33results about How to "Avoid less accurate" patented technology

The invention discloses a positioning and navigation system and method of a robot and relates to the technical field of robots, aiming at solving the problem that the orientation judgment accuracy isrelatively low when a UWB (Ultra Wideband) technology is used for positioning and navigating the robot. The system comprises a plurality of labels, a base station, an electronic compass, a gyroscope and a controller. The method comprises the following steps: calculating a position coordinate of each label in real time and judging whether position coordinate data of each label is reliable or not; if so, determining a relation between a mounting position of each label and the orientation of the robot and combining the position coordinate of each label to determine an orientation angle of the robot; if not, judging whether detection data of the electronic compass is reliable or not; if the detection data is reliable, calculating according to the detection data of the electronic compass and determining the orientation angle of the robot; initializing the detection data of the gyroscope to be 0; when the detection data is not reliable, determining the orientation angle of the robot according to the detection data of the gyroscope. The positioning and navigation system and method of the robot can be applied to a positioning and navigation process of the robot.

The invention discloses a homologous determination method and device for malicious files. The method and device are used for solving the problem of the lower accuracy of the homologous determination result in the prior art. The method comprises the steps that based on the characteristic behavior data information of each malicious file in the first class specified dimension, the category of each malicious file is obtained, and whether the malicious file in the corresponding category is the homologous malicious file or not is determined based on the characteristic behavior data information of each malicious file of each category in the second class specified dimension. Through the fusion analysis of the characteristic behavior data information in the multiple dimensions, the homologous malicious file is accurately judged and the accuracy of the homologous determination result is effectively improved. Moreover, after the determination of the category of the various malicious files, the homologous determination of the category of the various malicious files is conducted, so that the calculated amount of the homologous determination is decreased and the accuracy of the homologous determination result is further improved.

An embodiment of the present application provides a depth-of-field imaging module, comprising an imaging lens, a TOF sensor, an RGB sensor, a laser emitter, and an optical path conversion structure. The laser emitter is used to emit an infrared laser to that camera lens. The camera lens is used for transmitting infrared laser light to the target object. The reflected light beam reflected by the target object is received and transmitted to the optical path conversion structure. Reflected light includes infrared laser and ambient light reflected by the object. The optical path conversion structure is used for dividing at least part of the reflected light ray into a first target light ray and a second target light ray. A first target light is emitted to the TOF sensor and a second target light is emitted to the RGB sensor. The TOF sensor is used for obtaining depth information of a target object from an infrared laser in a first target light. The RGB sensor is configured to obtain RGB information of the target object from ambient light of the second target light. The present application can avoid the problem of low accuracy of the depth-of-field imaging module due to the use of a dualimaging structure.

The embodiment of the invention provides a battery management unit test method and system; the test method comprises the following steps: responding to a test command, obtaining a target test case ofa battery management system BMS, obtaining a test script corresponding to the target test case, wherein the test script contains simulation signal information; forming a simulation signal according tothe simulation signal information; outputting the simulation signal to the BMS, obtaining state information of the BMS after the simulation signal is inputted; obtaining the BMS test result accordingto the state information. The test method and system can solve the problems that an existing method enables test personnel to manually simulate the simulation signal, thus causing manpower wastes andlow testing efficiency.

The invention provides an image classification method and device, computer equipment and a storage medium, and belongs to the technical field of image processing. The method comprises the following steps: inputting a target image into an image classification model, classifying the target image based on the image classification model, and outputting a plurality of first confidence coefficients; determining second confidence coefficients of at least two object types corresponding to the at least two image clusters according to the first confidence coefficients of the at least two image clustersmeeting the target condition; and determining at least one object type corresponding to the target image according to the second confidence coefficients of the at least two object types. The technicalscheme is adopted, the problem that the accuracy is low due to the fact that the confidence that the target image belongs to the object type is directly output is avoided, the second confidence of the at least two object types is determined according to the first confidence of the at least two image clusters meeting the target condition, and then the at least one object type is determined, so that the accuracy of the at least one object type is high.

The invention relates to the technical field of image processing, and discloses a text paragraph structure restoration method, device and equipment and a computer storage medium, and the method comprises the steps: carrying out the recognition of a target image, and determining all textboxes in the target image and the textbox positions of all textboxes based on a recognition result of recognition; sorting the textboxes according to the textbox positions, and inputting text features of the textboxes into a preset deep learning model for training based on a sorting result of sorting; and combining the textboxes based on the training result of the training to obtain all text paragraphs corresponding to the target picture. According to the method, the text paragraph structure restoration accuracy is improved.

