240results about How to "High discrimination accuracy" patented technology

This optical receiving device for discriminating and recovering a data signal, which results from converting an optical signal input through a dispersion equalizer into an electrical signal and amplifying it to a pre-determined amplitude, by using a clock and data recovery circuit for discriminating a data signal at the decision point controlled to achieve the optimum position controls the dispersion characteristics of a dispersion equalizer so that the error count in the recovered data signal by using a clock and data recovery circuit will be minimized by controlling the eye pattern of the data signal which has been amplified to a pre-determined amplitude.

The invention discloses an intelligent early warning and fire extinguishing robot. The intelligent early warning and fire extinguishing robot is formed in the mode that a temperature sensor, a smoke sensor, a flame sensor, a pan-and-tilt camera, a two-freedom-degree mechanical arm, a fire extinguisher and a controller are installed on a wheel type vehicle, the two-freedom-degree mechanical arm is installed on the side of the fire extinguisher in a surrounding mode, and the temperature sensor, the smoke sensor, the flame sensor, the pan-and-tilt camera, the two-freedom-degree mechanical arm are connected with the controller in an electric loop mode. The intelligent early warning and fire extinguishing robot is high in anti-interference ability, good in space applicability, capable of monitoring fire flame videos of a tiny fire source at the beginning of a fire, high in fire early warning accurate and low in false alarm rate.

An analyzing means first calculates the blur direction and blur level of a digital photograph image based on edge widths and histograms of the edge widths obtained in each direction, and discriminates whether the image is a blurred image or a normal image. Then, it further calculates the camera shake level and blur width for the image discriminated to be a blurred image. A parameter setting means sets a one-dimensional correction mask and a two-dimensional correction mask based on the blur width. It also sets a correction level based on the blur level. It further makes an adjustment to the ratio between the one-dimensional and two-dimensional correction masks based on the camera shake level.

An emotion is decided based on both image and voice data, and then a decorated image or a substitute image is outputted. Further, a segment of voice signal is precisely determined for the analysis of the signal. Emotion analysis is conducted along with operations of extracting constituent elements of an image and continuously monitoring motions of the elements. A period during which no motion of lips is observed and a period during which no voice is inputted are used as a dividing point for voice signal, and an emotion in voice is decided. Furthermore, the result from the analysis of the image data and the result from the analysis of the voice data are weighted to eventually determine the emotion, and a synthesized image or a substitute image corresponding to the emotion is outputted.

A method of and a system for splitting a compound object using multi-energy CT data including a density and an atomic number measurements are provided. The method comprises: compound object detection; computing a two-dimensional DZ distribution of a compound object; identifying clusters within the DZ distribution; assigning a component label to each object voxel based on the DZ distribution clusters; and post-processing the set of voxels identified as belonging to each component.

The invention discloses an intelligent correcting method, comprising: dividing the answer area of an electronic version test paper or an electronic version answer sheet by using different identification areas according to the correct answers which are numbers, letters and symbols, and associating the page numbers, the title numbers and the answering area positions; at the same time, presetting thecorrect answers, then obtaining the handwritten answer handwriting information written in the answering area on the paper test paper or the paper answer sheet, finally identifying the answer handwriting information, comparing with the preset correct answers, and determining correct or not. The invention, which may predict in advance question types such as an objective question, a fill-in-the-blank question, and a short answer question with the correct answer of a number, a letter, and a symbol, narrows the recognition range, improves the recognition accuracy, and lightens the correcting burden of teachers by setting different types of answer areas, and promotes the transformation of the classroom teaching form in the case of reducing the negative emotions generated by the teachers duringthe teaching process, by manually correcting and counting a large number of daily assignments, thereby achieving a truly student-oriented precision teaching.

The embodiment of the invention discloses an expression recognition method. The expression recognition method comprises the steps that a face image to be recognized is input to a global depth convolution nerve network to determine a first expression and a first value of the face image, wherein the global depth convolution nerve network is obtained by deep learning a reference face image with different expressions; the face image is partitioned into multiple sub-images; feature information of the sub-images is compared with reference feature information to determine a second expression and a second value of the face image, wherein the reference feature information is determined by multiple reference sub-images of the reference face image with the different expressions; the expression of the face image is determined according to the first expression, the first value, the second expression and the second value. Correspondingly, the invention further provides an expression recognition system. According to the method and system, the face expression can be more accurately and more quickly recognized.

