system and image reconstruction method based on random projection. The reconstruction speed and reconstruction precision of hyperspectral data can be improved by using the system and the method. The system comprises a point-by-point scanning module, a beam splitter module, a first observation channel module, a second observation channel module and an image reconstruction module, wherein the first observation channel randomly generates an observation matrix A for a spectral information matrix constituted by received point spectral information, and performs compression coding on the spectral information by using the observation matrix A; the second observation channel randomly generates an observation sub-matrix for each piece of received point spectral information, and performs compression coding on the corresponding point spectral information by using the observation sub-matrix respectively; and the image reconstruction processing module implements reconstruction of a hyperspectral image according to the observation value sent by the first observation channel and the observation value sent by the second observation channel.

The invention belongs to the technical field of equipment safety assessment, particularly relates to a reconstruction method and device for a residual stress field of a ferromagnetic component, and aims to solve the problems that an existing reconstruction method for the residual stress field needs to treat a surface of the ferromagnetic component, the reconstruction period is long and the operation is complex. The method comprises the following steps: selecting to-be-measured points in an area, needing residual stress field reconstruction, of the ferromagnetic component; measuring each to-be-measured point through a nine-pole magnetic detection sensor to obtain corresponding first induction voltage and second induction voltage; obtaining a principal stress direction angle and a principalstress difference of each to-be-measured point; and obtaining the principal stress and the component stress of each to-be-measured point in an even number of rows in the point array by adopting a shear stress difference method, and reconstructing the residual stress field of the ferromagnetic component. According to the method, the surface of the ferromagnetic component does not need to be treated, the method has the advantages of being wide in application range, high in reconstruction speed and high in efficiency, and the quality of the ferromagnetic component and the safety and reliability of industrial equipment can be evaluated faster and more accurately.

The invention discloses a super-resolution structure illumination obvious micro-imaging method based on space-frequency domain hybrid reconstruction. The method comprises the following steps: 1) generating a stripe-shaped illumination light field; 2) illuminating and exciting a to-be-detected sample to generate a fluorescence signal by utilizing the stripe-shaped illumination light field under the conditions of different stripe directions and different phase shifting steps; 3) collecting fluorescence signals to obtain original fluorescence images excited by different stripe-shaped illumination light fields; and 4) if an original fluorescence image does not comprise a defocus background, processing the original fluorescence image which does not comprise the defocus background by using a space-frequency domain hybrid reconstruction algorithm to obtain a super-resolution image; and if an original fluorescence image comprises the defocus background, processing the original fluorescence image comprising the defocus background by using the space-frequency domain hybrid reconstruction algorithm to obtain a super-resolution image with a chromatography effect. According to the method, the calculation amount required in the reconstruction process is reduced by more than two times, and the reconstruction speed can be increased to more than 80 times of that of an existing super-resolution reconstruction algorithm.

The invention provides a resource creation method and device. A storage pool comprises K pieces of PD (Physical Disk) which are not dilated and M pieces of dilated PD. The method comprises the following steps: in a resource creation process, selecting multiple pieces of VD (Virtual Disk) used for creating resources from the K pieces of PD and the M pieces of PD, wherein the selected multiple pieces of VD are evenly distributed in the K pieces of PD and the M pieces of PD; and after resource creation is finished, carrying out equilibrium processing on the multiple pieces of VD in the K pieces of PD and the M pieces of PD to cause a use ratio of the K pieces of PD and a use ratio of the M pieces of PD to be kept at an effective range of an average use ratio. Through the technical scheme of the resource creation method and device, when any one piece of PD in the storage pool breaks down, a reconstruction process is triggered, at the time, the data of the failed PD is recovered on the basis of the data of all pieces of PD, an amount of the PD which participates in the reconstruction process is large, a reconstruction speed is greatly improved, reconstruction efficiency is not lowered by dilatation, the reconstruction pressure of the M pieces of dilated PD is lightened, and influence on business is reduced.

