80 results about "Least squares curve fitting" patented technology

The invention relates to a fish eye lens calibration and fish eye image distortion correction method which belongs to the field of digital image processing and is very suitable for occasions of vision navigation, mobile monitoring and so on. Since imaging using a large wide angle will cause serious distortion, generally, imaging correction is needed to facilitate human eye observation. The method provided by the invention is mainly technically characterized by establishing a space coordinate system 1 and employing a least square curve-fitting method to calibrate a fish eye image to determine the center andthe radius of a fish eye lens, establishing a cylindrical projection model in a space coordinate system UV, and setting one-to-one relationships among an image point q2 of the fish eye image, a point q1 on object sphere and a point q projected on a cylinder, wherein q1 is in vertical projection relation with a bottom plane on which q2 is located and line q1q parallels to line oq2, thus achieving the purpose of fish eye image correction.

The invention discloses a method and device for obtaining a wheel set tread contour line of an urban rail train. The device comprises two sets of 2D laser displacement sensors installed on the two sides of a rail in a mirror symmetry mode. The laser detection faces of the two sets of sensors are located on the same plane, and the integral device is lower than the rail surface. The method comprises the steps that after the sensors detect wheels at the same time to obtain detection point coordinates, the output points of the two sets of sensors are combined to a same coordinate system through coordinate transformation and coordinate translation; the x-coordinate value of the right end face of a tread is obtained, and a filtering window is established based on the x-coordinate value to remove interference points of measuring data; the x-coordinate value of the right end face of the tread is obtained, and measured data points are divided into K number sets according to the actual situations; least squares curve fitting is carried out on the data points of each number set; coordinate values of segment points are obtained, and each segment curve is smoothened into an integral tread contour line according to the lagrange multiplier method. The method and device have the advantages of being high in speed and precision, easy to operate, capable of carrying out online non-contact type measurement, and the like.

The invention is applicable to the field of image processing and provides an estimation method and system for a lung motion model. The method includes: extracting a control point in a lung three-dimensional CT image of a reference phase; using a tracking position at a previous moment of the control point as a prediction position and tracking the position of the control point in a continuous-discrete-phase three-dimensional CT image; performing least-square curve fitting on the continuous-discrete-phase pursuit path of the control point and determining the motion position of the control point at any continuous time; and constructing a three-dimensional deformation model of the lung at a current time according to the motion position of the control point at any continuous time and obtaining lung motion position estimation of any voxel of the lung of the reference phase at any time. The estimation method and system for the lung motion model adopts an image registration algorithm based on a mark point so that the method is small in calculation quantity, high in calculation speed and capable of estimating the motion model of the lung effectively.

The invention discloses a method and device for measuring X-ray thickness and belongs to the field of ray measurement. The method for measuring the X-ray thickness comprises the steps of (1) arranging an X-ray emitter and adjusting emitting parameters; (2) starting up the X-ray emitter and measuring energy data I0 of rays in the air where an object is not placed; (3) starting up the X-ray emitter, enabling X-rays to penetrate through a standard to-be-measured material with known thickness d and measuring the ray energy data I after the rays penetrates through the to-be-measured material; and (4) calculating correction parameters by means of the following methods for obtaining corresponding attenuation coefficients mu according to a formula, performing least-squares curve fitting calibration to obtain the correction parameters A, alpha and beta according to a formula mu(d)=A(e-alphad+beta) and performing thickness measurement of the to-be-measured material according to a formula. By means of the method, the problem of mismeasurement caused by 'beam hardening' of the X-rays is solved. The method has a correction function. The method and device for measuring the X-ray thickness is simple and easy to operate, high in accuracy and strong in reliability.

The invention discloses a special spectrum simulation illumination device for plant cultivation and a simulation method. In the invention, the design idea of single light source is broken through; a hexagonal extensible light source array is designed based on LED combination; control power supply is realized by signal chip computer control and inductance switch stabilized current; the configuration parameter of a light source is determined by a monochrome calibration method; and the simulation of a special spectrum is realized by a nonlinear least squares curve fitting. The device has the advantages of special spectrum simulation, light-reliable-efficient performances, convenient extension and strong commonality. The simulation test result shows that the invention can pointedly and flexibly simulate the visible light absorption spectrum of special plants, solves the technical problem that how to supply light efficiently and flexibly for biological culture and provides a new technical solution for biological scientific study, biological culture and fresh-keeping in special environment and the like.