The invention discloses a kind of recognition system of vehicle ID of ETC electron without stopping to charging, which includes IC card of vehicle-carrying radio frequency containing the basic information of vehicle and several information channels connect with several sensors, and the information channels is combined by anabranch model of play chess information and the ratiocination algorithm of Bayes's network to complete the precious identification for vehicle ID. The invention anabranches the information of vehicle ID in non-contact intelligent IC card based on the anabranch theory of playing chess information, and the several source information can reach the optimized reliable judgment by the competition and the corporation in the anabranch model, and identify the real ID of vehicle, and ensures the reliability of identification, and the precision can reach above 99.99%, and the invest is little, and the reliability is well. The recognition system of vehicle ID does not occupy the source of host computer and improve the work efficiency of host computer. The identifying device is small and easy to install.

An object is to provide an image detector for bank notes which enables the cost to be lowered. Accordingly this has: a light emitting device 31 (31X) which is arranged facing an image detection sensor 24 (24X), and which irradiates light of a plurality of different wavelengths ranges towards a bank note on a bank note transportation path 12, and detects light of the light which has transmitted through the bank note with an image detection sensor 24 (24X); a light emitting device 27 (27X) which is provided on the same side as the image detection sensor 24 (24X), which irradiates light of a plurality of different wavelengths ranges towards a bank note S, and detects light of the light which is reflected from the bank note with the image detection sensor 24 (24X); an image detection sensor 24 (24Y) provided on the opposite side to the image detection sensor 24 (24X); and a third light emitting device 27 (27Y) provided on the same side as the second image detection sensor 24 (24Y), which irradiates light of a plurality of different wavelength ranges towards the bank note S, and detects light of the light which is reflected from the bank note S with the image detection image detection sensor 24 (24Y).

Ultrasonic pipeline inspection apparatus which utilises two different wave modes (e.g. a horizontally polarised shear mode and a symmetric Lamb wave mode) to discriminate different defect types by comparing corresponding signals collected from defects. EMATs may detect the signals. The detected amplitudes of the collected signals from both wave modes are compared to calculate a ratio, which may be compared with a distribution of ratio values for known defects to evaluate the type of defect detected. The apparatus may be a sensor module mountable on a pipeline vehicle and having a plurality of sensors distributed around its periphery.

The invention discloses a fault diagnosis comprehensive positioning method for an intelligent distribution network, and belongs to the field of fault monitoring. Based on a fault indicator, a wirelesscommunication technology and a ground fault line selection technology, the method comprises the following steps: monitoring zero-sequence voltage of a transformer neutral point in real time, startingzero-sequence fault diagnosis when detecting that a zero-sequence voltage change value exceeds a preset value, and identifying whether the zero-sequence voltage change value is a single-phase grounding fault; if the zero-sequence voltage change value is single-phase grounding fault, performing grounding line selection through a fault diagnosis device in a station, and positioning the grounding fault on a certain phase of a certain line; when the single-phase earth fault is judged to be a stable fault, recalling the real-time current data of the fault indicators on the fault phase line, comparing the signals collected by the fault indicators, comparing the difference of the real-time current data before and after a fault point, judging the two fault indicators with abrupt changes, and thendetermining the specific section of the fault. The method can be widely applied to the field of power distribution network operation fault monitoring and positioning.

The invention discloses a spam mail filtering system and method based on clusters. The system comprises a cluster module, a mail training module, a characteristic base storage module and a mail filtering module. The method includes that S1.1 training mails and unmarked mails are obtained from a mail backup system; S1.2 a cluster preprocessing module preprocesses the mails; S1.3 a cluster analyzing module divides the preprocessed mails into different clusters; S1.4 a cluster central computing module computes vector expression of the clusters; S1.5 the training module studies marked mails in the clusters and updates characteristic bases corresponding to the clusters; S2.1 mails required to be filtered are obtained from a mail system; S2.2 a mail class attribute distinguishing module calculates the cluster most similar to mail content; S2.3 a mail characteristic extraction module conducts characteristic extraction on the mails to be distinguished; S2.4 a mail distinguishing module shows a distinguishing result according to the mail characteristics and the corresponding characteristic bases. The system and method has the advantages of being high in extraction speed, high in accuracy and good in effect.

The invention relates to a video-based automatic identification system for abnormal conditions of a fused magnesium furnace. The method includes the steps of acquiring fused magnesia production fieldcondition information through an image information acquisition module and transmitting the information to a sample generation module; generating a training sample by the sample generation module, performing label processing and manual classification label processing on images through label-image software, and transmitting the processed video and image information to a detection and classificationmodule; performing, by the detection and classification module, feature extraction and processing on the video and image information through a corresponding algorithm to obtain an intelligent detection and classification model of the condition identification system; and displaying, by a display unit, the novel uses the test video and image information in a visualized manner by using the intelligent detection and classification model processed by the detection and classification module. The video-based automatic identification system of the invention has obviously improved identification effects as compared to a system using non-visual information alone such as the current and voltage, is reasonable in hardware device structure, convenient in operation, low in cost and high in the accuracyof discrimination and can replace manual routing inspection.

