130results about How to "Improve refactoring effect" patented technology

The invention discloses a deep learning super-resolution reconstruction method based on residual sub-images; residual sub-images are effectively combined with deep learning method based on convolutional neural network, super-resolution reconstructed images are clearer, and reconstruction speed is higher. By increasing the depth of convolutional neural network, a network model acquired by learning has higher nonlinear expression capacity and image reconstructing capacity; in addition, by introducing residual sub-image process, preprocessing based on traditional interpolation algorithm is removed, and fuzzy effect due to the interpolation algorithm is avoided. By making ingenious use of residual sub-images, it is possible to transfer deep learning convolutional operation from high-resolution space to low-resolution space, and accordingly reconstruction efficiency of super-resolution algorithm is increased at the premise of improving super-resolution reconstruction effect.

The invention discloses an adaptive compressed sensing-based non-local reconstruction method for a natural image. The problems of serious reconstructed image information loss and the like in the prior art are mainly solved. The method is implemented by the steps of: (1) dividing an image into N 32*32 sub-blocks, obtaining a basic sensing matrix Phi' according to a basic sampling rate b and a sensing matrix Phi, and sampling a signal by utilizing Phi' to obtain a basic observation vector; (2) estimating a standard deviation sequence {d1, d2, ..., and dN} of the image according to the basic observation vector; (3) adaptively allocating a sampling rate ai for each sub-block according to the standard deviation sequence {d1, d2, ..., and dN}, and constructing an adaptive sensing matrix, and sampling the signal by utilizing the adaptive sensing matrix to obtain an adaptive observation vector; (4) forming an observation vector of each sub-block by using the basic observation vector and the adaptive observation vector; (5) obtaining an initial solution x0 of the image according to the observation vector; and (6) performing iteration by using x0, and reconstructing the original image until consistency with a finishing condition is achieved to obtain a reconstructed image x'. The method has the advantages of high image reconstruction quality, clear principle and operational simplicity, and is applied to the sampling and reconstruction of the natural image.

An autonomous assembling technique of application software based on ITRON includes container in charging of interaction between software application component and operation environment and in charging of setting up application component according to design policy, domain configuration file for describing hardware unit and software component in system, automatic assembly / disposal tool for providing visualized editor of domain configuration file and automatic assembly / disposed tool being used to generate domain configuration file as required by container.

The invention provides a real image blind denoising method based on a deep residual network. According to the method, an RGB spatial clear image set is selected through an image dataset, and an RGB spatial image group set is constructed through spatial transformation; images under multiple scenes are shot through multiple cameras, and real image groups are constructed according to real clear images and real noisy images shot by each camera under each scene, and a real image group set is constructed; multiple RGB spatial image groups in the RGB spatial image group set and multiple real image groups in the real image group set are randomly selected to construct an image training set, and a preprocessed image training set is obtained through preprocessing; remaining RGB spatial image groups in the RGB spatial image group set and remaining real image groups in the real image group set are used to construct an image test set; and the preprocessed image training set is used as input to construct an image denoising residual convolutional neural network, the neural network is trained in combination with residual learning and a batch normalization strategy, and the image test set is denoised. The method has the advantages that convergence speed is high, and the denoising effect is good.

The invention discloses a dictionary database-based adaptive image super-resolution reconstruction method in the field of image processing. The dictionary database-based adaptive image super-resolution reconstruction method comprises the steps of: adaptively selecting a matched dictionary from a dictionary database according to a characteristic vector of each low-resolution image block, if the matching fails, re-training to obtain a proper dictionary, updating the dictionary into the dictionary database, then carrying out super-resolution reconstruction on the blocks by using the dictionary to obtain image blocks with high resolution, and finally, recombining all blocks to obtain a high-resolution image. The dictionary database-based adaptive image super-resolution reconstruction method is test in a face image, results prove that the method is superior to a method using a single dictionary in term of the image reconstructing effect, training image blocks with higher matching degree can be screened out in a process of training a local adaptive dictionary; and since many matched image blocks exist, prior information of a training set is sufficient amd a reconstructing effect is greatly improved compared with that of the method using the single dictionary.