The invention discloses a non-convex compression preception optimization reconstruction method based on sketch representation and structured clustering. The method mainly settles a problem of inaccurate compression image reconstruction on the condition of low sampling rate. The method comprises the following steps of according to a sketch of an image, defining a sketchable block and a non-sketchable block, wherein the non-sketchable block comprises a smooth block and a patterned block, and the sketchable block comprises a unidirectional block and a multidirectional block; performing clustering based on sketching direction guidance on the unidirectional block; performing clustering based on a direction distribution characteristic on the multidirectional block; performing gray scale clustering on the smooth block and the pattern block; performing multi-measuring-vector observation on each kind of image blocks; and in reconstruction, executing a particle swarm optimization algorithm based on crossing and atom direction restraint according to multiple measuring matrixes, kind index and direction information of each kind of image blocks for obtaining a final reconstructed image. Compared with a TS-RS method and an NR-DG method, the non-convex compression perception optimization reconstruction method has advantages of high quality of the reconstructed image, high robustness and high suitability for reconstruction of a natural image.

The invention belongs to the technical field of semiconductors, and particularly relates to a software-defined on-chip system, a data interaction method and a system architecture, the system comprises a wafer-level interconnection substrate, a plurality of prefabricated members arranged on the wafer-level interconnection substrate, and a software-defined on-chip interconnection network used for interconnection among different prefabricated members, each prefabricated member is integrated with different function calculation particles according to field application function requirements; the prefabricated members follow a unified interface standard and a physical layer protocol specification, and data exchange is carried out between the prefabricated members through a wafer-level interconnection substrate and a software-defined on-chip interconnection network, so that different function calculation particles are recombined and / or rebuilt in a software-defined mode to adapt to different application requirements and task mapping requirements. According to the method, boundary conditions such as a design method, a calculation normal form, an implementation material and an integration mode of an existing integrated circuit are broken through, and flexible interconnection and functional reconstruction of different prefabricated members for field application are realized, so that complex and diversified application task requirements are met.

The invention discloses a three-dimensional magnetic flux leakage detection defect reconstruction method based on a generative adversarial network, and belongs to the technical field of nondestructivetesting. The method is characterized by comprising the following steps: step 1, acquiring a defect image by using a depth camera to obtain defect data; 2, adopting a magnetic flux leakage detection instrument, and obtaining magnetic flux leakage data through post-processing of the magnetic flux leakage data; 3, dividing all data into sample data and test data, dividing the sample data into training data and verification data; 4, matching the magnetic flux leakage data in the sample data with defect data; 5, obtaining a GAN final model; and step 7, obtaining a defect three-dimensional contour.According to the three-dimensional magnetic flux leakage detection defect reconstruction method based on the generative adversarial network, three-dimensional reconstruction is carried out on the bottom plate by using the GAN model, defect three-dimensional contour reconstruction can be rapidly carried out, the reconstruction speed is high, the stability is good, the precision is high, the robustness is good, and the reconstruction speed and precision of various irregular defects are facilitated.

The invention provides a quick reconfiguration method of live working site of a live working robot system, which comprises the following steps: establishing a standard working site parameter databaseof various distribution transmission lines, and constructing an index structure of the system database; Constructing standard three-dimensional model data taking origin coordinate system as reference coordinate system, establishing standard three-dimensional model of each component, and storing the modeling result data in standard three-dimensional model database; Construct standard working scenarios for live working tasks; The homogeneous transformation matrix of the origin coordinate system relative to the vision measurement coordinate system is obtained by establishing the vision measurement coordinate system of each component in the live working scene. The binocular camera attached to the end of the manipulator is used to measure the unstructured error of the live working scene, and the standard working scene model is modified and perfected to obtain the reconstructed scene which is consistent with the actual scene. The live working site constructed by the method coincides with the real site, and is convenient for the robot to carry out the related operation.

