152results about How to "Improve refactoring efficiency" patented technology

A video encoding/decoding apparatus includes a video encoder for encoding block mode information indicating whether a block mode of a block to be encoded in a macroblock is a skip mode, partition mode information and skip motion information including index information, encoding at least one of transform type information, CBP information, and delta quantization parameter information of the block according to a combination of the block mode information and the partition mode information, and encoding coefficient information of the block based on the encoded information of the block, and a video decoder for decoding a bitstream to reconstruct block mode information indicating whether a block mode of a block to be decoded is a skip mode, partition mode information indicating whether block is divided into subblocks, and skip motion information including index information, and reconstructing block based on the reconstructed information by using reconstructed skip motion information.

The invention discloses a multi-task super-resolution image reconstruction method based on KSVD dictionary learning, which mainly solves the problem of relatively serious quality reduction of the reconstructed image under high amplification factors in the existing method. The method mainly comprises the following steps: firstly, inputting a training image, and filtering the training image to extract features; extracting image blocks to construct a matrix M, and dividing the matrix M into K classes to acquire K pairs of initial dictionaries H1, H2...Hk and L1, L2...Lk; then, training the K pairs of initial dictionaries H1, H2...Hk and L1, L2...Lk into K pairs of new dictionaries Dh1, Dh2...Dhk and Dl1, Dl2...Dlk by utilizing a KSVD method; and finally, carrying out super-resolution reconstruction on the input low-resolution image by utilizing a multi-task algorithm and the dictionaries Dh1, Dh2...Dhk and Dl1, Dl2...Dlk to acquire a final reconstructed image. The invention can reconstruct various natural images containing non-texture images such as animals, plants, people and the like and images with stronger texture features such as buildings and the like, and can effectively improve the quality of the reconstructed image under high amplification factors.

The invention provides a method for accurately reconstructing a microcosmic finite element grid model of a dissimilar material on the basis of CT (computed tomography) images. According to the method, sequence CT images are acquired through industrial CT, and micro-structural information in the CT images is mapped onto the reconstructed finite element grid model on basis of digitization and threshold segmentation, so that any detailed structural information in the dissimilar material can be represented in the reconstructed model. The method improves the reconstruction accuracy by means of contrast-limited adaptive histogram equalization, median filtering and pixel interpolation, and improves the reconstruction efficiency through image cut and pixel combination. With the method, rectangular (two-dimensional) and cuboid (three-dimensional) unit grid models with higher finite element analysis accuracy are directly reconstructed, error accumulation during reconstruction, grid partition and other links of the existing geometric reconstruction method is avoided, and reconstruction accuracy and efficiency are improved. The method can be widely applied to fields such as performance prediction and optimization design of dissimilar materials.

The present application relates to a vehicle trajectory reconstruction method and system and an electronic device. The method comprises the following steps: step a, collecting road network data, and constructing a road network topology diagram according to the road network data; step b, setting starting and end points of the trajectory range, and selecting the monitoring points of the video monitor devices within the trajectory range according to the road network topology diagram; step c: performing vehicle data retrieval according to the video monitoring device monitoring points within the trajectory range, and performing trajectory reconstruction according to the vehicle data retrieval result. In a network of a large number of video surveillance devices that have been deployed, accordingto the road network data and point of interest data, an adaptive algorithm for selecting monitoring points is established by using network community discovery, shortest path algorithm and other algorithms, and data search is carried out on the basis of selecting monitoring points, which avoids time-consuming analysis and processing of redundant video monitoring devices data and improves the efficiency of vehicle trajectory reconstruction.

Device and method for inserting error correcting codes and for reconstructing data streams, and corresponding products

The present invention relates to a device (1) for inserting error correcting codes into data streams (11) intended to be sent in packets in a network (8), a device (9) for reconstructing data streams (13), and corresponding methods and products.

The insertion device comprises a unit (3) for identifying resynchronization indicators dispersed in the data streams and a unit (4) for determining successive segments in these streams, the segments forming partitions of the data lying between two consecutive resynchronization indicators. It also comprises a unit (6) for allocating a set of error correcting codes to each of these segments, these sets of codes having decreasing sizes between any two of the consecutive resynchronization indicators. Preferably, the device also comprises a data interleaving unit (7).

The embodiment of the invention provides a method and device for reconstructing an independent redundancy array of inexpensive disks. The method includes the steps of copying data on the fault disk in the independent RAID to the standby disk in the RAID, and enabling degradation data to be written into the standby disk, wherein the degradation data comprise data required to be written into the fault disk after the RAID is degraded. The data obtaining speed can be effectively quickened, the data reconstruction efficiency is improved, whole-disk data reading on the other member disks except for the fault disk is avoided in the data reconstructing process, the I/O pressure on the other member disks can be reduced, influences on normal business I/O are reduced, and then the spending of an RAID group system is reduced.