The invention discloses a method and system based on a differential phase contrast imaging reduction quantitative phase image, and relates to the field of the computer imaging. The stable method from a target image to a reduction quantitative phase image is established. The method is capable of firstly establishing a differential phase contrast imaging two-dimensional optical phase transfer function H(u), and establishing the relation between the H(u), a frequency domain function of a differential phase contrast image (the figure is as shown in the specification) and the quantitative phase information, finally executing the deconvolution operation to recover the quantitative phase information through the phase transfer function of the differential phase contrast imaging. The method is capable of further researching a method of quantitatively recovering the phase information under the asymmetrical illumination pattern formed by multi-axis division so as to enhance the reconstitution capacity of the phase information in the different directions, and using the mathematical optimization to reduce the frequency noise increase caused by the direct deconvolution. The method is capable of overcoming the defects that the traditional quantitative image acquisition operation is complicated and the imaging condition is rigorous, and the obtained image resolution is higher.

The invention discloses a seismic signal restoration method based on dictionary learning regularization sparse representation. The method comprises the steps that S1 a tensor product method is appliedto a tensor dictionary learning process to construct an objective function; S2 an alternating iterative algorithm is used to solve a tensor sparse coefficient; S3 a tensor dictionary is trained by the Lagrange dual method; and S4 the tensor dictionary and the tensor sparse coefficient are iteratively updated to reconstruct a missing seismic signal. According to the seismic signal restoration method based on dictionary learning regularization sparse representation, firstly a t-product operator is introduced into tensor decomposition, and accordingly a new objective function is constructed; the ADM algorithm and the Lagrange dual algorithm are used to solve the sparse coefficient and the tensor dictionary; sparse coding of the seismic signal is realized; sparse representation of tensor data and recovery of the missing seismic signal are finally realized; and the effect of seismic data reconstruction is improved.

The invention discloses a method and a device for transmitting data in a unidirectional network. The method comprises the following steps: if original data to be transmitted are incontinuous medium information, grouping the original data into a data packet with fixed length at a transmitting end, carrying out outer code coding on the grouped data packet, then carrying out inner code coding on thedata packet after outer code coding and afterwards, delivering the data packet after inner code coding into a transmission path; judging the type of the medium information according to the load type of coding data at a receiving end; and if the type of the medium information is the incontinuous medium information, carrying out inner code decoding on the data packet received from the transmission path and then carrying out outer code decoding on the data packet after inner code decoding to obtain the original data. By applying the invention, the end-to-end reconstruction capability of data transmission is reinforced, and the reliability of the data transmission of the unidirectional network is enhanced, thereby achieving the aim of enhancing the communication quality of the unidirectional network effectively.

The invention provides a video signal collection and reconfiguration system based on high-dimension compressed sensing. The video signal collection and reconfiguration system comprises a complex sensing matrix construction and optimization module, a sparse base matrix construction module, a video signal universe sensing module and a reconfiguration processing module. An optimized complex sensing matrix and a high-dimension sparse base are respectively generated from the complex sensing matrix construction and optimization module and the sparse base matrix construction module with matrix operations of Kronecker products. Projection of video signals on the matrix is generated for the optimized complex sensing matrix by the universe sensing module. The acquired data are finally decoded and reconfigured in the reconfiguration processing module. While synchronous compressing and sampling in time-space domain are provided, distributed progressive structure during video sampling is adopted; accuracy and efficiency of reconfiguration are improved by corresponding optimization of the sensing matrix; sampling efficiency of the video signal is greatly improved; 4 dB sampling gain is acquired as the maximum value with varied sampling compressing rate; and the video signal collection and reconfiguration system has high scalability.

The invention discloses a compressive sensing method based on principal component analysis and mainly solves the problem of low sampling efficiency in the prior art. The method comprises the following steps of: (1) taking z images from a gray natural image library, taking a 32*32 sub-block from each image which is taken at intervals of three pixels along the horizontal and vertical directions to form a training sample set x1, x2, ..., and xm, and training a full-rank observation matrix Phi(f) for the training sample set x1, x2, ..., and xm by using a principal component analysis method, wherein z is not less than 15 and not more than 25, and m is the quantity of training samples; (2) dividing an image which is required to be sampled into n 32*32 sub-blocks x1, x2, ..., and xn, acquiring an observation matrix Phi according to sampling rate s and the full-rank observation matrix Phi(f), sampling each image sub-block by using the observation matrix Phi, and thus obtaining an observation vector y; (3) acquiring an initial solution x0 of the image according to the observation vector y; and (4) iterating according to the initial solution x0 until iteration is in accordance with end conditions, and thus obtaining a reconstructed image x'. The compressive sensing method has the advantages of high sampling efficiency, high image reconstruction quality and clear principle, and is easy to operate and applicable to sampling and reconstruction of a natural image.

