71 results about "Linear least squares method" patented technology

A gesture recognition interface for use in controlling self-service machines and other devices is disclosed. A gesture is defined as motions and kinematic poses generated by humans, animals, or machines. Specific body features are tracked, and static and motion gestures are interpreted. Motion gestures are defined as a family of parametrically delimited oscillatory motions, modeled as a linear-in-parameters dynamic system with added geometric constraints to allow for real-time recognition using a small amount of memory and processing time. A linear least squares method is preferably used to determine the parameters which represent each gesture. Feature position measure is used in conjunction with a bank of predictor bins seeded with the gesture parameters, and the system determines which bin best fits the observed motion. Recognizing static pose gestures is preferably performed by localizing the body / object from the rest of the image, describing that object, and identifying that description. The disclosure details methods for gesture recognition, as well as the overall architecture for using gesture recognition to control of devices, including self-service machines.

A gesture recognition interface for use in controlling self-service machines and other devices is disclosed. A gesture is defined as motions and kinematic poses generated by humans, animals, or machines. Specific body features are tracked, and static and motion gestures are interpreted. Motion gestures are defined as a family of parametrically delimited oscillatory motions, modeled as a linear-in-parameters dynamic system with added geometric constraints to allow for real-time recognition using a small amount of memory and processing time. A linear least squares method is preferably used to determine the parameters which represent each gesture. Feature position measure is used in conjunction with a bank of predictor bins seeded with the gesture parameters, and the system determines which bin best fits the observed motion. Recognizing static pose gestures is preferably performed by localizing the body / object from the rest of the image, describing that object, and identifying that description. The disclosure details methods for gesture recognition, as well as the overall architecture for using gesture recognition to control of devices, including self-service machines.

An estimation apparatus of an object appearance frequency is provided. The apparatus estimates an appearance frequency of objects, such as pedestrians, in a predetermined estimation area. The apparatus calculates a matrix FPFP by searching an appearance frequency data in the past. In detail, an estimated result that is an output of an estimating module is expressed as a vector. Objects to be estimated are classified into a total of 12 kinds, such as a “man, woman, child, bike, unknown, dog” and a “right, left”. Feature vector of appearance frequency of pedestrians, i.e., objective variables of estimation, is expressed by 12th dimension vector space. Moreover, status information is used as explaining variables, which explain the feature vectors when a feature vector occurs. The regression relationship of the feature vector with respect to the status information vector expressed in 28th dimension is solved by the linear least squares method.

The invention provides a cable connector fault diagnosing method. The cable connector fault diagnosing method comprises the steps that the temperature of the outer surface of a cable connector and the temperature of a monitoring point (located at the zero potential position of the outer surface of an aluminum protective sleeve or the outer surface of a prefabricated rubber insulating layer) inside the cable connector are monitored in real time; the linear least squares method is used for identifying relevant parameters of cable connector transient state thermal circuit models through the temperatures obtained through real-time monitoring in order to obtain the time constant tau of the first order transient state thermal circuit model and the steady state temperature rise coefficient k of the first order transient state thermal circuit model; and fault diagnose is conducted on the cable connector by the utilization of the identifying result. By means of the cable connector fault diagnosing method, the four common faults of the cable connector can be effectively diagnosed, the defect that in many quality evaluating theories of simplified models, cable heat capacity and the transient state process of cable connector temperature rise are not considered is overcome, the demand for on-line fault monitoring of the cable connector is met, and the stability of system running is improved.

The method for estimating position of mobile station includes the following steps: utilizing the correspondent measurement of TOA of main base station and TDOA of other base stations corresponding to main base station to obtain related statistical information of TOA correspondent to main base station and TDOA correspondent to other base station; according to base station position information and related TOA and TDOA statistical informations constructing operation matrix, then using weighted linear least square method at twice to estimate position of mobile station, finally utilizing minium distance difference to select final position estimation valve. Said invention also provides its estimation equipment, including TOA and TDOA generator, information data base and position estimator.