The invention discloses an automatic compensation method for static drift of a sight line of a stabilized photoelectric sight system, and belongs to the field of automatic control. The method comprises the steps that an automatic tracking control loop is disconnected under a small view field of a sight tool of the stabilized photoelectric sight system; the tracking displacement deviation is counted in a certain period; the drift velocity of the sight line is calculated by using a least square curve fitting algorithm; the drift velocity is converted into corresponding control voltage; a gyro stabilization loop is reversely added to compensate drift motion; and a servo control system is controlled by a timer to automatically and repeatedly execute the operation until the drift velocity is less than a specified threshold. The method is achieved by completely adopting a software algorithm, solves the problems of slow manual drift compensation and low compensation accuracy, is simple in algorithm and high in portability, and is suitable for drift compensation of the sight line of the existing on-board stabilized photoelectric sight system.

The invention particularly discloses a visual identification-based precise shaft size measurement method, apparatus and system. The method comprises the steps of establishing a mapping relationship between a pixel size and an actual spatial geometric size of a to-be-measured precise shaft; obtaining multiple to-be-spliced precise shaft images; building a conversion model between a reference imageand the to-be-spliced images to perform conversion; based on CS and NSST algorithm combination, performing splicing on the converted to-be-spliced images, and performing fusion to form a whole image of the to-be-measured precise shaft; and for the fused whole image, performing pixel level edge tracking and preliminary locating by using a multilevel filtered interested edge detection method firstly, then based on Sobel operator and least square curve fitting combination, obtaining sub-pixel-precision edges, and according to the detected edges and the mapping relationship, calculating the spatial geometric size of the to-be-measured precise shaft. According to the method, the apparatus and the system, the image splicing data processing amount is effectively reduced; the detection efficiencyis improved; and real-time online detection can be performed.

The invention belongs to the technical field of embedded air data systems for aerospace vehicles, provides a monitoring and processing method for pressure faults of an embedded air data system, and aims to solve the problem of system final fault caused in case of pressure hole blockage, pipeline leakage, pressure sensor faults, acquisition and processing circuit faults, signal transmission faults and the like occurred in an FADS, or the problem of parameter operation divergence which finally causes impacts on flight control. A correlation matrix among pressure spots is determined by using a wind tunnel test and flight test data according to the distribution of pressure holes, and the correlation matrix is pre-assembled in the FADS; fault location is performed through a finally obtained pressure value, the pressure value of a fault point is estimated by using a least squares curve fitting method, and the pressured value obtained through evaluation is used for participating follow-up operation. The embedded air data system pressure fault monitoring and processing method provided by the invention can effectively process pressure hole blockage, pipeline leakage, pressure sensor faults, processing circuit faults, signal transmission faults, pressure value fault and the like, can provide reasonable atmospheric parameters to a mobile system, and can effectively improve the practicability of an embedded air data system.

The invention discloses a human-computer interaction method based on eye movement control. The method comprises the following steps of human eye pupil positioning and eye movement characteristic extraction. The step of pupil center positioning based on grey information comprises the following stages that: 1: human eye positioning; 2: pupil edge detection; 3: pupil center positioning. The step of eye movement characteristic information extraction based on visual point movement comprises the following steps that: calculating a fixation point deviation and visual point deviation mapping; on the basis of the displacement difference of two obtained adjacent frames of images, enabling eyes to move back and forth to calibrate points calibrated by coordinates on the screen, and utilizing a least square curve fitting algorithm to solve a mapping function; and after eye movement characteristic information is obtained, on the basis of the obtained eye movement control displacement and angle information, carrying out a corresponding system message response. By use of the human-computer interaction method based on the eye movement control, the energy consumption of the user can be reduced, theuser can be assisted in carrying out direct control, and efficient and natural interaction is realized.

