161results about How to "Improve recovery accuracy" patented technology

The invention relates to an underwater three-dimensional reconstruction device and method based on a structured light technology and a photometric stereo technology. The underwater three-dimensional reconstruction device provided by the invention comprises a cuboid-shaped frame, wherein a sliding rail is arranged at the top of the cuboid-shaped frame, a laser transmitter and a high-definition camera are arranged on the sliding rail, a rotatable circular sliding rail is arranged below the sliding rail, and the light ray pattern of the laser is a line pattern. At least six LED parallel shiny side light source lamps are arranged on the circular sliding rail and are white light lamps, and spaces between every two adjacent LED parallel shiny side light source lamps are equal. The underwater three-dimensional reconstruction device provided by the invention can carry out underwater laser structured light three-dimensional reconstruction and underwater photometric stereo three-dimensional reconstruction. The underwater three-dimensional reconstruction method provided by the invention comprises the following steps: calibrating equipment in the device and water quality conditions, detecting a target object in a target water area, namely carrying out laser structured light detection, photometric stereo detection and fusion of two detection results, and realizing high-accuracy three-dimensional imaging of the target area. The underwater reconstruction device and method provided by the invention can realize multiple underwater reconstruction methods; and compared with the previous methods, the object boundary restoring accuracy is greatly improved, and more accurate three-dimensional information is obtained.

The invention provides a method for Bayes compressed sensing signal recovery based on a self-adaptive measurement matrix and relates to the field of the information and communication technology. The method aims at solving the problem that an existing compressed sensing signal recovery method is low in accuracy. Based on the design of the self-adaptive measurement matrix in compressed sensing and combined with the Bayes compressed sensing algorithm, a design scheme of the compressed sensing method is obtained. The method is characterized in that the designed measurement matrix can be generated in a self-adaptive mode according to different signals, the purposes of determinacy and storage of the matrix are both achieved, and combined with the Bayes compressed sensing recovery algorithm of a relevant vector machine, the priority of a layered structure is introduced. The design scheme passes simulation verification, it is confirmed that the good signal recovery effect can be obtained, and the error range of recovered signals can be evaluated. The method is suitable for wireless signal transmission occasions in the information and communication technology.

A method, a device and a system for clock synchronization are provided. The method includes: under a first transfer mode, generating a first receiving timestamp by using a first link, and acquiring a first transmitting timestamp of a master clock device; under a second transfer mode, generating a second transmitting timestamp by using the first link, and acquiring a second receiving timestamp of the master clock device; and acquiring a time difference to the master clock device according to the first receiving timestamp, the first transmitting timestamp, the second transmitting timestamp and the second receiving timestamp, and realizing a clock synchronization with the master clock device by using the time difference. A slave clock device and a master clock device are provided in the embodiments of the present invention, as well as a system for clock synchronization.

The invention provides a Bayer color image interpolation method based on direction flag bits. The method includes the steps that firstly, unknown green components are restored, gradients of each pixel point in four directions are detected through edge detection operators, direction flag bit information of each pixel point is determined according to the detected gradients, the chromatic aberration coefficient of each pixel point and an adjacent pixel point is calculated, and interpolation restoration is conducted on the unknown green components according to the direction flag bit information and the chromatic aberration coefficient. After all the green components of a whole image are restored, all unknown red and blue components are finally restored based on a chromatic aberration method, and then color interpolation conducted on the Bayer image is finished finally. Compared with a traditional method that single-channel independent interpolation is conducted only by detecting horizontal and perpendicular edges, on the basis that the low complexity and the high image quality of an existing method are kept, interpolation accuracy is further improved, and the method is more suitable for human eye observation and provides a foundation for high-accuracy image restoration in the following process.

The invention discloses a refined restoration method for an ancient karst landform. According to the method, spectra inversion is carried out by using original seismic data and logging data of a target stratum, thereby obtaining a seismic reflection coefficient data volume; positions, at the reflection coefficient data volume, of a top boundary and a bottom boundary of the target stratum at the logging position can be calibrated according to the logging data, tracking is carried out automatically by using the calibrated top boundary and bottom boundary as seed points, thereby obtaining the positions of the top boundary and bottom boundary of the whole target stratum at the reflection coefficient data volume; subtraction of the top boundary and the bottom boundary of the whole target stratum is carried out to obtain a residual thickness of the target stratum; and then the ancient karst landform of the target stratum is restored by using the residual thickness method. With the method, the restoration precision of the ancient karst landform can be improved and the exploration economic benefits of the karst reservoir can be improved effectively.