The invention discloses a rapid testing method for optical fiber gyroscope scale factor parameters. The testing method comprises the following steps: fixedly mounting an optical fiber gyroscope on a rotating table and rotating along with the rotating table in an identical sine angle velocity, acquiring to obtain output digital amount of the optical fiber gyroscope and the rotation sine angle velocity of the rotating table at different time points, performing linear least square method fitting, calculating to obtain the optical fiber gyroscope scale factor parameters, testing a plurality of sine angle velocity periods, averaging output digital amounts of the optical fiber gyroscope under the identical sine angle velocity in each sine angle velocity period so as to obtain an average output digital amount of the optical fiber gyroscope under the identical sine angle velocity, and further performing linear least square method fitting to calculate the optical fiber gyroscope scale factor parameters. The testing method is short in testing time, the optical fiber gyroscope scale factor parameters can be rapidly evaluated, the testing equipment is lowly required, no special rotating table adaptive interface line is needed, the testing system is simple and reliable and high in security coefficient, and rapid testing on the optical fiber gyroscope scale factor parameters is achieved.

The invention discloses a steel member cylinder fitting algorithm which comprises the following steps: setting a plurality of measuring points on a to-be-measured section circle, arranging a high-precision laser instrument to acquire three-dimensional coordinates of the measuring points, and converting the three-dimensional coordinates measured at different measuring positions into a common coordinate system by adopting a transfer algorithm; calculating a fitting plane equation by utilizing the coordinate points after station transfer to obtain plane equation parameters, and taking the plane as an evaluation plane for the roundness of the section; projecting the coordinates of the measurement points to the evaluation plane, and calculating the projection coordinate of each point and the distance from each point to the evaluation plane; carrying out plane circle fitting on the coordinate points projected to the same plane by utilizing a linear least square method to obtain a fitting radius and a circle center of a circle; on the basis of plane circle fitting, calculating the cylinder roundness of a plurality of profile circles by adopting a least square method, namely carrying out cylinder fitting. A plurality of algorithms are integrated, so that the method has very strong pertinence on the dimension measurement of the large-scale cylindrical steel member.

The invention relates to a fusing method of temperature compensation of a humidity sensor. A characteristic function y=f(x) of the humidity sensor is given, wherein y represents the output of the humidity sensor, x represents the input of the humidity sensor, a temperature parameter is provided additionally, when t0<t1<t2<t3, according to the characteristics that a characteristic curve of the humidity sensor is linear in the middle and is nonlinear at two ends under the influence of the temperature, through a linear least square method, a linear equation f2(x, t) influenced by the temperature within a range of [t1, t2] is obtained; and when t0<t<t1 or t2<t<t3, error compensation caused by the temperature influence is carried out on f1(x, t) and f3 (x, t) through an RBF (Radial Basis Function) neutral network. The fusing method has the beneficial effect that an actual value which is closer to an output value of the humidity sensor is finally obtained.

The invention discloses a method for calibrating an accelerometer under an unsupported state, which is applied to an inertial navigation system with the capacity of initially aligning with a static base attitude. The method comprises the following steps of: firstly, simplifying an error model, repeatedly changing the carrier position, and coarsely and finely aligning with each position by using the static base to acquire the attitude information; and calculating the three-axis component of an acceleration of gravity g projected under a carrier coordinate system according to the attitude information, and recording the measured value of the accelerometer at each position. The invention provides a 10-position calibrating scheme, which comprises the following steps of: acquiring zero offset to degenerate the error model into a linear model by identifying the zero offset through a nonlinear least square method according to the calculated value and measured value of the projected component; and mounting an error and scale factor coupled matrix by identifying through a linear least square method. Therefore, the key parameters of the accelerometer are calibrated. The method has the advantages of capacities of realizing unsupported calibration of the accelerometer and effectively estimating the key parameters of the accelerometer with time-varying performance, along with simple process and high speed.