This invention relates to an automatic extraction method for image character dot with high noise, which comprises the following steps: firstly, preprocessing the collected digital image, then dividing the image automatically by method basing on image dividing in transition area or in background area to get the object character area, and then distilling the object character brink dot, finally working out the analysis format of the image character dot by iterative least square curved-fitting method.

The invention discloses a dynamic harmonic wave electric energy metering method based on multi-resolution short-time Fourier transformation. The method mainly comprises the steps of 1, conducting discrete sampling on a voltage signal and a current signal to obtain sampling sequence U (n) and I (n) of N points; 2, using a polynomial convolution window to weight the sampling sequence U (n) and I (n) and to conduct short short-time Fourier transformation to obtain a time frequency matrix FSTFT (g, j) of the signals; 3, using a least squares curve fitting method to solve all extreme points Km (g) in an amplitude matrix |FSTFT (g, j)|; 4, according to the extreme points Km (g), calculating the frequency, amplitude and phase of a fundamental wave and a harmonic wave and a duration set Tsm; 5, according to a power calculation formula, solving power of the fundamental wave and all harmonic waves, and then obtaining an electric energy metering result through a short-time Fourier transformation dynamic electric energy calculation formula. According to the dynamic harmonic wave electric energy metering method, based on the multi-resolution short-time Fourier transformation, a frequency domain characteristic and a time domain characteristic of the fundamental wave and the harmonic wave are analyzed, dynamic harmonic wave electric energy metering can be accurately conducted, and the requirement for modern electric power system dynamic electric energy metering is met.

The specification discloses a system and related method for obtaining electromagnetic resistivity logs of a formation having fixed depths of investigation that differ from the actual depths of investigation for the tool. The system and related method comprises calculating actual invaded zone and formation resistivities, and then using least-squares curve fitting techniques to find a function that models the relationship between the physical spacing of the transmitters and the depths of investigation. Using the function, it is possible to find a transmitter-to-receiver spacing that, though not actually used, would give the desired depth of investigation. This transmitter-to-receiver spacing is then used in combination with a second function, determined using least-squares curve fitting methods, which models the relationship between the transmitter-to-receiver physical spacing and the measured resistivity to obtain the resistivity reading at a fixed depth of investigation. This may be done for a set of desired depths of investigation, and may also be done for a set of elevations within the borehole, to create a fixed depth of investigation log that may be easily compared to other resistivity logs, such as those created by a wireline device.

The invention relates to a frequency calibration method, device and system for an oven controlled crystal oscillator. The method comprises the steps that preset test conditions are read and written in the crystal oscillator; output frequencies of the crystal oscillator under the different test conditions are read; least squares curve fitting is performed on relation arrays of the test conditions and the output frequencies to obtain polynomial functions about the test conditions and the output frequencies; the test condition corresponding to the calibration frequency of the crystal oscillator is calculated through the polynomial functions and is written in the crystal oscillator; the output calibration frequency of the crystal oscillator under the test condition corresponding to the calibration frequency is read and verified; if the output calibration frequency meets the range of frequency accuracy, calibration is over; or else, the calibration continues to be performed after coefficients of the polynomial functions are adjusted. According to the frequency calibration method, device and system, the frequency calibration of the oven controlled crystal oscillator can be automatically completed, measuring errors are reduced, and the purposes of performing frequency calibration automatically and accurately and enhancing production efficiency are achieved.

A time domain sensor system operable via a set of executable instructions storable in relation to a memory device for transforming output data by performing a nonlinear least-squares curve fitting, involving: a time domain sensor device having a mass-spring oscillator including a frame portion and a cantilever portion having a proximal end and a distal end, the cantilever portion coupled with the frame portion; and a plurality of proximity switches having a movable portion and a fixed portion in relation to the frame portion, each proximity switch of the plurality of proximity switches having at least one proximity tip, and each proximity switch configured to trigger in response to an acceleration experienced by the cantilever portion; and a processor operatively coupled with the time domain sensor device and configured to operate via the set of executable instructions for transforming the output data by performing a nonlinear least-squares curve fitting.

