165 results about "Standards characteristics" patented technology

The invention relates to a method for automatic evaluation based on generalized fluent spoken language fluency, which comprises the following steps of: acquiring speech data according to different ages and spoken language levels by using a speech input device; adopting an evaluating model based on characteristics of the generalized fluency and the machine learning training fluency; configuring a speech recognition system with corresponding parameters according to scripts of different subjects and genders of enunciators in the speech data; performing quantification on speech speed coherence, content understanding, advanced skills and reconstruction standard characteristics in the speech data to comprehensively extract the characteristics of the fluency from the speech data from the angle of expert assessment and evaluation; and adopting a decision tree method in regression fitting analysis and data mining to detect faults of abnormal fluency and grade and diagnose the fluency. The acquired score of the machine fluency can reach the level close to that of grading experts, and the relativity index exceeds that of 2 to 3 of general 5 experts; besides, the method has a high speed, and can be embedded into a spoken language automatic evaluation system to serve as an important module to evaluate fluency indexes in pronunciation quality.

The invention relates to an online automatic detection method of fabric defects based on machine vision and a device thereof, wherein the clear and integral standard fabric images without the defects of the fabric are shot on a batching and burling machine as a standard image template by utilizing an industrial camera in order to solve the problems of great labor intensity and poor detection quality of the traditional manual detection method; the corresponding characteristic parameters are extracted and stored into a computer; the range of the fabric defect detection characteristic parameters is set according to the requirements of users; the fabric images which are shot in real time are transmitted to the computer to be compared with the standard image template by utilizing the industrial camera and compared with the standard characteristic parameters; if the fabric characteristic parameters are not in conformity with the set requirements, the alarm is started by detection software; the defects are analyzed; and the printing of defects labels is executed. The invention has the advantages of high detection accuracy and high speed; and the labor intensity is greatly reduced.

The present invention provides an urban traffic system state visualization evaluation method based on traffic flow characteristics. Said method includes the following steps: A, in specific road section collecting real-time signal waveform whose continuous time is not less than three months, forming traffic flow state parameter; B, making traffic flow state parameter undergo the process of signal filter treatment so as to obtain traffic flow sample population; C, calculating average value and standard difference of all the samples in sample population, and using ratio value of standard difference and average value as sample discrete coefficient; D, according to traffic flow multidimensional state parameter strong aggregation characteristics forming and optimizing standard characteristic-phase plane graph for evaluation; and E, placing historical data, current data and predictive data of traffic system on the standard phase graph, and visually distinguishing running state of traffic system and its change tendency.

The invention discloses a method and a device for testing performance of a millimeter wave radar and a computer-readable storage medium. The method comprises the following steps: when the millimeter wave radar and a laser radar at preset mounting positions are in a dynamic scene, acquiring target characteristic data detected by the millimeter wave radar, and acquiring target point cloud data detected by the laser radar; according to the target point cloud data, determining coordinates of a target in a laser radar coordinate system, and according to a preset coordinate system converting relationship between the millimeter wave radar and the laser radar, converting the coordinates of the target in the laser radar coordinate system into coordinates of the target in a millimeter wave radar coordinate system to obtain standard characteristic data of the target; comparing the target characteristic data with the standard characteristic data of the target to determine an error of the target characteristic data detected by the millimeter wave radar. The performance of a vehicle-mounted millimeter wave radar is tested, so that the millimeter wave radar can be effectively used for accuratelydetecting an obstacle.

The invention discloses an information hiding method taking text information as a carrier, which is used for providing a hiding channel to secret information in the process of data transmission. The method comprises the following steps: self-defining a reserved code set according to a character coding standard characteristic; self-defining an L scale code according to a base number L of the reserved code set; self-defining a bijective function f of the L scale code to the reserved code set and a reverse function f<-1> thereof; self-defining a method of embedding reserved code character stringsin text information; embedding the secret information to a carrier of the text information and obtaining a secret object; and extracting the secret information from the secret object. The informationhiding technology takes the text information as the carrier and has strong versatility, large hiding capacity, good transparence and strong safety.