The invention discloses a hyper-spectral compression perception reconstruction method based on nonlocal total variation and low-rank sparsity, and mainly solves the problems in the prior art that reconstruction accuracy is low and the effect is poor after compressed sampling of hyper-spectral data. The hyper-spectral compression perception reconstruction method comprises the steps that 1. the hyper-spectral data are inputted and vectorized; 2. the vectorized hyper-spectral data are sampled so that sampling data are obtained; 3. initial reconstruction of the sampling data is performed; 4. the initially reconstructed data are clustered; 5. the sampling data are classified according to the type of image elements so that various types of sampling data are obtained; 6. a secondary reconstruction model is constructed; and 7. The secondary reconstruction model is solved according to various types of sampling data so that the optimal data of secondary reconstruction are obtained, and the data act as the final reconstruction data. The idea of nonlocal total variation and clustering is introduced on the basis of low-rank sparse reconstruction so that the hyper-spectral compression perception reconstruction method has advantages of high reconstruction accuracy and great effect and can be used for hyper-spectral data imaging.

The invention discloses a spectrum image processing method based on partition compressed sensing. The method mainly solves the problems of traditional spectrum image processing methods that sampling cost is high and coding complexity is high. The method comprises the steps of conducting classification and difference operation on a spectrum image according to inter-spectrum correlation coefficients, conducting sampling by means of the partition compressed sensing technology, improving a traditional measurement matrix, and recovering the original spectrum image by means of the compressed sensing reconstitution algorithm. The method solves the problems of traditional spectrum image processing methods that sampling cost is high and coding complexity is high, and can effectively improve the reconstitution quality of the spectrum image under the condition that the sampling rate is unchanged.

The invention provides a scalable video encoding system based on hierarchical structure progressive dictionary learning. The system comprises a system framework based on a hierarchical structure, a progressive dictionary learning module and a scalable video frame reconstructing module. According to the system, due to a scalable B frame prediction structure, reconstructed frames are added into dictionary training as reference frames of a finer layer, and the complexity of a super-resolution algorithm based on learning is reduced through a random gradient descending method. Through the system, consistency of video frame movement can be effectively kept, and meanwhile space and quality are scalable based on the system frame of the hierarchical structure.

The invention provides a compressed video capture and reconstruction system based on structured sparse dictionary learning. The system comprises a structured sparse dictionary learning module, a video signal sensing module and a reconstruction processing module. The structured sparse dictionary learning module firstly acquires a training set through a sub-space clustering method, then, a dictionary is acquired through a linear sub-space learning method and minimized-block-relevant block sparse dictionary learning method, the sensing module projects video signals in an image block mode, and acquired data are finally decoded and reconstructed in the reconstruction processing module. Compressed sampling is provided, the distributed progressive structure of the video sampling process is combined, the reconstruction accuracy and efficiency are improved for the special structure of a structured sparse dictionary matrix, the sampling efficiency of the video signals is greatly improved, reconstruction gains are acquired compared with other methods under different sampling compression ratios, and meanwhile the good expandability is achieved.

The invention discloses a hydrocarbon fuel internal reforming fuel cell and gas turbine combined power generation system for an airplane, and belongs to the field of airplane power generation systems. According to the system, on the condition that the flexibility of the airplane is met, a hydrocarbon fuel cell is applied to supply power to the system, and the efficiency of the system is improved by adopting the mode of combining the hydrocarbon fuel cell with a gas turbine; hydrocarbon fuel is high in volume energy density and low in storage and heat exchange quality penalty; waste heat of tail gas is fully utilized, fuel which is not utilized by the solid oxide fuel cell is fully utilized to be burned in a burning chamber, the hydrocarbon fuel provides a high-temperature high-pressure working medium for a turbine after being burned, the turbine does work to generate power, the power serves as electric power for the airplane, and then the fuel utilization rate is greatly increased.

