43results about How to "Addressing the drop in accuracy" patented technology

The invention discloses a method for positioning object step by step based on vision fusion, comprising collecting a first image containing target feature points to realize the coarse positioning of the target feature points, according to the coordinate information contained in the first image, collecting the second image containing the target feature point information, carrying out the precise positioning, obtaining the coordinate information contained in the second image, and obtaining the target feature point based on the coordinate transformation amount under the same coordinate system according to the conversion relationship between different coordinate systems, and locating the target feature point. At the same time, the invention also discloses an application, device and system forpositioning object step by step based on vision fusion. The scheme separates the positioning accuracy from the condition of large field of view, two different types of cameras are used to realize step-by-step positioning, which solves the problem that the positioning accuracy is insufficient when the workpiece moves in a large range; and when there are many screw holes, there is no need to take photographs repeatedly, which can reduce the number of photographs, improve the production efficiency, and has the advantages of high precision and high positioning efficiency.

The invention discloses a traffic space-time sequence multi-step prediction method and system, and a storage medium, and the method comprises the steps: 1, calculating an improved historical mean value based on historical data X1:T; 2, carrying out the downsampling of the historical data X1:T and the improved historical mean value through a 3D-CNN module; 3, performing multi-step prediction from asequence to a sequence model to obtain an encoder and a decoder of a model, wherein the encoder and the decoder respectively consist of ConvLSTM, fusing an attention mechanism into the decoder, and further performing up-sampling by using transposed 3D convolution; 4, performing multi-model fusion to obtain a final prediction result. The prediction method provided by the invention is higher in precision.

The invention discloses a system and method for compensation of track precision of an industrial robot based on kinematics analysis. A robot interconnection interface and a measurement interconnectioninterface are in communication connection with a robot controller and a measurement lower-layer interface respectively; the robot interconnection interface and the measurement interconnection interface are connected to an input end of a track bias computation module; output ends of the track bias computation module are connected to a parameter identification module and a precision evaluation module respectively; an output end of the parameter identification module is connected to the robot interconnection interface via a compensation control module; and a track control module is connected tothe robot interconnection interface. According to the method, a motion track bias model of the industrial robot is established; then, model parameter identification is carried out according to robot track data measured by measurement equipment; and finally, error data identified is used in a robot kinematic model for parameter compensation. The system and method disclosed by the invention have thebeneficial effects that a motion track bias of the industrial robot is compensated from the perspective of kinematics analysis; the motion track precision of the robot is increased; cost is low; andimplementation is easy.

The invention provides a pressure sensor with a small measurement error. An elastic membrane (3) is fixedly arranged at a combined part of an air cover and a shell; an iron core (4) is arranged at the center of the elastic membrane; an inductance coil (5) is wound on a framework of the shell; a circuit board (7) is arranged below the shell; the circuit board is provided with a memory device (71); the shell is fixedly provided with a frequency signal output terminal (8) and a memory signal input / output terminal (9); one end (81) of the frequency signal output terminal and one end (91) of the memory signal input / output terminal outward output signals; a middle part (82) of the frequency signal output terminal and a middle part (92) of the memory signal input / output terminal are fixed with the shell; the other end (83) of the frequency signal output terminal and the other end (93) of the memory signal input / output terminal are connected with the circuit board; and the memory device stores a frequency value obtained by outputting through the frequency signal output terminal under different pressures. The pressure sensor disclosed by the invention is applicable to being used as a water level pressure sensor.

The embodiment of the invention provides a remote sensing image water body extraction method and device, electronic equipment and a storage medium, and relates to the technical field of remote sensingimages. The method comprises the following steps: acquiring a to-be-measured remote sensing image, and preprocessing the to-be-measured remote sensing image; based on the water body index, the brightness index and the shadow index, carrying out preliminary extraction on the preprocessed water body area of the remote sensing image to be measured so as to acquire a pattern spot image; identifying the pattern spot image by using a pre-established water body distinguishing model so as to distinguish a water body pattern spot from a non-water body pattern spot; and obtaining a water body extraction result according to the distinguishing result, and removing interference by adopting a brightness index and a shadow index, thereby improving the precision of the water body extraction result, and solving the problem of precision reduction caused by interference in the existing method.