The invention discloses a health monitoring method for a numerical control machine tool, and the method comprises the steps: (1), executing a self-inspection G instruction, and collecting the operation state data; (2), building the mapping relation between the collected operation state data and the self-inspection G instruction, and forming an instruction domain mapping ensemble; (3), dividing the instruction domain mapping ensemble according to the self-inspection G instruction, extracting the instruction domain mapping under the self-inspection G instruction, obtaining the operation state data under the self-inspection G instruction, calculating an actual characteristic value according to the operation state data, comparing the actual characteristic value with a standard characteristic value, and obtaining a health index through calculation. The method improves the calculation basis of a key health index of the numerical control machine tool, the calculation method of the health index and the display mode of the health index, and can effectively solve a problem that the judgment of the health condition of the numerical control machine tool is not timely and accurate. Moreover, the obtained health index of the numerical control machine tool is displayed in a visualized manner, thereby improving the interaction effect of the numerical control machine tool with a person.

The invention discloses a non-intrusive load identification method and a non-intrusive load identification device. The non-intrusive load identification method comprises steps that electricity utilization behavior habits of users are counted to acquire first data; an electricity utilization load standard characteristic database is established; a voltage discrete signal and a current discrete signal sampled on a power supply entrance are acquired; the acquired voltage discrete signal and the acquired current discrete signal are preprocessed; load characteristics are calculated according to the transient voltages corresponding to various sampling points in the voltage discrete signal of the power supply entrance and the transient currents corresponding to the various sampling points in the current discrete signal of the power supply entrance; the load identification is carried out according to the load characteristics, and second data is acquired; the weighting of the first data and the second data is carried out, and the type of the load is identified. The technical problem of the prior art of the low accuracy of the load identification during non-intrusive load monitoring is solved.

The invention relates to a vehicle video characteristic extraction system and a vehicle video characteristic extraction method based on video structure description and belongs to the technical field of video processing. The system comprises a video streaming coding resolver, a vehicle bottom layer characteristic extractor, a vehicle video standard characteristic extractor, a license plate and vehicle logo divider, a license plate and vehicle logo recognizer and a video task support server. The method comprises the steps of firstly utilizing the video streaming coding resolver, the vehicle bottom layer characteristic extractor, the vehicle video standard characteristic extractor, the license plate and vehicle logo divider and the license plate and vehicle logo recognizer to extract vehicle characteristics in traffic monitoring video, then the video task support server uses a timestamp as a uniform identification to store information output by other modules to achieve information dynamic relevance, and the stored information are assembled to generate texts to be output to facilitate follow-up video access and quick index and location in processing to be capable of intelligently monitoring vehicle behaviors in road management in real time.

The invention discloses a distance measurement method and a device based on binocular vision. The method comprises steps of acquiring a first image acquired by a first light sensing element and a second image acquired by a second light sensing element at the same time; receiving a selected object, extracting a standard characteristic point of the selected object in the first image, carrying out violence matching on the second image according to the standard characteristic point and acquiring a matching characteristic point of the second image; according to position distribution of the standard characteristic point and the matching characteristic point, filtering the standard characteristic point and the matching characteristic point and establishing multiple line segments for connecting the standard characteristic point and the corresponding matching characteristic point after filtering; calculating length of each of the line segments and calculating multiple depth distances by means of binocular vision distance measurement according to the lengths of the filtered line segments; and calculating an average value of the depth distances and using the average value of the depth distances as a distance between the object and the sensing elements. Thus, while the distance between the object and a camera can be quickly obtained, calculation precision is ensured.