The invention relates to an automatic extraction method for rough vein work-piece characteristic points against a background of strong reflection, the characteristics of which are as follows: first, preprocessing the collected digital picture; second, segmenting picture through automatic picture segmentation method based on gray-scale gradient to acquisit target character area; then, extracting target character margin points; finally, calculating cleavage form of rough vein work-piece characteristic points through iterative least square curved-fitting method.

The invention discloses a wave height measuring method based on a monocular vision laser triangulation method. The wave height measuring method comprises the steps of collecting a water surface photocontaining laser light when there is no waves, preprocessing the image to obtain an edge and a contour of the laser light, extracting straight line features of the laser light, locating the image point coordinates of the intersection of laser light and water surface, obtaining a wave height corresponding to the image point coordinates according to the principle of direct laser triangulation methodwhen there are waves on the water surface by regarding the water surface without waves as a reference plane, repeatedly measuring to obtain the data about a plurality of sets of image point coordinates and wave heights, obtaining a relationship curve of the image point coordinates and wave heights by a least squares curve fitting method, obtaining the image point coordinates of the intersection of the laser light and the water surface by image processing during the measurement, and directly obtaining the wave height at this time by the relationship curve of the calibrated image point coordinates and the wave heights. The wave height measuring method provided by the invention belongs to non-contact measurement, and has fast measurement speed, high measurement precision and high efficiencywithout being easily affected by environmental factors such as water temperature.

The invention belongs to the technical field of weapon damage assessment, and particularly relates to a method for analyzing the distribution law of explosion loads on the surface of a steel member. The component surface explosion load distribution rule analysis method comprises the following specific steps: firstly, calculating the positive reflection peak pressure and the positive impulse of thecenter of an explosion facing surface; establishing a fitting formula of the positive reflection peak pressure and the positive impulse by utilizing a least square curve fitting method; determining acomponent surface distribution function; determining explosion load parameters of any point on the surface of the component; and combining the explosion load parameters of the central point of the component, namely explosion load reflection pressure P and impulse I, and determining the explosion load parameters of any point by utilizing the normalized distribution function sum, namely, after theexplosion load parameters of the surface of the component are determined by utilizing the method, performing evaluation analysis on the damage of the component by utilizing the obtained explosion loadparameters. The method has the advantages that the effect in the near-explosion explosion environment can be quickly and reliably calculated, and the like.

The invention discloses a non-flat road surface lane line detection method. The non-flat road surface lane line detection method comprises steps of S1, establishing a normal-lane-line-based model and extracting line characteristics of the lane line, S2, adopting an improved Hough transform lane-line straight line detection method to detect the lane line by combining with direction characteristics and intensity information of line characteristics, and S3, using a detected straight line to accurately position the lane line and adopting a weighted least square curve fitting method to completely extract the lane line of the image. The non-flat road surface lane line detection method uses the improved Hough conversion to perform straight line detection on the basis of the extracted line characteristics, adopts the weighted least square method to perform curve fitting, can accurately and robustly detect the lane line on a curve and a non-flat road. The non-flat road surface lane line detection method is simple in calculation and thus the calculation process has high practicability.

The invention relates to a method for collecting and calibrating multiple cameras of an LED (Light-Emitting Diode) display screen. The method comprises the following steps of: respectively collecting the tricolor light intensity values of each pixel point on the display screen at N collection cameras by utilizing collection equipment; respectively fitting tricolor light intensity value distribution quadratic polynomials according to test data sequences of the tricolor light intensity values of the pixel point by utilizing a least square curve fitting method; solving out the maximum value of each quadratic polynomial, namely the tricolor normal light intensity data of the pixel point; obtaining a correction coefficient matrix of any pixel point by utilizing the object brightness of the display screen and the normal light intensity data of each pixel point; loading the correction coefficient matrix of each pixel point by virtue of a display screen control system, and performing differentiation control on the pixel points of an LED, namely achieving an effect that the brightness of the full screen is consistent. By utilizing the method, seen from different viewing angles, the uniformity of the full screen is consistent, and the visual angle of the display screen is enlarged.