The invention discloses an exposure control method and device, electronic equipment and a storage medium, and the method comprises the steps: carrying out the face detection of an image through a face detection algorithm to obtain a face region; determining a pixel average value of three channels of Y, U and V of the face area, and determining a brightness value on an RGB domain according to the pixel average value of the three channels of Y, U and V; determining an exposure parameter adjustment amount according to the brightness value on the RGB domain and a preset target brightness value, and performing exposure control based on the exposure parameter adjustment amount. According to the embodiment of the invention, the exposure parameter adjustment amount is determined according to the brightness value on the RGB domain and the preset target brightness value, and exposure control is carried out based on the exposure parameter adjustment amount, so the accuracy of exposure control is improved; meanwhile, the problem of relatively low accuracy of exposure adjustment caused by the fact that exposure parameters actually act on the Raw domain is avoided.

The invention provides a regional crowd trajectory determination method and device, and the method comprises the steps of obtaining pedestrian trajectory data of a to-be-analyzed region, and first attribute information of a direction region corresponding to a camera image in the to-be-analyzed region; determining pedestrian flow between two service areas corresponding to the direction area according to the pedestrian trajectory data and the first attribute information; and determining an area crowd track corresponding to the to-be-analyzed area according to the pedestrian flow and the pedestrian flow direction represented by the direction area. In the above scheme, cross-camera identification is not needed, and only the regional crowd trajectory needs to be analyzed for the picture of eachcamera, so that the problem that the accuracy of determining the regional crowd trajectory is relatively low due to low accuracy of cross-camera tracking identification is avoided, namely, the accuracy of determining the regional crowd trajectory is improved.

The invention discloses a skeleton line matching-based two-dimensional irregular contour layout method, which comprises the following steps of performing polygonal approximation on contours of a mother board and a sample piece to be arranged, and extracting all skeleton lines of the sample piece; performing similarity check on all the obtained skeleton lines by calculating Hu moment matching, andcombining the skeleton lines with similar shapes to obtain skeleton lines with simplified contours; and then determining the layout sequence of the sample pieces by calculating the Hu moment between the simplified skeleton line and the contour of the mother board according to the descending order of the matching degree of the mother board and the sample pieces and the principle of enabling the height of the layout sample pieces to be the lowest, and finally completing layout of any two-dimensional contour. The method is simple in algorithm implementation, does not need to carry out rotation and iterative calculation on the sample piece for many times, can obtain an ideal layout result only through one-time matching, is high in layout filling rate, consumes less time, and has a very ideal technical effect.

The invention discloses a building positioning method, device and equipment and a computer readable storage medium, and the method comprises the steps: collecting the basic position data of a full user, and obtaining the rough position data of the full user based on the basic position data; obtaining static state data when the user state of the total user in the rough position data is a static state, and associating the static state data with a preset building user sample to obtain a learning sample set; performing model training on all ECIs in the learning sample set to obtain a building positioning decision model corresponding to each ECI; and determining a target building positioning decision model corresponding to a target user in the total users based on each building positioning decision model, and inputting the basic position data of the target user into the target building positioning decision model for training to obtain a building positioning result of the target user. Therefore, the accuracy of positioning the user to the building is improved.

The invention discloses a correlation recommendation method, device and apparatus and a computer readable storage medium. The correlation recommendation method comprises the steps: determining the confidence degree between all to-be-recommended books and a target book, and determining a first recommendation sequence according to each confidence degree; calculating a similarity value between each to-be-recommended book and the target book, and determining a second recommendation sequence according to each similarity value; alternately sorting the first recommendation sequence and the second recommendation sequence to obtain a third recommendation sequence; and adjusting the third recommendation sequence according to the exposure click rate of each to-be-recommended book to obtain a target recommendation sequence so as to perform associated recommendation on the to-be-recommended books according to the target recommendation sequence when the target book is recommended. The accuracy of book recommendation is improved.

The invention relates to the technical field of finance technology, and discloses an exception handling detection method, which comprises the following steps of: obtaining a preset configuration tableand input parameters corresponding to target logic, and detecting whether the target logic meets a preset detection condition or not according to the preset configuration table and the input parameters; if yes, whether abnormal configuration exists in an abnormal class in the preset configuration table or not is detected; and if the abnormal configuration exists, constructing an abnormal scene based on the abnormal configuration, obtaining a target exception corresponding to the target logic according to the abnormal scene, and outputting the target exception. The invention further disclosesan exception handling detection device and equipment and a computer storage medium. According to the invention, the effectiveness of exception handling detection is improved.