The embodiment of the invention relates to the field of computers and discloses a disease auxiliary diagnosis method based on image identification. The method comprises the steps of receiving an uploaded symptom image of a to-be-diagnosed patient; extracting to-be-diagnosed characteristics of the symptom image of the to-be-diagnosed patient by use of a deep learning algorithm; establishing a to-be-diagnosed characteristics set; comparing the to-be-diagnosed characteristics set with a standard characteristics set of a standard symptom image of each disease in a training gallery to acquire an image similarity between the symptom image of the to-be-diagnosed patient and the standard symptom image of each disease, and determining the probability of the to-be-diagnosed patient who suffers from each disease according to the image similarity. Through the above-mentioned mode, the embodiment can rapidly associate the to-be-diagnosed patient with one or more diseases and saves a lot of time for further definite diagnosis.

The invention provides a method for external optical compensation for an AMOLED, and the method comprises the following steps: S1, measuring the brightness difference between the pixels of a display screen through an image collected by the AMOLED; S2, adjusting a drive current according to the brightness first-gray standard characteristics described by a Gamma curve. The invention also provides a device for external optical compensation for the AMOLED. The beneficial effects of the invention are that the method improves the precision of compensation brightness measurement, is better in compensation effect, and is suitable for the detection of a large-size AMOLED.

The invention discloses an embedded high-speed online machine vision detection method, which comprises the steps of: acquiring the image signal of the measured object in the product line at high speed in real time; analyzing the acquired image of the measured object, separating and positioning the sensitive image area of easily appearing quality problem; picking up the characteristics of the sensitive image area of easily appearing quality problem; and the characteristics of the acquired sensitive image area are compared with the standard characteristics of the sensitive image area, and the measured object is estimated to be eligible or not according to that whether the similarity between two arrives the predicted value. In addition, the invention also discloses an embedded high-speed online machine vision detection device. Through the implementation of the invention, the production in the automatic production line can be detected at high speed, and the cost is low, which is conducive to the promotion and applications of wide range.

The invention discloses a guide system applied to a double-shield tunneling machine and a positioning method. The guide system is characterized by comprising a rear-view prism mounted on a duct piece wall, a total station mounted on the duct piece wall, an industrial computer mounted in a cab and provided with a guide system, a CCD measurement camera mounted on a support shield, LED standard characteristic points mounted on a front shield, and a laser target mounted on a tail shield. During the tunnel piercing construction, the CCD measurement camera extracts the variation of the light spot positions of the LED standard characteristic points, and compares the positions with the original positions calibrated in the CCD measurement camera, and thus the accurate position relation between the support shield and the front shield can be obtained. The guide system has the advantages that the laser beams sent by the total station directly irradiate the laser target mounted on the tail shield, and a special laser measurement window channel does not need to be reserved between the support shield and the tail shield, so that the measurement distance of the guide system is greatly prolonged, the station change frequency is reduced, and the construction efficiency is improved.

The invention discloses a cell characteristic vector based lithium ion cell configuration method, which comprises the following steps of: carrying out one-time 0.2C discharge on cells to be configured, followed by 2C charge and discharge, finally standing for a while, recording the 2C discharge capacity C0.5, the ratio CI of constant voltage capacity to constant current capacity during the 2C charge, the ratio V0 of recovery voltage to discharge lower limit voltage after the final step of standing, and the correlation coefficient rYS of the characteristic vector formed by the 2C constant current charge and 2C constant current discharge voltage values of the cells and the standard characteristic vector, determining an appropriate grading numerical range for grading according to the C0.5, CI, V0 and Rys, and configuring individual cells within the specified range. The method provided by the invention is used to fully present the difference between cells and screen out cells with similar consistency for assembling and usage, thus raising the performance of battery packs.