The invention provides a lake basin ancient landform restoration-based quantitative method. The method comprises the following steps of 1, establishing a high-precision isochronal stratigraphic framework in a recessed area; 2, selecting a marker bed, and determining the thickness of a two-section residual stratum in the recessed area; 3, adopting the mudstone sound wave time-difference method, andcorrecting the residual stratum thickness of a target layer to be the actual stratum thickness; 4, carrying out stratum inclination angle correction; 5, subjecting the actual formation thickness of arecovery layer to de-compaction treatment so as to determine the initial deposition time thickness; 6, restoring ancient water depth values at a plurality of well points through rock core analysis and thickness measurement, and converting the thickness of the stratum into the distribution of the paleo-water-depth according to the thickness correspondence relation between the paleo-water-depth andthe thickness of the initial sedimentary period, namely quantitatively restoring the lake basin shape. According to the invention, not only the influence problem of stratum dip angle, stratum erosionand stratum compaction effect on the thickness is solved, but also the lake basin shape can be quantitatively represented. The distribution of the paleo-water-depth is quantified. The precision problem when the lake basin shape is recovered is solved. The recovery precision of the lake basin shape is improved.

The invention discloses a novel offshore oil spill processing device based on double crawlers. The novel offshore oil spill processing device is characterized by comprising a ship (1), the ship (1) comprises a ship head (7) and a ship body (5) for containing a photoelectric oil-water separation chamber (2) and an oil tank (3), the ship head (7) is bottomless, a belt impeller (6) for guiding offshore spilled oil into the ship body (5) is arranged between two inner walls of the ship head (7), an oil spill weir (8) is arranged behind the belt impeller (6), and oil guide arms (12) are symmetrically arranged on the ship head (7); the oil tank (3) is arranged behind the photoelectric oil-water separation chamber (2). By means of the novel offshore oil spill processing device, sea surface oil dirt is rapidly removed, the cleanness and sanitation of ocean are protected, the oil dirt removing method is high in work efficiency, oil dirt is thoroughly removed, pollution is small, and the device can be used for removing sea surface oil dirt in a large area.

The invention discloses a transportation data loss recovery method based on tensor reconstruction. The transportation data loss recovery method based on the tensor reconstruction aims to resolve the problem that precision is low and loss in a plurality of days can not be processed when an existing traditional transportation data loss recovery method based on a vector or a matrix form is used for recovering loss data. The transportation data loss recovery method based on the tensor reconstruction comprises that (a) transportation data are set in a multi-dimensional tensor form, loss tensor data are expressed through marked tensor, (b) the tensor data are spread on each mode, the relevance of all modes is calculated, and the weight of each mode is obtained, and (c) an objective function of loss data value recovery is set up and the loss data value of the objective function is solved according to the set tensor data and the calculation of the weight of each mode. The transportation data loss recovery method based on the tensor reconstruction is based on a multi-dimensional tensor model, all transportation time-space information is contained, the relevance of multi-mode is fully utilized, at the same time the original structure of multi-dimensional properties and the like of the transportation data is maintained, recovery precision is obviously superior to the traditional recovery method based on the vector or the matrix form, and an extreme case of the loss of a plurality of days can be solved well.

A parameter channel error code resisting method of vocoder line spectrum includes carrying out odd-even calibration on parameter and writing calibration bit into most non-important parameter bit of vocoder at coding end, picking light voiced sound parameters and integrating said parameters with maximum post verification probability algorithm to carry out recovery estimation for raising parameter recovery accuracy by line spectrum then picking line spectrum to make relevant odd-even calibration on parameters at decoding end.

The invention provides an image restoration method based on an aerial TDI-CCD (Time Delay and Integration-Charge Coupled Device) imaging error vibration model. In the method, vibration frequencies are classified into low-frequency vibration and high-frequency vibration without regard to high-frequency vibration; the low-frequency vibration is taken as random vibration, and a low-frequency vibration model is established; the low-frequency vibration is decomposed into displacements in three direction of X, Y and Z on the basis of the low-frequency vibration model, and displacement expressions in the directions X, Y and Z are established; the high-frequency vibration is taken as the simple harmonic vibration of a set frequency in a set direction, and a high-frequency vibration model is established; and imaging errors generated by displacement changes caused by the low-frequency vibration in each direction are respectively corrected by adopting different image restoration algorithms in each direction when software image restoration is carried out. The method can more reasonably simulate the actual situation of aerial imaging, and the models established by the invention can improve the precision of the TDI-CCD image restoration.