The invention discloses a method and a device for calculating a polar coordinate system matrix in a two-dimensional reconstruction algorithm. The method comprises the following steps: S100: establishing a polar coordinate projection model of a two-dimensional field of view to be reconstructed; S200, calculating the overlapping area of the projection beam and the polar coordinate grid at the i-th projection angle as a system matrix element value amn, wherein m in the subscript is a detector primitive number and n is a grid numb; S300: traversing and calculating all overlapping areas under the projection angle to obtain the system matrix A= {amn} of the two-dimensional field of view to be reconstructed. The system matrix under other projection angles can be obtained by adjusting the positionof the system matrix element value in the system matrix according to the change of the projection angle without repeated calculation. The method can improve the precision of algebraic iterative two-dimensional reconstruction algorithm, and avoid recalculating the system matrix when the projection angle changes, and improve the computational efficiency. Another aspect of the invention also provides an apparatus for use in the method.

The embodiment of the invention discloses a reconstruction method for an aerodynamic model in a ground flutter test, and the reconstruction method achieves the reconstruction of the aerodynamic modelthrough modal shape excitation, physical coordinate and modal coordinate conversion and interpolation point polycondensation. In a reconstruction process, vibration mode excitation is carried out on amode participating in flutter; obtaining an aerodynamic force influence coefficient matrix under frequency domain and modal coordinates; the method comprises the following steps: performing conversion of physical quantities under modal coordinates and physical coordinates, performing optimal configuration on interpolation points through a genetic algorithm to realize polycondensation of the interpolation points, and finally fitting frequency domain aerodynamic force to a time domain through a rational function fitting mode to obtain an aerodynamic force model meeting the requirements of a ground flutter test. The reconstruction method provided by the embodiment of the invention has the advantages that the precision is high, the calculation efficiency is high, and the model reconstructionprocess is simple and feasible; the test precision of the ground flutter test in a transonic region is improved, so that the effect of expanding the application range of the ground flutter test is achieved.

The invention discloses an imaging method for expanding a view field range based on calculation of a ghost imaging penetration scattering medium. The method comprises two processes of modulation and detection of a structured light field and image recovery. The light field modulation and detection uses digital micromirror equipment to perform binary modulation on a light source, and a modulation matrix is a randomized differential Hadamard matrix; and after differential operation is performed on signals in the image recovery process, target information is recovered by using an association algorithm. Ground glass and a shielding object with a hole are introduced between a detector and an object, the effect of limiting the view field range of the shielding object or the small hole is counteracted through the scattering effect of the ground glass; penetrating scattering medium imaging is achieved and the effect of expanding the view field range is achieved to a certain degree. The method provides the application prospect for calculating ghost imaging in a penetrating scattering medium; and when the field of view existing in a detection section is limited by occlusion of an occlusion object, field of view expansion imaging can be achieved through the method. The cost is low, the reconstruction speed is high, and the reconstruction quality is high.

The invention discloses a compressed sensing data reconstruction method based on randomized Kaczmarz iteration. The method comprises the following steps: firstly, adaptively changing the weight of each row vector in a measurement matrix, and calculating weighted measurement matrix rows; secondly, updating original data vectors to be reconstructed through the weighted measurement matrix rows in themanner of sparsely randomized Kaczmarz iteration; thirdly, processing the updated original data vectors to be reconstructed through hard threshold operators, reserving first k' elements with the largest absolute values, and zeroing remaining elements; and lastly, obtaining a final reconstruction result when the difference between errors of two adjacent data reconstruction results is smaller thana threshold. The method provided by the invention has the advantages that the weight of the measurement matrix is adaptively adjusted according to data characteristics, and the vectors to be reconstructed are updated through the weighted measurement matrix, so that reconstruction effects for key elements in original signals are strengthened, the compressed sensing data reconstruction precision isimproved, and compressed sensing data reconstruction is accelerated.