The invention provides a satellite-borne wireless information system which adopts a high and low speed wireless network mixed multi-layer networking, is used for acquiring, processing, fusing and transmitting multi-source information in real time and implementing management and control on each functional unit of a satellite and comprises a satellite house-keeping computer, a subsystem lower computer and terminal nodes, wherein the satellite house-keeping computer is connected with the subsystem lower computer by a high speed wireless network; the subsystem lower computer is connected with each terminal node by a low speed wireless network; and the satellite-borne wireless information system adopts a high and low speed mixed wireless network system structure. Therefore, by the satellite-borne wireless information system provided by the invention, full coverage of a bus on a whole spacecraft is implemented, non-blind-spot monitoring and control on the whole spacecraft are completed, restraints of communication network wiring to a satellite system structure and configuration layout are eliminated, a modularization degree of a satellite and flexibility of configuration layout are promoted, and 10% to 15% of an overall weight of the satellite can be reduced, so that modularization design is effectively ensured, and the satellite-borne wireless information system is adaptive to novel tasks such as on-orbit maintenance and service and the like.

The invention discloses a time-space video compressed sensing method based on a convolutional network, and mainly aims at solving the problems that in the prior art, the video compression time-space balance and the video reconstruction real-time performance are poor. The scheme of the method comprises the steps that a training data set is prepared; a network structure of a time-space video compressed sensing method is designed; training and testing files are written according to the designed network structure; a network of the time-space video compressed sensing method is trained; and the network of the time-space video compressed sensing method is tested. The network of the time-space video compressed sensing method adopts an observation technology of simultaneous time-space compression are conducted simultaneously and a reconstruction technology of using 'time-space blocks' to enhance the time-space correlation, not only can real-time video reconstruction be achieved, but also the reconstruction result has the high time-space balance, the reconstruction quality is high and stable, and the network can be used for compressed transmission of a video and follow-up video reconstruction.

The invention provides a compressed video acquisition and reconstruction system based on a data driven subspace set. The compressed video acquisition and reconstruction system comprises a subspace set construction module, a sparse basis matrix construction module, a video signal sensing module and a reconstruction processing module, wherein a clustering method is utilized by the subspace set construction module to generate the subspace set, a linear subspace learning method is utilized by the sparse basis matrix construction module to generate a sparse basis corresponding to the subspace set, video signals are projected in a block mode through the sensing module, and obtained data are decoded and reconstructed in the reconstruction processing module at last. Compressed sampling is provided, meanwhile a distributed gradual model structure in a video sampling process agrees with the system, and reconstruction accuracy and efficiency of a sparse basis matrix special structure is also promoted. Sampling efficiency of video signals is greatly improved, reconstruction gains are obtain compared with other methods under different sampling compression ratios, and meanwhile the system has good expandability.

The invention discloses an image super-resolution reconstruction method based on a genetic algorithm and a regular prior model, and mainly solves a problem of poor quality of a reconstruction result of a traditional method. The image super-resolution reconstruction method comprises the following implementation steps: (1) learning one group of sub-dictionaries from a natural image; (2) obtaining the luminance component estimation X of a high-resolution image Xs after a low-resolution (LR) image is magnified through interpolation by three times; (3) constructing an initial population; (4) calculating the fitness value of each individual; (5) selecting and copying the individuals in a parent population; (6) successively carrying out cross and variation to the individuals of the parent population; (7) repeating the steps (5) and (6) for twenty times to obtain an optimal solution X'; (8) carrying out local optimization on the X' by utilizing the regular prior model; and (9) repeating steps (3) to (8) for four times to obtain a luminance component X2 of a high-resolution image, and finally combining the high-resolution image. Image edges and texture information can be favorably kept, and the image super-resolution reconstruction method can be used for image identification and target classification.

The invention provides an underwater robot and a multifunctional underwater operation device, and relates to the technical field of robots. The underwater robot comprises a head cabin, at least one standard cabin, a tail cabin and a plurality of clamping fasteners for connecting of the different cabins; the clamping fasteners comprise first connecting pieces and second connecting pieces, each first connecting piece is provided with a first end and a second end which are opposite, and each second connecting piece is fixed to one end of the corresponding first connecting piece; the second end ofeach first connecting piece is provided with a protruding part, and a groove part which is matched with the corresponding protruding part is formed between the first end of each first connecting piece and the corresponding second connecting piece; and when the clamping fasteners are installed on the standard cabin, the outer wall of each first connecting piece is flush with the outer wall of thestandard cabin, and the corresponding second connecting piece is embedded in the standard cabin. The multifunctional underwater operation device comprises the underwater robot, both the underwater robot and the multifunctional underwater operation device have the advantages of reliable connection, simple mounting and good reconfiguration between modules.