The invention discloses a superimposed CSI (Channel State Information) feedback method based on a deep learning large-scale MIMO (Multiple Input Multiple Output) system. A user side reads a segment ofCSI and a segment of uplink data; spectrum spreading processing is carried out on the CSI to obtain a spread spectrum sequence, and digital modulation is carried out on the uplink data to obtain an uplink modulation sequence; weighted superimposition is carried out on the spread spectrum sequence and the uplink modulation sequence to obtain a superimposed sequence, the user side transmits the superimposed sequence, and a base station side receives to obtain a receiving sequence; models HDNet are constructed, and the models HDNet comprise a CSI estimation model and an uplink data detection model; model parameters are initialized, the models HDNet are subjected to offline training, and after error convergence, the models are stored; and a signal is received online to obtain an online receiving sequence, an estimated value of the CSI is recovered by the trained models HDNet according to the online receiving sequence, and a detection value of the uplink data is detected out. According tothe invention, feedback does not require additional spectrum cost, recovery accuracy of the feedback CSI can be improved, and processing complexity of the system is reduced.

The invention discloses a multisource signal collaborative compressed sensing data recovery method comprising the steps as follows: 1, an aggregation node in a wireless sensor network acquires historical data of each sensor node, and computes a sparse structure information matrix by using the historical data and a first optimization equation; 2, the aggregation node receives a compressed sensing measurement result matrix that is to be processed and is transmitted by each sensor node; and 3, the aggregation node performs a data recovery operation on the compressed sensing measurement result matrix to be processed of each sensor node by using the sparse structure information matrix computed in the step 1, wherein the recovery data is a recovery result corresponding to current data to be transmitted of each sensor node after compressed sending measurement. According to the method provided by the invention, the sparse structure information matrix is trained by the compressed sensing measurement result matrix of the historical data, and the training process considers information loss during the compressed sensing measurement process, and thus the subsequent data recovery accuracy is improved.

The invention provides a method and device for recovering deleted records of a SQLite database file. The method for recovering deleted records of a SQLite database file comprises the following steps: analyzing the SQLite structure; obtaining the SQLite file encoding mode and the page size information; creating the base library of the field number and type of each table and realizing the association of the free pages with the base library; judging whether the free-block is covered or not and checking the integrity of the free-block; classifying the missing cases combined with the known data parameters which can be extracted from the free-block; and refactoring the control header information of the deleted records quickly and realizing the recovery of the SQLite deleted records. According to the invention, the base library of the field number and type of each table is created; the integrity check of the free-block is performed, so that the recovery accuracy rate is increased; the Rowid width of the free-block is calculated by means of the Rowid of the adjacent data unit that has not been deleted; the first four bytes in the memory structure of the free-block are calculated and reconstructed by means of the free-block size. Therefore, the data recovery of the SQLite deleted records can be realized generally, quickly and accurately by means of the invention, and the recovery rate and the recovery accuracy can be improved compared with the traditional method.

The invention discloses a channel estimation method based on compressed sensing and deep learning, a medium and equipment. The method comprises the steps that a base station end of an orthogonal frequency division multiplexing system sends a signal to a user end in a downlink in a comb-shaped pilot frequency form; the user side in the downlink obtains the sent pilot frequency receiving signal y and feeds back the signal y to the base station side; the base station end of the orthogonal frequency division multiplexing system performs channel estimation based on an ASJOMP algorithm of compressed sensing according to the obtained receiving pilot signal y by using the structured sparse characteristic of a time delay domain sparse channel to obtain an initial estimation channel, builds a noise reduction neural network based on deep learning, and trains a DnNet network by using an existing sample to obtain a network parameter theta; and de-noising the obtained initial estimation channel according to the DnNet network to obtain finally estimated channel state information. A lightweight network is adopted, so that the training process is faster, the calculated amount is reduced, and finally, an accurate CSI estimation value is obtained.

The invention provides an image fragment restoration method. The method includes the steps: 1, acquiring head fragments from a to-be-restored storage medium and forming a head fragment collection according to a head signature corresponding to an image in the format needing to be restored, and simultaneously obtaining a non-head fragment collection of the image from the to-be-restored storage medium; and 2, matching each head fragment in the head fragment collection with a corresponding next non-head fragment in the non-head fragment collection, combining the current head fragment with the next non-head fragment thereof, if a file obtained after combination is a complete image file, outputting the complete image file, and otherwise, adding the file obtained after combination in the head fragment collection to serve as a new head fragment. The image fragment restoration method has the advantages of high restoration accuracy and efficiency, and is applicable to restoration of image fragments in various formats and high in practicality.