The invention discloses a space target track interruption correlation method based on trace forecast, and belongs to the technical field of track correlation. A space target has the characteristics ofhigh density, low target relative movement speed, poor divisibility and the like, the problems that in a tracking process, targets are easily lost, and the same target is repeatedly subjected to track initiation and is given with a plurality of batch numbers are easily caused, in order to effectively improve continuity and stability of tracking on the space target and improve target tracking precision and effective tracking batch numbers, the speed and position of the space target are separately subjected to first-order least square curve fitting and second-order least square curve fitting toimprove the precision of a trace forecast initial value point, the position of the trace forecast initial value point is obtained by variable-order curve fitting, the target acceleration of the timeis estimated according to a spatial objective dynamic equation, the movement locus of the space target is forecasted through the estimated acceleration by a Euler extrapolation forecast method, and the problem of track interruption correlation is solved by the forecast result of the trace of the space target.

The invention discloses a real-time method for eliminating influences of temperature on low-frequency sampling indexes in bridge online safety monitoring. The method is clear in principle and easy toimplement, can efficiently eliminate the influences of the temperature effect from actually measured data of a bridge monitoring system, obtains responses purely caused by live loads, and meets the real-time pre-warning and state evaluation requirements of the bridge online safety monitoring system. The method comprises the main steps that firstly, according to the actually measured data of n low-frequency sample indexes before a current moment, least square curve fitting is conducted on the actually measured data to obtain a fitting formula of the variation trend of the actually measured data; secondly, after the fitting formula is obtained, the temperature effect of the current moment is predicted by using the formula; thirdly, the predicted temperature effect is eliminated from the actually measured data of the current moment. The method is easy to operate, efficient, accurate and suitable for real-time temperature effect modification of the low-frequency sampling indexes (such as displacement and the inclination angle) of all bridge online safety monitoring systems.

Provided is a power compensation optimizing method used in electric energy meter replacement in a live working state based on least square method curve fitting. The method comprises the following steps of: connecting the voltage wire of a three-phase electric energy meter field calibrator with a loop in parallel and connecting the current wire of the three-phase electric energy meter field calibrator with the loop in series before an electric energy meter is replaced; when the electric energy meter is replaced, reliably connecting the secondary current test terminals of the replaced electric energy meter in a short circuit manner and measuring a current value by using a clamp ammeter in order to determine whether reliable short-circuit connection is achieved, starting a stopwatch to record replacing time; acquiring active power discrete points p(tn) within a time segment from t1 to t2 by using the three-phase electric energy meter field calibrator when replacement begins, acquiring a power curve function p(t) by using a least square method curve fitting method according to discrete point sampling values, and computing total active power lost in a meter replacing process by using a relation between power and time. The method improves the accuracy of metering results and guarantees the economic interests of a power seller and a power buyer.

The present invention relates to a curve fitting method based on the least squares method. The method establishes a linear model on the basis of a nonlinear fitting curve, uses an iterative calculation method to provide coefficients in the function of the fitting curve, and substitutes the coefficients into The theoretical value obtained after fitting the function has the characteristics of the minimum sum of squares of errors, and the method can accurately provide the coefficients of the fitting function by means of precise iterative calculation, so that the sum of squares of errors of fitting is minimized; the present invention avoids The minimization problem existing in the least square method in the linear transformation process is solved, and the curve fitting method provided by the invention has high fitting precision, strong logic, easy programming, simple method and easy realization.

The invention discloses a VSG inertia and damping coefficient measuring method based on nonlinear least square curve fitting, the method analyzes that the system is on the off-line, The time domain unit step response functions of different operating conditions and damping states are analyzed qualitatively. The method of solving inertia and damping coefficients by using dynamic response indexes isqualitatively analyzed. A load rejection test for inertia measurement of synchronous generators is used for reference, and a self-verifiable comprehensive measurement method is obtained by combining power step test. By synthesizing the two experimental methods, the influence of inaccurate synchronous power value on parameter measurement is avoided, which provides a new method for external characteristic measurement of VSG.