The invention discloses a bidimensional compressed sensing image acquisition and reconstruction method based on discrete cosine transformation (DCT) and discrete Fourier transformation (DFT), belongs to the technical field of designs of measurement matrixes and optimization of reconstruction matrixes in the compressed sensing process and provides a method for firstly determining the measurement matrix and a sparse matrix and then optimizing the reconstruction matrix. In a measurement stage, the 0-1 sparse matrix is adopted; in a reconstruction stage, a Gaussian matrix is adopted; and therefore, an after-optimization method capable of easily implementing hardware and guaranteeing a signal reconstruction effect can be realized. The method comprises the following steps of: performing row vector orthogonal normalization and column vector unitization on the reconstruction matrix obtained by the (i-1)th iteration calculation through ith iteration, optimizing the reconstruction matrix on the basis of maximum values of absolute values of relevant coefficients among row and column vectors, the convergence stability of row vector modules and the number of rows and the number of columns which obey the Gaussian distribution, and finishing the after-optimization on measurement data subjected to one-dimensional and two-dimensional sparse transformation and the measurement matrixes by calculating a transitional matrix and a proximity matrix. The method lays a foundation for the compressed sensing from theoretical research to industrialization.

The invention provides a heterogeneous filter bank filtering method based on a tree structure; the method comprises the following steps: using a dual-channel FIR quadrature mirror filter bank as a basic module, and employing the tree structure to build a heterogeneous filter bank; deriving reconstruction conditions of the whole tree structure heterogeneous filter bank, and resulting the whole system complete reconstruction conditions as to design a prototype filter satisfying specific conditions; using an iteration passband cut-off frequency to enable the designed prototype filter to satisfy reconstruction conditions. In a prototype filter special designing process, the method can employ a discrete weight square error criterion; compared with other methods, the criterion can better realize the passband flat characteristics and bigger stop band attenuation; the design scheme can effectively reduce the iteration complexity, can obtain bigger stop band attenuation and excellent system reconfiguration performance, thus providing important application values for a radar broadband channelization receiver and voice and picture signal processing.

The invention relates to a monitoring scene-oriented image super-resolution method and device, and a storage medium, and belongs to the technical field of image processing. The method comprises the steps of inputting a target low-resolution image into a pre-trained feature mapping network to obtain high-dimensional features located in a target feature space; inputting the target low-resolution image and the high-dimensional features into a pre-trained image reconstruction network to obtain a high-resolution image; the problems that a low-resolution image synthesized through an existing image super-resolution method based on deep learning is different from a real low-resolution image, and generalization is poor can be solved. The feature mapping network learns a feature mapping relationshipin advance; and the image reconstruction network is trained by combining the output result obtained after the feature mapping network, so that the domain difference between the synthesized low-resolution image and the real low-resolution image is further reduced, and the image reconstruction effect is improved.

The invention relates to a compressed sensing reconstruction method suitable for microgrid harmonic wave monitoring. The compressed sensing reconstruction method includes the steps that it is supposed that theta=phipsi, initialization on fundamental wave filtering is carried out, fundamental wave filtering is carried out, fundamental wave contents in compressed sampling values are filtered, parameter initialization is carried out on a spectrum projection gradient method, the compressed sampling values yharmonic of harmonic components serve as input amount with the spectrum projection gradient method, and sparse vector estimated values sharmonic of the harmonic components are reconstructed to reconstruct microgrid harmonic wave original signals x. By means of the compressed sensing reconstruction method, fundamental wave filtering is carried out on the microgrid harmonic wave compressed sampling values to obtain the sparse vector estimated values of the fundamental wave contents and microgrid harmonic wave compressed sampling values (only containing the harmonic wave components) after the fundamental components are filtered, the harmonic signal reconstruction effect is effectively improved, and the compressed sensing reconstruction method is suitable for microgrid harmonic wave monitoring.

The invention discloses an interactive C language based multi-bus modularization robot controller. The controller is characterized by comprising N extension equipment, N first sensor execution units, a main controller and an interactive C language programming module, wherein the N first sensor execution units are connected with extension ends of the N extension equipment respectively, and N is a natural number more than 0; the input end of the main controller is connected with the input ends of the N extension equipment; and the control end of the main controller is connected with the interactive C language programming module by a network bus. Due to the adoption of the structure, the controller avoids the defect of high requirement on the programming language capability of a student; andmeanwhile, due to the ideas of cascade connection and modularized design, the extensibility and the reconstruction of the robot are improved.