A method of recovering timing information in a packet network is disclosed wherein a modulation scheme is used to transport additional information required for clock recovery between the sender and receiver across the network.

The invention discloses an electric quantity data recovery method based on joint weather information matrix decomposition. Weather data is used for providing context information, and joint matrix decomposition is used for recovering missing electric quantity data. An electric matrix behavior date is listed as a user, and a weather matrix behavior date is listed as weather attribute. The electric quantity matrix and the weather matrix share data information, and the feature matrix of the date is shared after decomposition. A joint decomposition optimization target is designed, a joint decomposition optimization model is established, the optimization model is solved on the basis of gradient descent iterative update, and an electric quantity recovery algorithm based on the joint matrix decomposition is designed. An experiment result indicates that the method exhibits higher recovery accuracy than an NMF (Nonnegative Matrix Factorization).

The invention discloses an MIMO channel state information feedback method based on anti-fitting deep learning, and belongs to the field of communications. The method comprises the following steps that: firstly, an AOCN model is constructed, a channel matrix is divided into a real part and an imaginary part which are then input into an encoder of a user side, the encoder comprises a convolution layer and a full connection layer, data reaches a receiving end through a feedback link after being encoded, a decoder at the receiving end comprises an anti-fitting layer, a full connection layer, a RefineNet layer and a convolution layer, and finally a predicted channel matrix is output. After the AOCN model is constructed, offline training is performed on the model, model parameters are initialized firstly, the model is stored after error convergence, and finally, the trained and stored AOCN model is subjected to channel state information prediction online. According to the invention, the recovery precision of the information matrix can be further improved, the transmitting end of the system is ensured to obtain accurate channel state information, and the communication quality of the system is improved.

The invention belongs to designing and manufacturing of impacting-resistant platforms and relates to a three-freedom-degree angular-displacement-free impacting-resistant platform. The three-freedom-degree angular-displacement-free impacting-resistant platform comprises a parallelogram angular-displacement-free device and an impacting-resistant rubber block. According to the impacting-resistant platform, a rubber elastic component serves as an impacting-resistant buffer, and a parallelogram angular-displacement-free platform serves as an angular-displacement-limiting mechanism. The parallelogram corner-free platform is simple in structure, the vulcanization process of the rubber impacting-resistant rubber block is mature and has been widely used for various types of equipment; and meanwhile, because of the nonlinear characteristics of a rubber material, reliability is high. Because an oblique metal block is used in the buffering rubber block to decompose the impacting energy, buffering under the large impacting load can be achieved.

The invention relates to a medium and small-sized umbrella-type airship and a control method thereof. The airship comprises a soft airship body, wherein the soft airship body comprises a skin used for keeping a shape and a plurality of airbags arranged in the skin; the skin is of an umbrella-shaped structure; an umbrella cavity is formed below the skin; the airbags are symmetrically arranged in the skin; an umbrella top hole communicated with the umbrella cavity is formed in the top end of the skin; the top of the skin is connected with a guide umbrella through an auxiliary connecting cable; and the lower part of the skin is connected with an effective load cabin through a main connecting cable. When in use, the airship is put into use mainly in a high-altitude inflating expansion way, the airship body is unfilled with air, folded and compressed to have a smaller volume and carried to high altitude for jettison by a high-altitude aircraft or other equipment, the guide umbrella is used for driving the airship to open, and inflating equipment is used for inflating the airship body until certain pressure intensity is reached at the same time. The high-altitude inflating expansion way is particularly suitable for jettison of more airships in a certain domain range, and can effectively avoid side wind influence in a jettison process to enable the airship to accurately reach an appointed region.

The invention relates to the technical field of imaging and spectrum recognition, in particular to a monolithic integrated image sensing chip and spectrum recognition equipment. A light modulation layer of the chip is integrated on an image sensing layer; the light modulation layer comprises at least one sub-modulation layer arranged in the thickness direction of the light modulation layer; each set of pixel confirmation module comprises modulation units and induction units; at least one group of modulation units are distributed on the light modulation layer along the surface of the light modulation layer; at least one group of induction units are distributed on the image sensing layer; each group of modulation units and at least one group of induction units are correspondingly arranged along a vertical direction; a signal processing circuit layer is connected to the lower surface of the image sensing layer and is electrically connected with each sensing unit. According to the chip andthe equipment, the monolithic integration of a light splitting part and a wafer is realized; hyperspectral imaging can be realized; the failure rate of a device is favorably reduced; the yield of finished products of the device is improved; and the chip and the equipment are stable in performance and are not easily influenced by an external environment.