The invention discloses a high-performance soft decision decoding method based on a QC-LDPC code, and mainly solves the problems of poor decoding performance and difficulty in implementation of an existing algorithm. According to the implementation scheme, the method includes: carrying out MATLAB simulation on the relation between the signal-to-noise ratio and the multiplicative factor according to the characteristic that the value of the multiplicative factor is related to the signal-to-noise ratio, and obtaining the corresponding relation between the signal-to-noise ratio and the multiplicative factor; performing least square curve fitting on the corresponding relationship to obtain a relational expression alpha (s) of the multiplicative factor and the signal-to-noise ratio; in a decoding iteration process, updating a check node message by utilizing a multiplicative factor alpha (s) and a minimum sum decoding algorithm; updating a variable node message; judging the decoding message in combination with the check node updating information and the variable node updating information; and after the maximum number of iterations is reached, stopping decoding, and outputting a decoding result. According to the method, the decoding performance of the QC-LDPC code is improved, the method is convenient to implement, and the method can be used for a communication system scene with a highrequirement for information transmission reliability.

The invention discloses a zero-second resistance value computing method. The zero-second resistance value computing method comprises solutions of zero-second resistance values of a stator winding and a rotor winding, and is characterized by comprising the following steps of: (1) programming a program by using MATLAB software, and (2) executing the program: drawing a least square fitting figure of the stator winding, drawing a least square fitting figure of the rotor winding, computing a zero-second resistance value and a fitting error of the stator winding, and computing the zero-second resistance value and the fitting error of the rotor winding. According to the invention, the zero-second resistance can be quickly solved by using a least square curve fitting method, a programmed simple program can be applied for many times, and therefore the zero-second resistance value computing method is an effective computing method characterized in convenience and high accuracy and without influence of human factors.

The invention discloses a high voltage equipment state predicting method. The method comprises the following steps: (1) actually detecting historical data of gas components of at least one kind of transformer oil dissolved gas and / or SF6 gas in high voltage equipment, namely, detecting data of gas components in various periods on the condition that the equipment is an electrification mode; (2) utilizing the historical data in the step (1) and adopting a least squares curve fitting method to legally obtain the component data of corresponding gas in the high voltage equipment at certain time in future, namely, obtaining a characteristic curve changing along with time, and eventually obtaining long-time discharging or overheating stability data of the equipment, wherein the numerical value of the characteristic curve possibly continuously tends to be stable; (3) through a vector supporting machine, utilizing the component data, of the corresponding gas in the high voltage equipment at certain time in future, obtained in the step (2) to predict the failure state type of the high voltage equipment. The high voltage equipment state predicting method has the advantages of being low in dependence on experiences, strong in anti-interference ability and generalization ability, high in accuracy and capable of solving a multi-classification problem.

The invention relates to a design method of an approximate flat-top main beam for a microwave wireless energy transmission system. The method comprises the following steps: (1) calculating a basic field strength pattern at a far field of a transmitting antenna according to an aperture field distribution function; (2) constructing the flat-top main beam field intensity pattern function Eflat-top(theta); (3) defining the description function of the flat-topped main beam and determining the objective function of the least square curve fitting; (4) using the least square method, and obtaining thefield intensity pattern function Eflat-top(theta); (5) calculating beam collection efficiency and edge coning on the receiving antenna; (6) using the following optimization model to optimize the objective function to obtain the optimal beam collection efficiency and the optimal coning value of the receiving antenna edge; (7) obtaining a design variable X; (8) according to the design variable X andthe Pareto front-end diagram, obtaining the optimal beam collection efficiency and the coning value of the receiving antenna edge as well as the flat-top main beam field intensity pattern function Eflat-top(theta).

The invention discloses a road network data automatic quality inspection method based on a high-resolution remote sensing image, and the method automatically achieves the quantitative quality inspection of a road track, and assists the updating and optimization of road data. The method comprises the following steps: 1) according to road morphological characteristics, breaking at road characteristic points to generate a simple road section structure; 2) constructing a universal road cross section model according to the road section structure characteristics; and 3) acquiring image data block data corresponding to the road section structure. And 4) establishing an image block coordinate system, and carrying out template matching between the road cross section model and the image to obtain aroad matching point. And 5) after RANSAC least square curve fitting error calculation, obtaining an actual road extraction result in the image. 6) matching the actual road extraction result with the road section structure similarity, and 7) performing comprehensive evaluation in combination with the similarity matching result of the road section structure in each road to screen out a'problem road'for road data quality evaluation and further data modification and updating.