The invention discloses a minimization maximum design method for alternating DFT modulation filter banks. The requirement for oversampling of reconstructed characteristics and frequency characteristics of the DFT modulation filter banks is converted into a function relative to prototype filter coefficients, then a design problem of alternating the DFT modulation filter tanks is converted into a constrained optimization problem, and finally minimization maximum algorithm is adopted for obtaining the optimal coefficient of the prototype filter. An aliasing component compensation mode is introduced into the design of oversampling of the DFT modulation filter banks for the first time for controlling aliasing errors, so that aliasing errors are effectively constrained, the reconstruction performance is improved, and the high-aliasing problem caused by stopband attenuation is avoided.

The invention provides a compressed video capture and reconstruction system based on a data drive tensor subspace. The compressed video capture and reconstruction system comprises a tensor sparse base structure module, a video signal sensing module and a reconstruction processing module, wherein the tensor sparse base structure module utilizes a tensor subspace learning method to generate a sparse base matrix corresponding to the tensor subspace, the video signal sensing module projects a video signal in a tensor block mode to obtain an observed value, and the reconstruction processing module receives the sparse base matrix and the observed value and performing decoding reconstruction on all dimensionalities of a tensor signal respectively. The compressed video capture and reconstruction system provides compressed sampling, meanwhile conforms to a distributed progressive structure in the video sampling process, and also improves the reconstruction accuracy and efficiency of the special structure of the tensor sparse base matrix. The compressed video capture and reconstruction system greatly improves the video signal sampling efficiency, obtains reconstruction gain at different sampling compression rates compared with other methods, and meanwhile has good expandability.

The present invention provides a blind reconstruction method under a modulation broadband converter based on the sparse Bayesian, and is used for the technical field of reconstruction of compressed sensing signals. The problem is solved that a reconstruction method under a current modulation broadband converter is poor in reconstruction performance when the signals contain noise. The method comprises the steps of: multiplying input sparse signals by a pseudo-random sequence, performing low-speed sampling and filtering operation for the signals obtained through multiplying, constructing an observation matrix to show the signals to a representation of compressed sensing, adopting the sparse Bayesian method to estimate the signals in the recovery, and obtaining a variance [gamma] of the inputsparse signals through iteration by employing an EM algorithm to complete reconstruction of the sparse signals. In the condition that the signal-to-noise ratio of the signals is -15dB, compared to the prior art, the reconstruction method provided by the invention can reduce the steady-state mean square error value above 75% so as to effectively improve the reconstruction performance. The blind reconstruction method can be applied to the reconstruction field of the compressed sensing signals.

The invention discloses a dynamic reconfigurable universal ground measurement and control equipment based on a communication protocol, and a signal input and output control method thereof. Through adopting universal, intelligent and miniature design, the dynamic reconfigurable universal ground measurement and control equipment has the advantages of simple structure, uniform configuration, high compatibility and high maintainability, can realize dynamic reconfiguration from a module level to a single-machine level, and is convenient for function expansion as well as maintenance and replacementto meet the testing requirements of different stages and different working conditions, improves equipment universality and model compatibility, improves the aerospace ground test efficiency and mission reliability, adopts universal, miniature and modular design for measurement and control equipment through integrating the measurement and control requirements and measurement and control resources,realizes the equipment combination configuration oriented to functional requirements, satisfies the usage requirements of fast maintenance and replacement, integrally applies the backplate bus technology, satisfies the cascading expansion from the module level to the single-machine level and the test requirements of different stages and different working conditions, adopts an independent bus for monitoring health of board card status, and improves the intelligence level and health management level of the measurement and control equipment.

The invention provides a device for improving the reconstruction performance of a virtual disk group. The device comprises a reconstructed virtual disk group VDG selection module and a VDG reconstruction module. The reconstructed virtual disk group VDG selection module is used to select a VDG needing reconstruction when the number of VDGs in reconstruction supported by a system is not exceeded currently. When there is a VD in reconstruction in a PD where the VDs of the selected VDG needing reconstruction are located, the VDG is not reconstructed. When there is no VD in reconstruction in the PD where the VDs of the selected VDG needing reconstruction are located, the VDG is added to a reconstruction queue. The VDG reconstruction module is used to reconstruct VDGs in the reconstruction queue. By adopting the scheme, the memory consumption and the impact on the business are reduced on the basis of improving the reconstruction performance.