The invention provides a vehicle networking data processing method, a server and a terminal. The method comprises the following steps of acquiring vehicle-running historical data by the server; training the vehicle-running historical data and acquiring a program script used for analyzing a user driving behavior; and sending the program script to the terminal. In the technical scheme of the invention, through the server, the vehicle-running historical data is firstly acquired, then, the vehicle-running historical data is trained and the program script used for analyzing the user driving behavior is acquired, and finally, the program script is sent to the terminal. According to the program script and the vehicle-running data, the user driving behavior is dynamically analyzed. A problem thatalgorithm precision is reduced because algorithm updating is not timely is solved. Accuracy of user driving behavior analysis is guaranteed, and further, extra cost brought by terminal firmware updating is reduced.

The invention relates to a two-degree-of-freedom self-decoupling large-rigidity large-stroke piezoelectric rapid servo tool rest which comprises a tool, a base and a piezoelectric ceramic driver. An installation hinge set composed of a transverse deformation flexible hinge and a longitudinal deformation flexible hinge is arranged between the other end, opposite to the cutter, of the cutter adjusting and controlling vertical rod and the base, and the two piezoelectric ceramic drivers are located on the two sides of the cutter correspondingly. The piezoelectric ceramic drivers comprise a first piezoelectric ceramic driver for outputting longitudinal driving force and a second piezoelectric ceramic driver for outputting transverse driving force, and the two piezoelectric ceramic drivers are respectively provided with a two-stage lever amplification mechanism with a decoupling function. By the adoption of the scheme, the two-degree-of-freedom self-decoupling large-rigidity large-stroke piezoelectric rapid servo tool rest is high in precision, the stroke is enlarged, and the machining stability is improved.

The invention discloses a ball screw prestretching mechanism and a prestretching method thereof. The ball screw prestretching mechanism comprises a ball screw pair, a front locking module and a back locking module, wherein the front end and the back end of the ball screw pair are respectively fixed by the front locking module and the back locking module; the back locking module includes a back end bearing, a back end bearing sleeve and a back end bearing base coating the ball screw pair in sequence from inside to outside; the back end bearing sleeve is coated with a stretching unit; the stretching nut is near the outer side of the back end bearing base; and a rotation preventing pin is inserted in the back end bearing base, and is butted against the outer surface of the back end bearing sleeve after penetrating through the back end bearing base. The ball screw prestretching mechanism can obtain the beneficial effect of preventing the blockage phenomenon of a ball screw by a bearing inner ring due to a bearing locking nut stretching the ball screw when locking a ball screw bearing.

The invention provides a forestry pest identification method and system based on a feature selection module. The method comprises the steps of obtaining a forestry pest image; according to the obtained forestry pest image and a preset forestry pest recognition network model, obtaining a forestry pest classification result, and performing forestry pest recognition; wherein the forestry pest recognition network model is obtained by training a ResNet network embedded with a feature selection module through an expanded forestry pest image training set; obtaining the feature selection module by combining a feature selection unit and residual mapping. According to the invention, a classic ResNet feature extraction network is combined with a designed feature selection module, so that the network can perform feature extraction on purpose, select pest feature information in input data and enhance the pest feature information, the problems of high complexity of a pest picture background and high similarity of part of pests can be effectively solved, and the identification precision is effectively improved.

The invention discloses a fast target positioning method based on ground object coordinate information, which determines the attitude deviation of the satellite according to the position information of the reference point and the orbit information of the satellite; and determines the attitude information of the satellite based on the attitude deviation Perform deviation compensation; determine the position information of the target point according to the compensated attitude information. The invention also discloses a device for quickly locating a target based on ground object coordinate information, a satellite carrying device and a storage medium.

The present invention provides a data processing method for the Internet of Vehicles, a server and a terminal, comprising: the server acquires driving history data; trains the driving history data to obtain a program script for analyzing user driving behavior; sends the program script to the terminal . In the technical solution of the present invention, the server first obtains the driving history data, then trains the driving history data to obtain a program script for analyzing the user's driving behavior, and finally sends the program script to the terminal. By dynamically analyzing the user's driving behavior according to the program script and driving data, the problem of algorithm accuracy decline due to untimely algorithm update is solved, the accuracy of the analysis of user driving behavior is ensured, and further, the update terminal firmware is reduced. additional costs.