The invention discloses a method for improving an erasure code based storage cluster recovery performance. The method comprises that a reconfiguration control node sends reconfiguration read requests to reconfiguration source nodes and sends reconfiguration write requests to reconfiguration destination nodes simultaneously, all reconfiguration source nodes implement reconfiguration read optimization according to reconfiguration read requests, all reconfiguration destination nodes implement reconfiguration write optimization according to reconfiguration write requests simultaneously, all reconfiguration source nodes read data participating in reconfiguration of this time from a disc respectively and divide the data into a plurality of data blocks respectively, all reconfiguration source nodes transmit the plurality of data blocks to corresponding reconfiguration destination nodes simultaneously through networks respectively, all reconfiguration destination nodes adopt a reconfiguration algorithm to reconfigure received data blocks coming from all reconfiguration source nodes respectively, and all reconfiguration destination nodes return data blocks obtained through reconfiguration to a client side respectively. By the aid of the method, the problem of transmission of a large amount of additional data caused by existing centralized decoding methods can be solved.

The invention discloses an uplink signaling-free NOMA system multi-user detection method based on deep learning. The method comprises the three steps of conducting modeling on received signals, conducting sparse representation on the received signals and training a deep structure for sparse signal recovery. According to the method, multi-user detection is conducted by using the structured compressed sensing theory by means of the sparsity of users in an NOMA system and the relation existing among transmitting signals of different time slots, the sparse reconstruction problem is solved by meansof the deep learning theory, low-complexity and high-performance multi-user detection can be achieved, and multi-user detection is achieved by using the structured compressed sensing theory by combining the correlation of users among the transmitting signals of different time slots; and the underdetermined inverse problem of signal recovery is solved by means of the deep learning theory, and thereconstruction precision and the reconstruction efficiency of a sparse reconstruction algorithm are improved from the three aspects including parameter optimization, sparse representation training andneural network training on the basis of the deep learning method.

The invention is a domain uniqueness-based method for realizing high efficiency structural light, producing a high efficiency coding phototemplate for a light projector in the structural light system and using the phototemplate for 3D reconstruction and comprising the steps of: (1) selecting different colors to compose a code set; (2) selecting elements from the code set, expressing each optical lattice by one color, making special code combination to produce the phototemplate and composing adjacent optical lattices into a coded word, which has the domain uniqueness; (3) corresponding the coded words to input and output beams of the light projector one to one; (4) corresponding the obtained image to the phototemplate to calculate the 3D coordinates. And a device for realizing the high efficiency structural light comprises light projector, vidicon and microprocessor, where the microprocessor comprises color set module, coding module, code word decoding module and 3D coordinates calculating module. The invention has high 3D sensing speed, high reconstructing efficiency and low use cost.

The invention provides a reconstruction method and device of an enhanced dedicated channel, belonging to the technical field of communications. The device disclosed by the embodiment of the invention comprises an acquiring module, a demodulating module, an obtaining module and a reconstructing module. The method comprises the following steps of obtaining antenna data; demodulating the antenna data to obtain the enhanced dedicated channel data of an offset user; obtaining a likelihood ratio of the enhanced dedicated channel data of the offset user and performing hyperbolic tangent judgment to obtain a judgment result; and reconstructing the antenna data of the offset user according to the judgment result to obtain the reconstruction data of the offset user. The embodiment of the invention can be used for reconstructing by obtaining the likelihood ratio and performing the hyperbolic tangent judgment, so that the reconstruction efficiency is improved.

The invention discloses a multi-point statistical three-dimensional geological model automatic reconstruction method based on a local search strategy. The method is characterized by directly using a two-dimensional transverse and longitudinal geological section to reconstruct a three-dimensional geological model, and according to the spatial cross relation of a known section, dividing a three-dimensional modeling space into several subareas, wherein each subarea is surrounded by N (the N is greater than or equal to 1 and is less than or equal to 6) subsections; for a unknown attribute positiondistributed in each subarea, firstly, according to the position, acquiring the N surrounding subsections, and then, according to a set neighborhood parameter and a known data point, customizing dataevents at X, Y and Z directions, using the corresponding data event to scan the subsection at the corresponding direction, and acquiring N condition probability density function; and then, using a reasonable probability fusion strategy to acquire joint probability density distribution, and finally acquiring the attribute value of the position. The method can be popularized and applied in various kinds of three-dimensional geological information systems, geographic information systems and geologic modeling and simulation systems.