The invention discloses an omni-directional side lobe suppression method for synthetic aperture radar imaging processing. The method comprises the following steps of: arranging two-dimensional echo signals obtained by synthetic aperture radar into one-dimensional data; determining an imaging plane region position; constructing into a measuring matrix corresponding to an imaging plane; and finally, processing the one-dimensional echo data by using a linear least square method in combination with a measuring matrix, thereby realizing omni-directional side lobe suppression of a synthetic aperture radar imaging result and obviously reducing the energy leakage. Thus, the energy leakage of the imaging result of the processed satellite-borne echo data is the minimum, and further, the aim of suppressing side lobe is fulfilled. The omni-directional side lobe suppression method has no special requirement on target scattering intensity, can be used for imaging processing of any scene target and has a wide application range.

A chromatographic analyzer is provided for facilitating curve fitting by means of the linear least-square method for a chromatogram that contains a plurality of overlapping peaks. A chromatographic data processor executes fitting processing to each peak in an arbitrary time region having the plurality of peaks of the chromatogram starting from the front side or back of the time region, and the processed peaks are subtracted from the chromatogram in the time region so that the plurality of peaks in the chromatogram can be separated from one another.

The invention discloses a device and method for improving azimuth measurement precision in a coherent wind finding laser radar. The method comprises the following steps: at first, calibrating an electronic compass; then under the control of a control circuit board, driving the electronic compass to rotate by a rotary platform, recording the angle of the rotary platform and read azimuth of the electronic compass; preprocessing the azimuth of the electronic compass, fitting the azimuth by a linear least square method, calculating the intercept and slope of the linear functional relationship; and finally according to the angle of the rotary platform, calculating the azimuth. Because the rotary platform adopts a worm and gear structure; a control circuit board and a motor driver are used to obtain an angle, wherein the precision of the angle is higher than electronic compass; and thus the provided device and method can improve the azimuth measuring precision of a coherent wind finding laser radar. Moreover, the device does not need extra equipment or human; only needs the devices in a coherent wind finding laser radar, the precision improvement is automatically performed under the control of an algorithm; the method and device are convenient, fast, and intelligent, and the precision is high.

A chromatographic analyzer is provided for facilitating curve filling by the linear least-square method for a chromatogram that contains a plurality of overlapping peaks. A chromatographic data processor executes fitting processing to each peak in an arbitrary time region having the plurality of peaks of the chromatogram starting from the front or back side of time region and the processed peaks are subtracted from the chromatogram in the time region so that the plurality of peaks in the chromatogram can be separated from one another.

The invention provides a CO2 concentration layering inversion method and system of a multi-wavelength laser radar. The method comprises the following steps: S100, layering a CO2 concentration inversion integration route in a vertical direction according to the influence of weather and carbon circulation on CO2 concentration; S200, respectively establishing a corresponding CO2 concentration multi-wavelength inversion model and a weight model for each layer, and obtaining a CO2 column concentration calculation model of each layer; S300, performing linear equation aggregation according to the CO2column concentration calculation model of each layer, obtaining an observation equation adopting the CO2 column concentration of each layer as a to-be-calculated parameter, calculating the CO2 columnconcentration of each layer by adopting a linear least square method, and calculating the CO2 concentration of each layer according to the CO2 column concentration. According to the CO2 concentrationlayering inversion method and system of the multi-wavelength laser radar, the actual distribution state of CO2 in atmosphere is considered, and a CO2 concentration inversion result with higher accuracy can be obtained.

The invention discloses a method of structure fracture and ground settlement deformation decomposition based on force source inversion. Firstly, time-sequence interference processing is carried out onan SAR image of a study area to obtain a corresponding annual average deformation rate or a deformation time sequence; then a deformation area comprehensively influenced by ground settlement and structure fracture is selected for relevant parameter setting of an inversion model; then gridding is carried out on the inversion model, a compression amount of each grid is calculated through a Green'sfunction and a linear least square method with an additional constraint condition, and the Green's function is combined to obtain ground settlement caused by groundwater exploitation by forward modeling; and finally, ground settlement deformation obtained by restoration is removed from ground surface deformation obtained by InSAR technology, and ground surface deformation caused by structure fracture is obtained. The method has guiding significance for disaster formation modes, disaster prevention and disaster reducing of geological disasters in research areas where structure fracture deformation and ground settlement deformation comprehensively appear and generate influences together.