A F-P pressure sensor with a composite dielectric film disclosed by the present invention comprises an upper ferrule with a F-P resonant cavity, a lower ferrule which is axially fixedly connected with the upper ferrule and inserted into the lower ferrule. A ferrule for the fiber, and a pressure diaphragm covering the port of the F‑P resonator. At least one layer of composite dielectric film is grown on the back side of the pressure diaphragm and the silicon wafer at the bottom of the groove of the F-P resonator, and the composite dielectric film is SiO 2 / Ta 2 o 5 Thin film, Si 3 N 4 / Ta 2 o 5 thin film or SiO 2 / Ta 2 o 5 / Si 3 N 4 A multilayer film in at least one combination form. The invention has the advantages of simple structure, convenient and quick manufacture, low cost and high thermal stability. In the present invention, a composite dielectric film is grown on the back of the pressure diaphragm and the bottom surface of the upper ferrule groove, and the F-P resonant cavity is formed by connecting the pressure diaphragm with the upper ferrule to match the reflectivity of the system. This combined composite dielectric film is not only thermally stable At the same time, it can be combined according to the design requirements to obtain the reflectivity that meets the system requirements.

The invention provides a signal waveform converting circuit. The signal waveform converting circuit comprises an electronic switch unit and a differential motion output unit, wherein the electronic switch unit converts an input signal into two in-phase output signals, and the signal waveform converting circuit outputs a difference value between two in-phase output signals and converts the input signal into a direct current signal; as the input signal is not required to be converted into a steamed bread wave and then converted into the direct current signal, the problem of precision decrease caused by introducing of a diode when the input signal is converted into the steamed bread wave is solved, and the problem of long delay caused by introducing of a capacitor when the steamed bread wave is converted into a direct wave signal is solved.

The invention relates to the field of human face key point detection, and discloses a method for selecting a human face frame in a human face key point detection module, including obtaining a human face detection model and a key point model; Face frame coordinates and the real face frame coordinates of N pictures to obtain the face frame coordinate scaling ratio; use the face frame coordinate scaling ratio to the face frame coordinates of each picture in the test data set output by the trained face detection model Perform optimization to obtain the optimized face frame coordinates of each picture in the test data set; perform coordinate transformation on the optimized face frame coordinates to obtain new face frame coordinates, and cut out from the original image according to the new face frame coordinates Face frame, input the cropped face frame into the trained key point model to obtain key point coordinates. The present invention can effectively solve the problem that the accuracy of key points decreases due to inconsistent selection standards of human face frames in human face detection and key point detection modules.

The invention discloses an industrial robot track precision compensation system and method based on kinematics analysis. The robot interconnection interface and the measurement interconnection interface are respectively connected to the robot controller and the measurement bottom layer interface. Both the robot interconnection interface and the measurement interconnection interface are connected to the input terminal of the trajectory deviation calculation module, and the output terminals of the trajectory deviation calculation module are respectively connected to the parameter identification module and The accuracy evaluation module, the output terminal of the parameter identification module is connected to the robot interconnection interface through the compensation control module, and the trajectory control module is connected to the robot interconnection interface; the method establishes the motion trajectory deviation model of the industrial robot, and then uses the robot trajectory data measured by the measuring equipment to carry out Model parameter identification, and finally the identified error data is used for parameter compensation of the robot kinematics model. The invention compensates the motion track deviation of the industrial robot from the perspective of kinematics analysis, improves the motion track accuracy of the robot, has low cost and is easy to implement.

Embodiments of the present application provide a water body extraction method, device, electronic device, and storage medium for remote sensing images, which relate to the technical field of remote sensing images. The method includes acquiring a remote sensing image to be tested, and preprocessing the remote sensing image to be tested; and preliminarily extracting the water body region of the preprocessed remote sensing image to be tested based on a water body index, a brightness index and a shadow index, to obtain Spot image; use the pre-established water body discrimination model to identify the spot image to distinguish the water body spot and the non-water body spot; according to the discrimination result, obtain the water body extraction result, and use the brightness index and the shadow index to remove the interference, so as to The accuracy of the water body extraction result is improved, and the problem that the accuracy of the existing method is reduced due to interference is solved.