The invention discloses a pesticide identification method based on terahertz theoretical simulation spectrum. The method comprises the following steps of: obtaining a pesticide sample absorption spectrogram from the angle of theoretical simulation spectral analysis, and establishing a sample standard characteristic absorption peak fingerprint database X; collecting the terahertz time-domain spectrum of the unknown sample to be detected by use of a terahertz time-domain spectrum device; calculating the experimental absorption spectrogram of the unknown sample to be detected by use of the Fresnel formula; performing second-order differential processing of the experimental absorption spectrum of the unknown sample to be detected to obtain the second-order derivative absorption spectrogram of the unknown sample to be detected; extracting the characteristic absorption peak Y of the second-order derivative absorption spectrogram of the unknown sample to be detected; and performing the best match screening of the characteristic absorption peak Y of the second-order derivative absorption spectrogram of the unknown sample to be detected according to the established sample standard characteristic absorption peak fingerprint database X, so as to identify the unknown sample. The method disclosed by the invention reduces the difficulty and complexity of data processing, establishes a pesticide sample fingerprint database and improves the identification efficiency and accuracy.

The invention provides an unread message pushing displaying and sorting method and device. The method comprises the following steps: obtaining characteristic information in a plurality of received unread messages; carrying out primary analysis comparison on the characteristic information and first standard characteristic information stored in a user defined information pushing database so as to display positions of pushing the unread messages; when the matching degree between the characteristic information of the unread messages and the first standard characteristic information is less than a first preset threshold, carrying secondary analysis comparison on the characteristic information of the unread messages with second standard characteristic information stored in a third-party defined information pushing database; adjusting the display positions of pushing the unread messages according to the result of the secondary analysis comparison; and displaying the unread messages according to the adjusted pushing display positions. According to the unread message pushing displaying and sorting method and device provided by the invention, the users can more rapidly find the more important information pushing, so that the flexibility of the unread message sorting is improved, and the use of the users is convenient.

Provided are a breaker mechanical characteristic performance testing method and system. The breaker mechanical characteristic performance testing method includes that the standard characteristic quantity values with the reserved groups located on the same two-dimensional plane coordinate with characteristic quantity waveforms are selected by collecting the characteristic quantity waveforms of the mechanical characteristics in the breaker acting process, the distance between the characteristic quantity values in the characteristic quantity waveforms matched with the standard characteristic quantity values and the standard characteristic quantity values is calculated, and the offset distance of the characteristic quantity values is determined; and the distance between the standard characteristic quantity values and the original point of the two-dimensional plane coordinate is calculated, and the normalized offset distance is determined by dividing the offset distance of the characteristic quantity values by the distance between the standard characteristic quantity values and the original point of the two-dimensional plane coordinate. Users can obtain the values of the offset degree and the similarity of the characteristic quantity values according to calculation, and thereby evaluating the breaker mechanical performance and forecasting potential fault hidden dangers of the breaker mechanical performance. In addition, the breaker characteristic quantity value deterioration degree can be evaluated in real time according to the calculated offset distance of the characteristic quantity values.

A recognizing and positioning method of ground buildings belongs to the imaging automatic target recognition field, which aims at solving the problem of recognition and positioning from different viewpoints and from different scales, and different heights, and to be used in forward looking ground buildings. The invention constructs a ground building standard feature library in advance. The sequence includes: an enhanced image procedure, a background suppression processing procedure, a gray level merge procedure, a feedback and segmentation procedure, a vertical bar feature detection procedure, and a quadratic character matching procedure. The invention further extracts characteristic quantity to match with the standard characteristic, considers recognizing the veins and scene information of the buildings, and recognizes and positions the forward looking ground buildings, aiming at the characteristic of ground buildings, and making use of mathematical morphology to extract structure information of image. The novel method has high precision of recognition, good reliability, and is used in fields such as urban planning, supervision, aircraft contact navigation, collision-avoidance to recognize the forward looking ground buildings of different viewpoints, different scales and different heights.