The invention discloses a method for compensation fitting a differential signal used for extracting a peak value of dispersion confocal. The method comprises the following steps: fitting a differential signal of a normalized discrete spectral response signal by a linear least square method to obtain an initial peak wavelength; further obtaining an error between the initial peak wavelength of the normalized discrete spectral response signal and an ideal peak wavelength, and performing first derivative deformation on the error; solving a differential equation of the first derivative deformationto obtain an error between the initial peak wavelength of the fitted and normalized discrete spectral response signal and the ideal peak wavelength X; and further obtaining a fitted value of the idealpeak wavelength of the normalized discrete spectral response signal; thereby performing rapid measurement of dispersion confocal by utilizing a fitted value of the ideal peak wavelength of the normalized discrete spectral response signal to ensure the accuracy of rapid measurement of dispersion confocal.

The invention discloses a time synchronization method based on a linear weighted least square method. The time synchronization method comprises the following steps: S1, arbitrarily selecting a node as a root node during initialization, and broadcasting a synchronous data packet; S2, if a non-root node does not receive the synchronous data packet within the appointed time, ensuring that the node is self-defined as the root node; S3, enabling the non-root node to construct a fitting straight line according to the linear weighted least square method to obtain a clock drift a and a clock drift b; and S4, adjusting the local clock according to the clock drift a and the clock drift b obtained in step S3, and ensuring that the non-root node similarly broadcasts the synchronous data packets after synchronization with the root node. The existing linear least square method is based on the assumption that the observed noises are consistent, according to the time synchronization method disclosed by the invention, a larger weight value is given to a small observation quantity of the observed noises to ensure that the clock drift can closer to the oscillator frequency change of a real sensor node.

The invention discloses an external parameter calibration method and system for a single-line laser radar and a visible light camera, and the method comprises the steps: collecting time and space synchronous furnace charge image data and point cloud data, correcting a furnace charge image according to a camera distortion model, and further extracting a laser scanning center line; determining a two-dimensional coordinate corresponding to a local extreme value of a laser scanning center line in an image coordinate system, and performing interpolation on the sparse point cloud data to obtain a three-dimensional coordinate of the local extreme value of the point cloud data in a radar coordinate system; according to the feature point pairs of the two-dimensional coordinate corresponding to the local extreme value of the laser scanning center line in the image coordinate system and the three-dimensional coordinate of the local extreme value of the point cloud data in the radar coordinate system, utilizing a linear least square method to obtain an external parameter matrix of the camera and the single-line laser radar; the technical problem that the existing laser radar and camera external parameter calibration precision is low is solved, the single-line laser radar and the visible light camera can be calibrated online in real time, and the external parameter matrix is corrected.

The invention provides a dynamic measurement method and instrument of laser spots. The dynamic measurement method comprises the following steps of: (1) respectively detecting received light intensities by all pixels of a CCD (Charge Coupled Device); (2) marking light intensity data detected by all the pixels as a matrix I, and obtaining Sigma x, Sigma y, xc, yc, A and I0 by utilizing a linear least square method on the basis of a laser spot light intensity distribution formula according to the actually-measured light intensity data matrix I and corresponding coordinate values of horizontal and vertical axis of all the corresponding pixels, wherein xc and yc respectively refer to coordinate values of the horizontal and vertical axis of a light spot center, Sigma x and the Sigma y respectively refers to 1 / e radiuses of the horizontal and vertical axis, A refers to amplitude and I0 refers to dark noise; (4) obtaining a modified vector D of a reference vector P(k) by using a Newton iteration method, and calculating P(k+1)=P(k)+D; and (5) repeatedly executing the step (4) till the iteration frequency reaches a preset value K or the error of the reference vector P(k) which is obtained by iteration for kth times is smaller than a preset tolerance value, and taking all elements of the reference vector P(k) at the time as a measured laser beam parameter. The invention further provides a corresponding dynamic measurement instrument of laser spots. The dynamic measurement method and instrument of the laser spots have the advantages of high measuring speed, high precision and large dynamic range.