The invention discloses a clothes humidity detection method and a clothes hanger, which are used for solving the problem of untimely collection caused by relatively low clothes drying degree sensing accuracy in the prior art. The clothes humidity detection method comprises the steps that a first instruction is sent to at least one first humidity sensor arranged on a clothes hanger multiple times,wherein the first instruction is used for requesting to obtain first data collected by the at least one first humidity sensor; a plurality of first data are received, and a first curve is determined according to the first data, wherein the first curve is used for indicating the change trend of the humidity of the clothes along with time; and the drying degree of the clothes is determined accordingto the first curve.

The invention provides an unmanned aerial vehicle flight control method and device, electronic equipment and a storage medium, and relates to the technical field of aerial photography. The method comprises the steps: acquiring the shot images of multiple historical time periods before the current moment, wherein each historical time period comprises multiple frames of shot images; respectively determining matching information of the multiple frames of shot images in each historical time period, wherein the matching information is used for representing the overlapping degree of the multiple frames of shot images; and adjusting the flight speed of the unmanned aerial vehicle at the current moment according to the matching information of the multiple frames of shot images in each historical time period. According to the scheme, the unmanned aerial vehicle can fly at the optimal flight speed to shoot the image, so that the overlapping rate of the shot image meets the requirement, the problem that the accuracy of the shot image obtained by flying at the fixed flight speed is low is solved, the accuracy of the aerial survey shot image is effectively improved, and the quality of image three-dimensional reconstruction is further improved.

The invention discloses a material ramming quantity statistical method, device and equipment and a computer readable storage medium, and the method comprises the steps: determining a target material pipe which needs to be subjected to material ramming quantity statistics in a plurality of material pipes based on an input selection instruction, and determining the type of target feed received by the target material pipe; determining the flow velocity of the target feed in the target feed pipe based on the type of the target feed; calculating the total feeding amount of the target feed pipe in a preset feeding duration according to the flow velocity, and detecting whether the total feeding amount is matched with a standard feeding amount corresponding to the target feed or not; and if the total feeding amount is matched with the standard feeding amount, outputting the total feeding amount. Therefore, the accuracy of the obtained material beating amount is improved.

The invention discloses a building positioning method, device, equipment and computer-readable storage medium. The building positioning method of the invention includes: collecting basic position data of all users, obtaining rough position data of all users based on the basic position data; obtaining rough positions Static state data when the user status of all users in the data is in a static state, correlating the static state data with preset building user samples to obtain a learning sample set; performing model training on all ECIs in the learning sample set to obtain The building positioning decision model corresponding to each of the ECIs; determine the target building positioning decision model corresponding to the target user in the total number of users based on each of the building positioning decision models, and input the basic location data of the target user to the target building The positioning decision model is trained to obtain the building positioning result of the target user. Thereby, the accuracy of the user's positioning to the building is improved.

The present invention relates to the technical field of image processing, and discloses a text paragraph structure restoration method, device, equipment and computer storage medium. The method includes: identifying a target picture, and determining all texts in the target picture based on the recognition result of the recognition. The text box and the text box position of each of the text boxes; sorting each of the text boxes according to the position of each of the text boxes, and inputting the text features of each of the text boxes to the preset based on the sorting result of the sorting The deep learning model is trained; based on the training results of the training, the text boxes are merged to obtain all the text paragraphs corresponding to the target picture. The invention improves the accuracy of text paragraph structure restoration.

The present invention provides a device testing method, device and electronic device. A server simulates an abnormal data link corresponding to a data link abnormality identifier, and buries the data link based on the buried point data and a preset standard corresponding to the data link abnormality identifier. The comparison results of the point data are obtained to obtain the data buried point test results. That is, when simulating a data link abnormality scenario, the present invention verifies whether the data embedding method is normal by whether the result fed back by the server is the preset standard data corresponding to the data link abnormality identifier. Since the present invention directly performs the data burying test under the abnormal data link scenario, the problem of low accuracy of the data burying test caused by the abnormal data link can be avoided.

A testing method and system for use with a battery management unit: in response to a test command, acquiring a target test case of a battery management system (BMS) (S101); then acquiring a test script corresponding to the target test case, the test script comprising simulation signal information (S102); thereafter, generating a simulation signal according to the simulation signal information (S103); then outputting the simulation signal to the BMS (S104); acquiring state information of the BMS after input of the simulation signal (S105); thereby obtaining a test result of the BMS according to the state information (S106). Therefore, the testing method and system for use with a battery management unit may solve the problem in the existing technology of wasted labor and low test efficiency which are caused by means of testing personnel manually simulating a simulation signal.