The invention discloses a double-constant current source structure-based high-resolution fast time interval measurement circuit. Based on a time-to-amplitude converter, a double-constant current source structure composed of a discharging constant current source and a charging constant current source is adopted, and a charging and discharging capacitor is charged through the charging constant current source, so that the voltage of the charging and discharging capacitor can return to initial voltage. According to the double-constant current source structure-based high-resolution fast time interval measurement circuit of the invention, the charging constant current source has two kinds of working states; and under the control of on-off control signals, the magnitude of the charging current of the charging constant current source can be changed, namely, the charging constant current source optionally generates micro current according to measuring process requirements in an initial preparation measuring phase, a discharging phase and an ADC sampling stage, and so as to provide stable initial voltage for the charging and discharging capacitor in the initial preparation measuring phase, and after measurement is completed, the charging constant current source optionally generates high current to charge the charging and discharging capacitor, so that the voltage of the charging and discharging capacitor can be restored rapidly, and therefore, time interval measurement speed can be improved. With the double-constant current source structure-based high-resolution fast time interval measurement circuit adopted, the time interval measurement speed can be improved greatly with measurement accuracy ensured, and therefore, random equivalent sampling waveform reconstruction efficiency can be further improved.

The present disclosure provides a motion vector encoding/decoding method and apparatus using a motion vector resolution combination, and an image encoding/decoding method and apparatus using the same. A motion vector encoding apparatus include: an encoding cost calculator for calculating an encoding cost according to a motion vector resolution for each motion vector resolution combination including one or more motion vector resolutions; a resolution combination determiner for selecting a motion vector resolution combination by using the encoding cost calculated for each motion vector resolution combination; and a motion information encoder for outputting encoded motion vector data generated by encoding a motion vector determined according to the motion vector resolution combination selected, as a bitstream. Accordingly, an efficient motion vector resolution combination is selected adaptively and a motion vector is determined and decoded based on the motion vector resolution combination, thus making it possible to improve compression efficiency.

The invention discloses a code reconstruction method and device, computer equipment and a storage medium, and the method comprises the steps: firstly, obtaining to-be-classified code data, carrying out the feature extraction of the to-be-classified code data, and obtaining to-be-classified sample features of the to-be-classified code data; inputting the to-be-classified sample characteristics of the to-be-classified code data into a preset code classification model to obtain a target code reconstruction type of the to-be-classified code data; obtaining a corresponding target reconstruction script according to the target code reconstruction type of the to-be-classified code data; and finally, carrying out code reconstruction on the to-be-classified code data by adopting the target reconstruction script. Not only is the code reconstruction efficiency improved, but also automatic classification and reconstruction of to-be-classified code data of different code reconstruction types are realized.

The invention provides a voice signal reconstruction method and apparatus, wherein the method comprises: acquiring a sample voice signal and converting the same into a log-domain spectrum; extracting the content factor of each frame of sample voice signal by using a pre-trained content factor extraction model; extracting the speaker factor of each frame of sample voice signal by using a pre-trained speaker factor extraction model; training a content spectrum reconstruction model and a voiceprint spectrum reconstruction model based on the content factors and the speaker factors; inputting a target voice signal into the voiceprint spectrum reconstruction model to obtain a voiceprint spectrum in the target voice signal, and inputting the target voice signal into the content spectrum reconstruction model to obtain a content spectrum in the target voice signal; and adding the content spectrum and the voiceprint spectrum to obtain the reconstructed log-domain spectrum of the target voice signal. The voice signal reconstruction method and apparatus can decompose and reconstruct the voice signal based on the content and the speaker, has stronger correlation with tasks, and has high reconstruction efficiency. The reconstructed voice signal is relatively similar to the original voice signal.

The invention discloses an independent redundant disk array data reconstruction method, device and equipment. The method comprises the following steps of: after the RAID independent redundant disk array is added to the storage pool, after the RAID formatted storage unit is added to the logical volume, setting the used flag on the storage unit in the logical volume; when a disk fault exists in theRAID memory is detected, setting a fault flag on all storage units corresponding to the fault disk; traversing all the storage units corresponding to the fault disk; if the currently traversed storageunit is simultaneously provided with the used flag and the fault flag, reconstructing the current traversed storage unit, and after the reconstruction is completed, cancelling the fault flag of the reconstructed storage unit, and continue to traverse the next storage unit; otherwise, directly traversing the next storage unit until the fault disk traversal is completed. According to the independent redundant disk array data reconstruction method, only the used parts of the fault disk are reconstructed, the unused parts are not reconstructed, so that the reconstruction time is shortened, and the reconstruction efficiency is improved.

