24 results about "Centroid algorithm" patented technology

The invention discloses a treadmill user center track identification method based on Kalman filtering, and belongs to the technical field of ultrasonic measurement. An ultrasonic sensor array installed above a running belt is used for collecting user position data, an improved weighted centroid algorithm is adopted for calculating a position observation value, and the algorithm is combined with an exponential attenuation item and a power weight to improve the positioning precision. A second-order motion state model containing position and speed is established, adaptive estimation of observation noise covariance is realized based on statistical characteristics of an innovation sequence, and process noise covariance is dynamically adjusted according to residual sequence changes. A Kalman filtering framework is adopted for state prediction and updating, and real-time estimation of the position and the speed of the user is achieved. The method accurately recognizes the transverse position track of the user in the running process, adapts to different noise environments and motion modes, and has the advantages of being high in precision, high in adaptability and good in reliability.

The invention relates to a memory data line training method, a memory controller, a processor and electronic equipment, and belongs to the field of electronic circuits. The method comprises the steps that a two-dimensional eye pattern corresponding to a target data line of the memory is acquired, the two-dimensional eye pattern comprises a plurality of test points located in a two-dimensional space formed by a voltage dimension and a time delay dimension, and each test point corresponds to a combination of a voltage value and a transmission time delay value; determining the center of mass of the two-dimensional eye diagram; and based on the coordinates of the centroid in the voltage dimension and the time delay dimension, determining a reference voltage configuration value and a transmission time delay configuration value corresponding to the target data line. By introducing the centroid algorithm in mathematics, automatic selection of the optimal equilibrium point of the eye pattern is realized, the centroid reflects the equilibrium center of the effective area of the eye pattern in geometric and physical distribution, and the centroid is taken as the sampling center point, so that natural equilibrium between the time domain and the voltage domain can be realized; therefore, the influence of external interference on the eye pattern shape is reduced to the minimum.

The application relates to a method for processing muon detection data of a triangular prism scintillator. S1: obtaining uncorrected energy deposition information and uncorrected hit position of each muon event in each layer detection unit; S2: based on the actual structure parameters of the detector, performing geometric correction on the traditional centroid algorithm; S3: based on the statistical relationship between the energy deposition of the muon in the scintillator and the track length, establishing a signal channel gain coefficient model, and correcting the energy response of each detection channel; S4: based on the fitting results of the muon track length and the energy of each channel, determining the energy consistency standard, and performing rejection processing on the muon events determined as abnormal; S5: performing muon track reconstruction and imaging, evaluating the position resolution of the detector, and if the convergence condition is reached, ending, otherwise returning to S2 for iteration. The technical scheme can improve the muon track reconstruction accuracy and imaging quality through geometric structure correction, energy correction and abnormal energy event screening.

The application relates to the field of space laser communication, in particular to a face array APD multi-spot overlapping state simultaneous identification method and system based on code division multiple access, which comprises the following steps: three slave optical terminals respectively emit three beacon light beams to a master optical terminal; the three beacon light beams pass through a converging lens and converge to a 5*5 face array APD detector photosensitive surface to generate 25 APD pixel photoelectric current signals; the 25 APD pixel photoelectric current signals are input to a signal conditioning circuit, and a digital signal corresponding to each APD pixel photoelectric current signal is output; the digital signals are all input to a preprocessing unit to obtain a summary signal; the preprocessing unit inputs the summary signal to a code tracking loop to obtain an output signal; integrated processing is conducted on the output signal output by the code tracking loop and the digital signal output by the signal conditioning circuit to obtain an integrated processing result; and a centroid algorithm is used to solve the spot center position of the integrated processing result. The number of code tracking loops, the complexity and the scale of the circuit system are reduced.

The invention relates to a digital simulation platform, method and device of a wavefront sensor and a storage medium. The system comprises a Zernike polynomial calculation module, a micro-lens array simulation module, a centroid extraction module and a wavefront reconstruction module. The Zernike polynomial calculation module is used for analyzing and generating incident wavefront by adopting a Zernike polynomial, the micro-lens array simulation module is used for simulating a micro-lens array by adopting a mixed diffraction propagation model to segment the incident wavefront to generate a light spot image, and the centroid extraction module is used for extracting the light spot image by adopting a weighted centroid algorithm. The light spot image acquisition module is used for acquiring a light spot image and calculating a mass center position of the light spot image in combination with global and local dynamic threshold mechanisms, and the wavefront reconstruction module is used for establishing a relationship between the mass center position and a wavefront slope, solving a Zernike coefficient by using a numerical optimization algorithm, generating wavefront aberration distribution and outputting a reconstructed wavefront phase. Through the modular design, the independence of each function is ensured, the simulation efficiency is improved, and the data generation capability and the system flexibility are enhanced.