The invention discloses a related entropy-based Torr-M-Estimators basic matrix robust estimation method. A Torr-M-Estimators algorithm performs weighting processing on a residual error of each matching pair to reduce the influence of an exterior point of the relatively large residual error on an estimation process of a basic matrix, and different weight functions have different optimization results; an information theory-based related entropy function has relatively great advantage in robustness and is more remarkably improved in precision and robustness; data normalization is performed by utilizing a Hartley method, an initial value of the basic matrix is obtained by a linear least square method, a related entropy is a Torr-M-Estimators weight function, a rank of the basic matrix is limited to be 2 by eigenvalue decomposition, and a loop stop condition is determined by a minimum iterative error and a maximum iterative frequency; and an experiment shows that the method has very high estimation precision and robustness and is an important improved method in the basic matrix solving step in the fields of three-dimensional reconstruction, motion estimation, matching tracking and the like.

The invention discloses a sine distortion image correction method used for a confocal endoscope, and relates to the technical field of medical instruments. The method comprises the steps that step 1:imaging of a ronchi grating is performed by using the confocal endoscope, and the Gaussian half-height width of each stripe is extracted by using a Gaussian fitting method; step 2: fitting of the Gaussian stripe half-height width is performed by using a linear least square method so as to acquire a fitting curve representing the degree of image distortion and acquire the correction coefficient ofcurve fitting; and step 3: the distortion image is corrected according to the correction coefficient and then the corrected image is acquired. The technical effects that the structural information ofsubcellular level resolution can be provided in an in-vivo way in real time, the method can be used for small animal experiments to shorten the drug research and development period and reduce researchand development input and thus the method has great application prospect and value can be achieved.

A chromatographic analyzer is provided for facilitating curve fitting by means of the linear least-square method for a chromatogram that contains a plurality of overlapping peaks. A chromatographic data processor executes fitting processing to each peak in an arbitrary time region having the plurality of peaks of the chromatogram starting from the front or back side of the time region, and the processed peaks are subtracted from the chromatogram in the time region so that the plurality of peaks in the chromatogram can be separated from one another.

The invention discloses a fault diagnosis method for catalyzing and cracking a main fractionator based on a double-layer method. A double-layer fault detection and diagnosis scheme for a rectificationprocess of the fractionator is provided, the rectification process is monitored by using a non-linear model which is developed at the early stage, a corresponding simplified linear model is established, and an abnormal source a measurement value of which has deviation is identified. According to a fault diagnosis algorithm, based on the linear model, a linear least square method is adopted for estimating fault parameters in rectification. A double-layer diagnosis structure can effectively capture the change of the fault parameters, the deviation is sensitively detected, the specific reason offaults can be judged from the change trend of the parameters, and the structure is more effective under comparison with a structure based on a pure non-linear model.

The invention discloses a detonator delay elimination method for coal mine shock wave velocity inversion and is suitable for the technical field of coal mining and coal mine safety. The method is characterized by, to begin with, marking and analyzing P-wave first arrival time of a shock wave recorded by a probe; calculating distance between each shot point and the probe; carrying out relation fitting through a linear least square method to obtain relation between the P-wave first arrival time and the distance and calculating detonator delay; and finally, eliminating the detonator delay to obtain corrected P-wave arrival time. The method can eliminate the detonator delay, so that actual propagation time of the P wave in a coal-rock mass can be calculated accurately, and accuracy and reliability of coal mine shock wave velocity inversion are improved; and the method is reliable in result, wide in application range and easy for computer programming.

The invention relates to a QPSO (quantum-behaved particle swarm optimization) algorithm based image edge detection method. The method includes: forming an image edge detector by four ANFIS (adaptive neuro-fuzzy inference systems) sub-detectors and a postprocessing module; before using the method to perform edge detection on an image, constructing a training image artificially, independently training the four sub-detectors by means of QPSO and a LSE (linear least square method), and determining parameters in a system; when the four sub-detectors are all train, forming the image edge detector by the sub-detectors with the postprocessing module, and detecting the edge of the test image. The image edge detection method has the advantages that even if the test image is polluted by noise, edge information in the image can be extracted effectively without image filter preprocessing.