The invention discloses a method and a device for achieving RAID reconstitution. The method comprises the following steps: identifying that a bad sector exists in a disk of a RAID group; locking sectors of other disks except the disk where the bad sector locates in the RAID group whose location is the same as that of the bad sector, and sectors of the fault disk which causes a RAID reconstitution; respectively obtaining a first data in the sectors of the locked other disks and a second data in the sectors of the locked fault disk; verifying the data in the bad sector based on the obtained first data and second data as well as the priority of the RAID group; storing the verified data into a preserved disk sector in the disk where the bad sector locates, verifying a proof test value based on the verified data and the obtained first data, and storing the proof test value into a sector of a backup disk, wherein the location of the sector of the backup disk is the same as that of the bad sector. The technical scheme increases the efficiency of RAID reconstitution.

The invention discloses a built-in fingerprint reconstruction accelerator. The accelerator comprises a frame collector, a frame buffer unit, a fingerprint reconstruction unit and a global control logic. The frame collector is connected with the frame buffer unit by a write address bus set and a write data bus set; the frame buffer unit is connected with the fingerprint reconstruction unit by a read address bus set and a read data bus set; the global control logic is respectively connected with the frame collector, the frame buffer unit and the fingerprint reconstruction unit by a control bus; and an image buffer area is connected with the fingerprint reconstruction unit by an internal bus. The invention also provides a method for realizing dynamic reconstruction in the processing course. The accelerator and the method have the beneficial effect of lowering the excessive consumption of the system resources such as a host processor, a memory and the like under the condition of enhancing the fingerprint reconstruction efficiency, which causes balance configuration of software and hardware resources in a SoC system structure and achieves global optimization of the system.

An image processing method and an image processing device includes: receiving image data of a user transmitted by an endoscope, storing and splicing to obtain spliced image data; determining the block corresponding to the spliced image data, determine the position of the block in the stored three-dimensional image contour of the user, reconstructing the spliced image data at the corresponding determined position in the three-dimensional image contour of the user to obtain reconstructed three-dimensional image data and update the currently stored three-dimensional image model of the user, according to the stored three-dimensional image frame database; and displaying the updated three-dimensional image model of the user. The three-dimensional image of the oral cavity can be reconstructed without continuous and orderly scanning of the oral cavity, the dynamic display can be performed, the user interaction effect can be improved, and the self-service three-dimensional true color model selection can be supported.

The invention discloses a wind turbine bearing fault diagnosis method based on rapid multi-separation dictionary learning, and the method comprises the following steps: obtaining a wind turbine bearing signal, reconstructing a one-dimensional wind turbine fault signal to a high-dimensional phase space through phase space reconstruction, obtaining a matrix of a reconstructed signal, and forming a bearing fault diagnosis original feature set Y; partitioning the sample matrix, and dividing the sample matrix into L classes according to different classes through a K _ means classifier; optimizing and updating the ith class of matrix signals through separable dictionary learning, and outputting sparse coefficient separable dictionaries Ai and Bi; repeating the above steps to obtain a set of allsparse coefficients and dictionary matrixes, and performing reconstruction through a 2D _ OMP algorithm; performing integration processing on the signal Yi to obtain Y ', and outputting a signal Y' reflecting a fault characteristic so as to obtain non-stationary fault information of the wind turbine bearing; according to the method, the local spatial correlation in the high-dimensional signal is maintained, and the reconstruction efficiency is greatly improved.

The invention relates to a method for quantitatively reconstructing a complex boundary contour of an interlayer debonding defect based on an infrared detection image signal, which comprises the following steps of: firstly, building an infrared thermal imaging temperature field measuring system, processing and manufacturing a multilayer structure flat plate test piece containing the debonding defect; then measuring a transient temperature signal of an upper surface area of the flat plate test piece by using the built infrared detection system, and calculating a differential temperature variablerate signal image as a reconstruction characteristic signal; and secondly, based on a simulation calculation method of a transient temperature field numerical model and a defect reconstruction inverse algorithm, providing a parameterization method of the debonding defect complex boundary contour of a Fourier series approximation radial method so as to achieve efficient and accurate quantitative nondestructive evaluation of the debonding defect complex boundary contour of the multilayer structure. The method can efficiently and accurately reconstruct the boundary contour of the debonding defect, has the advantages of simple principle, convenient and easy operation, small data volume, high quantitative efficiency and the like, and can be widely applied to quantitative evaluation of the interface debonding defect of multilayer welding mechanisms in nuclear power and chemical industry.

The invention relates to an evolutionary password chip. The evolutionary password chip is manufactured by using the characteristic of infinite reconstruction of a programmable chip and the evolutionary design characteristic of intelligent calculation to dynamically design the algorithm of the key part of the password in real time. The password chip comprises the programmable chip and each part module of a DES algorithm arranged in the chip; each part module of the DES algorithm is connected with an intelligent calculation part; the intelligent calculation part comprises an adaptive value calculating module, an evolutionary operator module and a true random number generator which is connected with an evolutionary processing part by the evolutionary operator module. The algorithm of the key part of the password is realized by an intelligent calculation method.