The invention discloses an accurate positioning system and method for mine personnel and mobile terminal equipment, the system can automatically correct a reference error caused by roadway deformation or equipment displacement by introducing a dynamic baseline real-time measurement technology, a multi-source data fusion mechanism is adopted, the arrival time difference and a received signal strength value are synchronously utilized, and the positioning accuracy of the mine personnel and the mobile terminal equipment is improved. And resolving is carried out through a weighted centroid algorithm, so that the robustness and the positioning stability of the system are remarkably improved. An inertial measurement unit is innovatively integrated, a signal intensity value is intelligently compensated by sensing a walking direction of a person and combining a human body shielding model, a measurement error caused by human body shielding is effectively reduced, a multipath interference index is extracted by analyzing a channel pulse response, an algorithm weight is dynamically adjusted accordingly, and the accuracy of human body shielding is improved. According to the method, environment self-adaptive positioning optimization is realized, the technical bottlenecks of insufficient precision and poor environment adaptability of the traditional positioning technology are effectively solved, and reliable technical guarantee is provided for safe production of mines.

The invention discloses a remote fault on-line positioning processing method, system and platform based on a floating tap sliding window and a storage medium, and aims to realize instrument-free and on-line automatic positioning of a remote fault of a high-speed link by configuring a floating tap in a decision feedback equalizer into a slidable detection window. The detection limitation that a standard fixed tap is only about 5cm is broken through, the effective diagnosis range is extended to 30cm through a parameter self-adaptive window sliding strategy, and a back plate and a key area at the tail end of a cable are completely covered. The spatial resolution is improved to a submillimeter level by combining energy detection with a centroid algorithm; through an original distance energy double-criterion mechanism, false peak values generated by multiple reflections can be effectively identified and inhibited, and only real physical faults are reported. Existing hardware resources of the chip are utilized in the whole process, external testing equipment or service interruption is not needed, and an efficient and low-cost automatic solution is provided for mass production testing, on-site operation and maintenance and monitoring of the high-speed link.

The application discloses a kind of element pin detection methods based on machine vision, comprising the following steps: step one: using industrial camera obtains the 3D point cloud of the detected element including pin;Step two: based on the 3D point cloud obtained in step one, obtain height map;Step three: in height map, select a height reference value, and form a selection interval on the height reference value according to the set height difference value;Step four: the point cloud of the sampling point in the selection interval determined in step three is used to form a gray scale image;Step five: in the gray scale image, the center of gravity of the pin is obtained using Blob analysis algorithm and center of gravity algorithm;Step six: use the center of gravity of the pin obtained in step five to judge defects, so use height map to set selection interval to select effective point cloud, so noise point cloud can be filtered, avoid noise point cloud adversely affect gray scale image, to form the real gray scale image of pin, and then judge the defects of pin.

The invention relates to a livestock and poultry positioning and tracking method and device based on the Internet of Things, equipment and a medium. The method comprises the following steps: acquiring historical position and relative distance data through an Internet of Things terminal deployed on livestock and poultry, and constructing a group topological relation model; when a terminal failure is detected, M normal terminals with the highest correlation degree are dynamically screened based on the model to form a collaborative node set; fusing the real-time position data of the cooperative nodes and the association strength parameters in the model, and calculating the position coordinates of the livestock and poultry where the invalid terminal is located in real time through a weighted centroid algorithm; and finally, performing state prediction and observation updating on the continuous position estimation points based on a Kalman filtering algorithm, and generating a smooth and continuous motion track during a terminal failure period. According to the scheme, a dynamic cooperation mechanism of group intelligence is utilized, space correlation characteristics among individuals are converted into a positioning compensation basis, complete trajectory tracking can still be maintained when equipment fails or signals are interrupted, and the robustness and data continuity of a livestock and poultry positioning system in a complex environment are improved.