The invention discloses a method and a system for identifying a linear bar code. The method comprises the following steps of acquiring a image and examining the bar code in the acquired image; locating the bar code in the image to determine the border of the bar code, estimating the blurring degree of the image; according to the blurring degree, choosing a corresponding standard characteristic vector set from a plurality of standard characteristic vector sets, each of which corresponds to a blurring degree scope and includes a plurality of standard characteristic vectors; based on the border of the bar code, extracting the testing characteristic vectors corresponding to all the bar code data characters of the bar code from the image; and utilizing the testing characteristic vectors corresponding to all the bar code data characters of the bar code and the chosen standard characteristic vector set to decode and obtain the bar code value of the bar code. By setting different standard characteristic vector sets according to different blurring degree scopes, the method for identifying the linear bar code can still decode the bar code correctly under the condition of a blurring image.

The invention discloses a vehicle monitoring and tracking method, system and server. The method comprises the steps that the server receives a vehicle tracking request message sent by a terminal device, coordinates of a reference position are set as coordinates of a geographic position in the message, and a first course and a second course are started; a target camera is determined in the first course, a target vehicle is searched for, characteristic vectors of last M characters in a license plate area of the target vehicle are extracted, after it is judged that the percentage of the similarity of the obtained M characteristic vectors and M standard characteristic vectors corresponding to the last M characters in license plate information is larger than a threshold value, image data and the identification of the target camera which the image data belong to are recorded, and prompts are sent out; in the second course, network information sent by the terminal device is monitored, the first course is terminated after a tracking confirmation message is received from the terminal device, the coordinates of the reference position are updated to be the position coordinates in the message, and the first course is restarted. The vehicle monitoring and tracking method, system and server reduce the workload of monitoring personnel and improve the intelligent degree of the monitoring system.

A method for detecting form and position tolerance of components comprises the following steps: (a) acquiring a chart file of a component to be detected and characteristic elements to be detected which are selected from the chart file of the component to be detected by a user; (b) performing triangular gridding on the chart file of the component to be detected, and acquiring point cloud data of the chart file of the component to be detected; (c) acquiring standard characteristic elements; (d) extracting the point clod which is fitted into characteristic elements from the point cloud data of the chart file of the components to be detected; (f) calculating the form and position tolerance between the characteristic elements and the standard characteristic elements; and (g) outputting a form and position tolerance analysis table. By adopting the invention, the form and position tolerance of components can be automatically detected.

Disclosed are a method and device for obtaining web page category standards, and a method and device for categorizing web pages. Tag contents are extracted from sample pages; standard characteristic terms are extracted from the tag contents; on the basis of the standard characteristic terms extracted from the tag contents, a list of standard categories and standard weights of standard characteristic terms, i.e. the standard for web page categories, are obtained; web pages to be categorized are categorized on the basis of the standard.

The invention relates to a positioning device of an automatic charging robot, which comprises a control processing module, a storage module and a shooting module, wherein the storage module is used for storing standard characteristic values of electric vehicles in various models and standard control parameters of the charging robot corresponding to the standard characteristic values; the shooting module is used for shooting the to-be-charged vehicles and transmitting pictures to the control processing module; and the control processing module performs image processing on picture information to obtain actual characteristic values, compares and analyzes the actual characteristic values and the standard characteristic values to obtain control deviations, obtains actual control parameters according to the standard control parameters and the control deviations, and accomplishes positioning control on the charging robot according to the actual control parameters. The positioning device has the technical benefits of automation, no labor participation, construction cost saving and high charging efficiency.

The invention discloses an LED screen dead pixel monitoring and early warning method and system. The method comprises the following steps: a. setting the standard characteristic value of the dead pixel amount of a target LED display screen; b. detecting the display data of each pixel point of the target LED display screen; c. according to the display data, calculating the practical characteristic value of the dead pixel amount; d. comparing the practical characteristic value with the standard characteristic value; when the practical characteristic value is more than the standard characteristic value, outputting a comparison result; and e. according to the comparison result, generating early warning signal data, and giving out early warning. The system is correspondingly provided with a detection unit, a control unit, an early warning unit and a regulation unit. When the practical characteristic value of the dead pixel amount of the target LED display screen exceeds the preset standard characteristic value, the invention gives out early warning so as to ensure lower replacement difficulty and lower replacement and maintenance cost.