The invention discloses a positioning and navigation system and method of a robot and relates to the technical field of robots, aiming at solving the problem that the orientation judgment accuracy isrelatively low when a UWB (Ultra Wideband) technology is used for positioning and navigating the robot. The system comprises a plurality of labels, a base station, an electronic compass, a gyroscope and a controller. The method comprises the following steps: calculating a position coordinate of each label in real time and judging whether position coordinate data of each label is reliable or not; if so, determining a relation between a mounting position of each label and the orientation of the robot and combining the position coordinate of each label to determine an orientation angle of the robot; if not, judging whether detection data of the electronic compass is reliable or not; if the detection data is reliable, calculating according to the detection data of the electronic compass and determining the orientation angle of the robot; initializing the detection data of the gyroscope to be 0; when the detection data is not reliable, determining the orientation angle of the robot according to the detection data of the gyroscope. The positioning and navigation system and method of the robot can be applied to a positioning and navigation process of the robot.

The invention discloses an environment map generation method and device based on ultrasonic waves. The method comprises the following steps: judging whether a plurality of ultrasonic measurement values have conflicts or not; if so, forming a conflict sequence by at least two ultrasonic measurement values with conflicts, and initializing the confidence coefficient of each ultrasonic measurement value in the conflict sequence; according to the initialization confidence of the current ultrasonic measurement value in the conflict sequence and the initialization confidence of the ultrasonic measurement value conflicting with the current ultrasonic measurement value, updating the confidence of the current ultrasonic measurement value to obtain the target confidence of the current ultrasonic measurement value; removing the ultrasonic measurement value with the minimum target confidence in the conflict sequence; executing the above process circularly until no conflict exists in the remaining ultrasonic measurement values, and obtaining a target ultrasonic measurement value; and updating the historical environment map according to the current pose information determined by the pose information of the vehicle at the previous moment and the current running state of the vehicle and the target ultrasonic measurement value.

The embodiment of the invention provides a training method, an image quality detection method and device and a medium. The training method specifically comprises the following steps: determining a standard image containing a target; determining a test image corresponding to the standard image; determining a matching degree between the standard image and the test image by using a target recognition model; determining a label of the test image according to the matching degree; training a quality detection model according to the test image and the label of the test image; the quality detection model is used for determining a quality score corresponding to the test image; in the training process of the quality detection model, loss information is determined according to the label and the quality score, and parameters of the quality detection model are updated according to the loss information. According to the embodiment of the invention, the marking cost of the test image can be saved, and the accuracy and other performances of the quality detection model can be improved.

The invention discloses a homologous determination method and device for malicious files. The method and device are used for solving the problem of the lower accuracy of the homologous determination result in the prior art. The method comprises the steps that based on the characteristic behavior data information of each malicious file in the first class specified dimension, the category of each malicious file is obtained, and whether the malicious file in the corresponding category is the homologous malicious file or not is determined based on the characteristic behavior data information of each malicious file of each category in the second class specified dimension. Through the fusion analysis of the characteristic behavior data information in the multiple dimensions, the homologous malicious file is accurately judged and the accuracy of the homologous determination result is effectively improved. Moreover, after the determination of the category of the various malicious files, the homologous determination of the category of the various malicious files is conducted, so that the calculated amount of the homologous determination is decreased and the accuracy of the homologous determination result is further improved.

The invention relates to the technical field of financial science and technology, and discloses a data processing method, which comprises the following steps of: acquiring initial data of a pluralityof clients in a system data source according to an input data requirement, and performing data analysis processing on each initial data to obtain a plurality of index attribute data of each client; performing customer grouping processing on each customer according to a preset clustering model to obtain a customer group corresponding to each customer, and determining a group type of each customer group according to each index attribute data; and determining a target customer group corresponding to the data demand according to each group type, performing linear regression operation on the targetcustomer group based on a preset linear regression equation, and determining a target customer corresponding to the data demand according to an operation result of the linear regression operation. The invention further discloses a data processing device and equipment and a computer storage medium. According to the invention, the accuracy of screening target customers is improved.