This invention discloses a machine vision-based method and system for monitoring the seal of an internal floating roof, belonging to the field of machine vision technology. It includes: acquiring laser stripe images collected by a monitoring unit; performing binarization and connected component analysis on the laser stripe images to delineate the region of interest, and extracting the center line of the laser stripes with sub-pixel precision using a grayscale centroid algorithm; establishing a triangulation model based on a pre-calibrated camera intrinsic parameter matrix and laser plane equation, converting the center line into a three-dimensional contour point cloud of the seal, and performing time-series filtering; retrieving the reference contour point cloud corresponding to the circumferential encoder angle, calculating the normal deviation value and tangential gap value; determining the fault type based on multi-level comparison results with preset thresholds, and generating a fault spatial index by fusing circumferential azimuth angle and liquid level data. This invention obtains high-contrast stripes under extremely dark lighting and metal reflection environments in storage tanks, enabling quantitative measurement of the three-dimensional deformation of the seal and preventative maintenance.

The application provides a strabismus nine-orientation image classification method and system, which comprises the following steps: based on a preset recognition algorithm and a pre-constructed coordinate system, boundary recognition is performed on a pre-acquired eye image to obtain initial left eye boundary box coordinates and initial right eye boundary box coordinates, a deviation angle is calculated and the eye image is corrected according to the deviation angle to obtain target left eye boundary box coordinates and target right eye boundary box coordinates; based on a preset centroid algorithm and the corrected eye image, target left eye iris centroid coordinates and target right eye iris centroid coordinates are calculated; a preset prediction model is used to obtain a nine-orientation distribution probability of the eye image; based on preset nine-orientation mutual exclusion requirements and the nine-orientation distribution probability of the eye image, the eye image is classified to obtain a strabismus nine-orientation image classification result. The application realizes accurate positioning of the iris position of the eye image, reduces irrelevant information interference, and classifies the iris by combining the mutual exclusion characteristics among the nine orientations, thereby improving the reliability of the eye image classification.

The invention relates to the technical field of optical measurement and signal processing, and discloses a high-precision center control target scoring system which comprises a laser phase shift interference array, a high-speed phase demodulation unit and a center control processing unit. According to the system, a laser phase shift interference array is used for forming a laser sensing surface behind a target surface, and the laser sensing surface is used for detecting air refractive index change caused in a transient wake of a projectile after the projectile penetrates through the target in a non-contact manner and converting the change into an instantaneous phase shift signal. And the central control processing unit accurately calculates the coordinates of the impact point through a weighted centroid algorithm based on the peak value of the signal. In addition, the system also constructs a phase disturbance characteristic vector by extracting a plurality of characteristic parameters of the phase signal, and performs mode matching with an expert knowledge base to realize synchronous diagnosis of the flight state of the projectile. The method does not depend on a special target plate, is high in anti-interference capability, not only realizes high-precision target scoring, but also provides an additional ballistic analysis function, and has the characteristics of high stability and low operation cost.

The invention relates to the technical field of sensor detection, in particular to a non-contact human body posture detection device based on an ultrasonic array. Comprising a shell, an ultrasonic transmitting unit, a two-dimensional ultrasonic receiving array, a signal conditioning module, a distance detection and multi-mode adaptive control module, a dual-mode automatic calibration module, an STM32 single-chip microcomputer control module, an output interface module and a power supply module. The ultrasonic transmitting unit is used for transmitting a 40kHz-60kHz pulse signal by adopting a dynamic beam forming technology; the two-dimensional receiving array obtains spatial amplitude distribution information; the multi-mode self-adaptive control module dynamically adjusts the gain and the threshold by combining the distance and the signal-to-noise ratio; the dual-mode automatic calibration module corrects channel errors through starting / periodic calibration; the single-chip microcomputer module combines an energy centroid algorithm and an attitude feature library to realize attitude recognition with response less than or equal to 50ms. The device has high detection precision and strong privacy, has power consumption of less than or equal to 2W, and is suitable for the scenes of old-age nursing, game interaction, smart home and the like.