The invention discloses a method applied to simplification of a random-configuration bistatic synthetic aperture radar echo frequency spectrum. The method comprises the following steps: initializing parameters, discretizing a GBSAR frequency spectrum phase, carrying out approximation of frequency spectrum phase space function, weighing and summating sum of frequency function of the frequency spectrum to obtain a simplification expression. The method applied to simplification of the random-configuration bistatic synthetic aperture radar echo frequency spectrum has the advantages that the linear fitting of the space composite function and frequency composite function in the GBSAR frequency spectrum phase is carried out by a discrete linear least squares method, the GBSAR frequency spectrum is simplified from the polynomial sum to the binomial sum, efficient focusing and imaging are implemented according to the simplification equation by an efficient frequency domain image processing technology, and the GBSAR frequency spectrum is simplified with high efficiency and high accuracy.

The invention discloses an enhanced camera-assisted positioning method based on a received signal strength ratio, and belongs to the technical field of wireless communication. The method comprises thefollowing steps: firstly, constructing a scene comprising K transmitters and a to-be-positioned receiver, and positioning by using four coordinate systems; obtaining the optical power received by thePD from each LED by using the channel direct-current gain between each LED and the PD, and simplifying the optical power; then respectively calculating an incident angle of each LED under a 3D cameracoordinate system; calculating the distance ratio between the receiver and each LED by using the optical power and the incident angle corresponding to each LED, and calculating the distance between the receiver and each LED by using the included angle between the two LEDs and the receiver according to the cosine law; finally, for the deployment of only three LEDs, performing 2D position estimation and 3D position estimation according to a linear least square method to obtain the position coordinates of the receiver. The method has the advantages of high coverage rate, low complexity and highpositioning precision.

The invention belongs to the technical field of electronic information, relates to a convex optimization method for determining target three-dimensional coordinates by utilizing one-dimensional directfinding, and aims at solving the problem that the nonlinear least square method needs three-dimensional search and the pseudo-linear least square method needs no less than 6 observation stations in the background art of one-dimensional direction finding-based three-dimensional positioning. Through converting the one-dimensional direction finding-based three-dimensional positioning problems into convex problems, the nonlinear constraint relationship between variables is utilized, so that the positioning precision is improved, the three-dimensional coordinate search is avoided, and the positioning timeliness is improved; and under the condition that the observation stations are less than 6, the three-dimensional positioning for targets is realized by utilizing one-dimensional direction finding results and orientation measurement of linear arrays.

The invention provides an in-orbit calibration method when pointing angle error of a satellite-borne laser altimeter is nonconstant. The method comprises the following steps: determining expression forms of laser pointing angle system error of the satellite-borne laser altimeter in the pitch and roll directions, and constructing a vector to be estimated; constructing a laser foot point observationequation when the angle system error is nonconstant; estimating an initial value by utilizing a linear least square method; carrying out iterative convergence estimation by utilizing a nonlinear least squares LM algorithm; calculating error components of the laser pointing angle in the pitch and roll directions at the current moment; and calculating a laser foot point precise coordinate after thepointing angle system error is compensated and calibrated in the current measurement time, and finishing in-orbit calibration of the pointing angle error of the satellite-borne laser altimeter. Compared with the existing pointing angle error calibration method, the method can carry out calibration when the pointing angle error is nonconstant, which cannot be finished through the existing method;and the method can be compatible with the existing situation that the pointing angle system error is constant, and is higher in universality.

The invention belongs to the technical field of the electronic information, and relates to a method for quickly determining target three-dimensional coordinate by utilizing one-dimensional direction-finding. The invention aims at a problem that the three-dimensional search calculated amount of the nonlinear least square method in the background technology of performing the three-dimensional positioning by utilizing the one-dimensional direction-finding and the problem that the pseudo-linear least square method information utilization is insufficient; by using an one-dimensional result of an observation station and own azimuth angle measurement of a linear array of the observation state, the three-dimensional positioning on the target is realized through an iterative searching way, the three-dimensional coordinate searching is avoided, and the three-dimensional coordinate of the target is quickly determined.