The embodiment of the invention provides an interference detection method and device, electronic equipment and a storage medium, and relates to the technical field of communication. The method comprises the following steps: receiving channel state information reference signals (CSI-RSs) reported by at least one candidate terminal and at a plurality of moments; according to the CQI of the coincident sub-band in the CSI-RS at the multiple moments, determining the number of times that the coincident sub-band reported by each candidate terminal in the at least one candidate terminal is interfered; and determining an interfered sub-band in the overlapped sub-bands according to the interfered times of the overlapped sub-bands reported by each candidate terminal in the at least one candidate terminal. According to the embodiment of the invention, the problem of low accuracy caused by determining the interfered sub-band based on the CQI at a single moment can be avoided, a foundation is laid for solving the problem of low throughput caused by the interfered sub-band, and the service level of a network is improved.

The invention provides an equipment test method and device and electronic equipment, and a server simulates an abnormal data link corresponding to a data link abnormal identifier, and obtains a data buried point test result based on a comparison result of buried point data and preset standard buried point data corresponding to the data link abnormal identifier. According to the method, when the data link abnormal scene is simulated, whether the data buried point mode is normal or not is verified by determining whether the result fed back by the server is the preset standard data corresponding to the data link abnormal identifier or not. According to the invention, the data buried point test is directly carried out under the data link abnormal scene, so that the problem of low accuracy of the data buried point test caused by the data link abnormity can be avoided.

The invention provides a low-altitude multispectral remote sensing water quality monitoring method for a surface water body. The low-altitude multispectral remote sensing water quality monitoring method comprises the following steps: S1, acquiring a spectral curve of a monitored water body and corresponding water quality index data; s2, on the basis of the spectral curve and the water quality index data, selecting a wave band for monitoring, and establishing a multispectral remote sensing water body extraction model and a water quality inversion model; s3, performing outdoor acquisition work of low-altitude remote sensing data; s4, preprocessing the original data acquired by low-altitude remote sensing; s5, performing data processing through the multispectral remote sensing water body extraction model and the water quality inversion model; and S6, carrying out visual expression on a processing result. The method fills the vacancy of water quality monitoring in satellite remote sensing monitoring and ground monitoring methods, can be applied to daily conventional water quality monitoring, and can quickly obtain the surface water quality condition of the surface water body.

An embodiment of the present application provides a depth-of-field camera module, including a camera lens, a TOF sensor, an RGB sensor, a laser emitter, and an optical path conversion structure. The laser transmitter is used to emit infrared laser to the camera lens. The camera lens is used to transmit the infrared laser to the target object; receive the reflected light reflected by the target object and transmit the reflected light to the optical path conversion structure. Reflected light includes infrared laser light reflected by the target object and ambient light. The optical path conversion structure is used to divide at least part of the reflected light into first target light and second target light; emit the first target light to the TOF sensor and emit the second target light to the RGB sensor. The TOF sensor is used to obtain the depth information of the target object from the infrared laser in the first target light. The RGB sensor is used to obtain the RGB information of the target object from the ambient light of the second target light. This application can avoid the problem of low accuracy of the depth-of-field camera module due to the dual-camera structure.

The application provides an image classification method, device, computer equipment and storage medium, belonging to the technical field of image processing. The method includes: inputting a target image into an image classification model, classifying the target image based on the image classification model, and outputting a plurality of first confidence levels; determining at least two Second confidence levels of the at least two object types corresponding to the image cluster; according to the second confidence levels of the at least two object types, at least one object type corresponding to the target image is determined. The above technical solution avoids the problem of low accuracy caused by directly outputting the confidence that the target image belongs to the object type, and determines the second confidence of at least two object types through the first confidence of at least two image clusters that meet the target conditions. Confidence, and then determine at least one object type, so that the accuracy of the at least one object type is relatively high.

The invention discloses a method and equipment for determining a transmission line standard device and calibration equipment. The method comprises the steps of determining a to-be-calibrated band of to-be-calibrated equipment; selecting a first frequency in the to-be-calibrated band, and calculating a channel combine corresponding to the first frequency; determining the thickness of the transmission line standard device corresponding to the to-be-calibrated band according to a set phase change interval and the channel combine corresponding to the first frequency; and determining the transmission line standard device according to the thickness. The thickness of the transmission line standard device is determined according to the set phase change interval and the channel combine corresponding to the first frequency included in the to-be-calibrated band, and the transmission line standard device is determined based on the determined thickness, so that the thickness of the determined transmission line standard device is obviously increased, and the problem of low accuracy of calibration on a vector network analyzer with a TRL (Transistor Resistor Logic) method due to deformation of the transmission line standard device can be effectively solved.