The application discloses a weak pressure measurement method based on a stable buoyancy system and an image centroid algorithm, and belongs to the technical field of precise mechanics measurement. The method first constructs a self-stable buoyancy rotation platform supported by high-viscosity silicone oil and an optical lever amplification light path, and eliminates mechanical friction by means of the difference between the gravity moment and the buoyancy moment; a weak pressure to be measured is applied to make the platform produce a slight rotation angle, the rotation angle is amplified into linear displacement of a reflected light spot through the optical lever principle, and a video is collected; red channel extraction and multi-frame superposition and averaging are carried out on the video frames to suppress liquid surface fluctuation noise; a threshold segmentation and a morphological operation are used to lock a light spot connected domain, a weighted pixel centroid algorithm is used to calculate the light spot center coordinates; and the light spot displacement is converted into the pressure value to be measured according to a calibrated linear response model. The application realizes high-precision measurement of weak pressure in the order of micro-newton, and the measurement resolution is better than 1 muN.

The invention provides a geometric level automatic measurement method and system based on CCD encoding and decoding, and the method comprises the steps: scanning a bar code ruler through a CCD sensor, and obtaining a bar code digital signal; obtaining a current measurement distance to generate a corresponding interpolation step length, and performing sub-pixel linear interpolation on the bar code digital signal to obtain an interpolated bar code digital signal; extracting a wave crest and a wave trough of the interpolated barcode digital signal, and screening wave crest data meeting a matching condition based on a wave crest interval to obtain a wave crest section; based on the wave crest segment, the wave trough and the standard bar code, performing bar code matching through a matching algorithm to obtain an initial bar code sequence; correcting the initial barcode sequence based on a centroid algorithm to obtain a target barcode sequence; and decoding the target barcode sequence based on the coding rule to obtain a geometric level measurement result so as to complete automatic geometric level measurement. According to the method, sub-pixel-level positioning precision is provided by adopting bar code matching, and the accuracy of geometric leveling results is improved.

This invention discloses a dynamic multidimensional rapid track association method and device based on millimeter-wave traffic radar. The method includes: acquiring point cloud data collected by millimeter-wave traffic radar; clustering the point cloud data using the DBSCAN clustering algorithm to obtain point cloud clustering results; using the centroid algorithm to agglomerate the points in the point cloud clustering results to obtain agglomerated points; and using an association algorithm based on the optimal association coefficient to perform association calculations between the agglomerated points and currently managed tracks or historical points to obtain matched track information.

This invention discloses a method, apparatus, device, and storage medium for wireless optical communication beam alignment. The method includes: acquiring a real-time beacon image; calling a target detection model to perform target detection; and outputting the beacon beacon bounding box and corresponding detection score; determining an initial region of interest (ROI) based on the beacon beacon bounding box, determining an expansion factor based on the detection score, adaptively expanding the ROI, and performing boundary enhancement processing on the expanded ROI to obtain an enhanced ROI image; using a centroid algorithm to perform sub-pixel-level localization of the beacon center position in the enhanced ROI image to obtain the sub-pixel centroid coordinates of a single beacon beacon; comparing the sub-pixel centroid coordinates with the center coordinates of the camera imaging plane to obtain a deviation; calculating a control quantity based on the deviation to drive an actuator to adjust the beam direction, thereby achieving closed-loop beam alignment. This invention enables fast, stable, and sub-pixel-accurate beacon light recognition and beam alignment in complex environments.

The application discloses a treadmill user central trajectory recognition method based on Kalman filtering and belongs to the technical field of ultrasonic measurement. User position data is collected by an ultrasonic sensor array installed above a running belt, and an improved weighted centroid algorithm is used to calculate position observation values, the algorithm combines an exponential decay term and a power weight to improve positioning accuracy. A second-order motion state model containing position and velocity is established, adaptive estimation of observation noise covariance is realized based on statistical characteristics of a new sequence, and process noise covariance is dynamically adjusted according to residual sequence changes. State prediction and updating are carried out by using a Kalman filtering framework, and real-time estimation of user position and velocity is realized. The application accurately recognizes the lateral position trajectory of a user during running, is suitable for different noise environments and motion modes, and has the advantages of high precision, strong adaptability and good reliability.

The application discloses a kind of real-time tracking method and system of space discontinuous weld based on laser vision, it is related to robot intelligent welding technical field, including steps: real-time acquisition laser stripe image of the region to be welded;Laser stripe image is carried out denoising processing to obtain denoising stripe image, extract image region of interest in denoising stripe image, extract stripe image center point in image region of interest using gray centroid algorithm;Based on stripe image center point, segmented fitting is carried out, and first fitting straight line segment and second fitting straight line segment are obtained;Image intersection point is mapped from image coordinate system to robot coordinate system based on calibration mapping relationship, and predetermined positioning image welding point is identified as welding skip point, and continuous welding center fitted by weld feature image is identified as continuous welding point.The method provided by the application has the beneficial effects of identifying predetermined positioning welding point and continuous welding point in advance to guide welding arm to carry out welding operation, which is beneficial to the automation of welding operation and improves the quality of welding.

The application discloses a high-dynamic smearing star image exposure center time image plane positioning method, and belongs to the field of telescope optical detection image processing. The method comprises the following steps: firstly, filtering a high-dynamic star map through an anisotropic Gaussian kernel to obtain an enhanced image and a direction image, and then accurately extracting the exposure center position of a star image based on the enhanced image and the direction image; performing linear fitting on each extracted star image by using a least square method to obtain a motion trajectory of the star image; and calculating an integral midpoint along the motion trajectory of the star image as the exposure center position of the star image by using the gray level accumulation sum of the normal direction of each point on the trajectory. The application realizes accurate pixel position determination of a smearing star image under a high-dynamic background, can effectively deal with the problem that the star image center position is difficult to accurately obtain under complex smearing conditions caused by high-dynamic star image smearing and platform turntable jitter, significantly increases the star image exposure center time image plane positioning precision compared with a traditional centroid algorithm, and finally improves the performance of a star tracker.

The application discloses a touch screen positioning and smoothing filtering method, which comprises the following steps: collecting a group of touch picture line data containing multiple frames; for each frame data, adopting a centroid algorithm to calculate the predicted coordinates of each frame touch point; from the second frame data, calculating the distance between the predicted coordinates of the current frame touch point and the predicted coordinates of the last frame touch point as a target distance; judging whether the target distance is between the minimum distance threshold and the maximum distance threshold; if yes, putting the predicted coordinates of the current frame touch point into a queue, and taking the mean value of the predicted coordinates of each frame touch point stored in the queue as the smoothed coordinates of the current frame touch point; if not, taking the predicted coordinates of the last frame touch point or the current frame touch point as the smoothed coordinates. The application has low calculation amount, which can be accepted by a single-chip microcomputer, and has excellent smoothing effect.

The present application relates to a kind of neural network hardware centroid calculation methods for cross bar position-sensitive anode, belong to neural network model training and data processing technical field, the technical problems that centroid algorithm exists are solved, processing speed is influenced, it is difficult to guarantee the calculation accuracy, hardware is difficult to realize, its method includes the measured data set of anode readout signal generation, matlab is made into data set by simulation data, data set randomly generates training set and test set, select U-Net neural network as model, using python carries out model training, exports model parameter;U-Net neural network is combined with the exported model parameter and deployed to FPGA, algorithm deployment uses FPGA chip, obtains the FPGA chip with centroid calculation function.The centroid calculation method is used for the hardware solution of cross bar position-sensitive anode centroid.

The invention provides a fiber grating sensor wavelength demodulation method and system based on an FPGA, belongs to the field of fiber sensing technology and hardware acceleration, can realize high-speed, high-precision and low-power consumption multichannel FBG wavelength demodulation, and meets the reliability requirement of spaceflight embedded deployment. By utilizing the parallel processing capability of the FPGA, simultaneous processing of multichannel spectral data is realized, and the demodulation speed is greatly improved; a pipeline architecture is adopted, so that data acquisition, processing and output can be continuously carried out, and the system throughput is remarkably improved; the innovative adaptive threshold algorithm improves the accuracy and reliability of peak detection; a centroid algorithm realized by hardware ensures high precision and stability of wavelength calculation; the whole scheme has the characteristics of low power consumption and high integration level, and is very suitable for embedded application.

The application provides a visitor management system for an Internet of Things community, comprising: a visitor terminal for visitor identity registration and access application; a proprietor terminal for the proprietor to receive access requests and perform remote authorization; and an access control execution unit arranged at a community entrance for collecting biological feature information of the visitor and verifying access rights; the application introduces a dynamic calculation formula of time deviation and credit score, realizes fine control of the access rights of the visitor, avoids security risks caused by static authorization, establishes a dynamic generation mechanism of a visitor 'black list' and 'white list', encourages the visitor to behave in a standard manner, uses homomorphic encryption technology to ensure the safety of biological feature data in the transmission and processing process, and uses an RSSI weighted centroid algorithm to realize effective monitoring of the trajectory of the visitor without increasing expensive positioning hardware.