A analog / digital modulation type discrimination circuit 1 discriminates whether a reception signal is an analog modulation type, a linear modulation type or a non-linear modulation type by digital modulation type. In case where the reception signal is discriminated to be the analog modulation type, an analog modulation type discrimination circuit 2 discriminates whether it is an AM signal or an FM signal among the analog modulation type. In case where the reception signal is discriminated to be the linear modulation type by digital modulation type, a linear modulation type discrimination circuit 3 discriminates whether it is a BPSK signal, a QPSK signal, a π / 4-shift QPSK signal, an 8-PSK signal, an M-ary PSK signal of multi-level exceeding 8-levels, a 16 QAM signal or a n M-ary QAM signal of multi-level exceeding 16-levels among the linear modulation type by digital modulation type. In case where the reception signal is discriminated to be the non-linear modulation type by the digital modulation type a non-linear modulation type discrimination circuit 4 discriminates whether it is an M-ary FSK signal, a 2-FSK signal, an MSK signal or a GMSK signal among the non-linear modulation type by the digital modulation type.

The value of the pixel at the same position of each of the training input image as well as a plurality of training feature images is inputted into the discrimination device, which learns in such a way as to reduce the error between the output value obtained from the discrimination device and the value of the pixel at the aforementioned pixel position in the training output image. At the time of enhancement processing, the feature image is produced from the image to be processed, and the values of the pixels of these images at the same position are inputted into the discrimination device, thereby outputting the enhanced image wherein the value outputted from this discrimination device is set as the value of the pixel at the aforementioned pixel position.

An information processing apparatus includes a face detecting unit that detects a face area included in image data, a face-component detecting unit that detects a face component from the face area detected by the face detecting unit, and a line-of-sight discriminating unit that executes line-of-sight discrimination processing for a face image from which the face component is detected by the face-component detecting unit. The line-of-sight discriminating unit executes processing for discriminating whether a line of sight of the face image data from which the face component is detected is in a positive state in which a line of sight is directed in a camera direction or a negative state in which a line of sight is not directed in a camera direction according to collation processing for a line-of-sight discrimination dictionary in which learning data including classification data corresponding to the respective states are stored and input face image data.

The invention relates to the field of milk powder quality detection and more specifically relates to a method for distinguishing milk powder original production places through an elemental analysis-stable isotope mass spectrometry. The method is characterized by comprising the following steps: S1, obtaining samples of milk powder imported from a plurality of Countries; S2, respectively performingstable carbon and nitrogen isotope detection on a plurality of samples, performing stable hydrogen and oxygen isotope detection and performing analysis; S3, performing microelement analysis on the plurality of samples; S4, utilizing a principle component analysis method for performing principle component analysis on elements in the plurality of the samples and primarily distinguishing the samplesof the milk powder from different original production places; S5, utilizing original production place information of the plurality of samples and analysis data in the S3 and S4 and discriminant analysis to perform statistical modeling to obtain an original production place discrimination model coefficient matrix and corresponding original production place discrimination prediction accuracy; S6, bringing unknown samples into the model to perform original production place discrimination and finally judging original production place attributes. By means of the method, original production place traceability and discrimination can be primarily performed on the milk powder; thus, convenience is brought to checking of milk powder quality and fighting of counterfeit and shoddy products.

The invention discloses a telecommunication fraud method for virtual number segments. The method includes the following steps that: the telecommunication fraud case data of the police and the communication data of an operator are obtained; data matching is carried out, so that an initial training set can be obtained; a random oversampling method is adopted to obtain positive and negative class equilibrium samples; the positive and negative class equilibrium samples are trained, so that a prediction model can be obtained; and the virtual number segment incoming call number communication characteristics of a user are introduced into the prediction model, so that a judgment result can be obtained. According to the method of the invention, the basic training samples are generated according to the existing communication data and the case data of the public security bureau; an Adaboost model is adopted to train samples based on the basic training samples, so that the telecommunication fraud judgment model can be obtained; calls are judged according to the judgment model; and therefore, with the method adopted, whether a virtual number segment is a telecommunication fraud can be judged accurately and fast according to the virtual number segment. The method has the advantages of high judgment accuracy, quick and convenient calculation and high applicability.

The invention relates to an application layer DDOS (distributed denial of service) attack and defense method, which is easy to realize, is low in time complexity and high in accuracy, is transparent to users, and has no influence on the user visit. The method provided by the invention is divided to two stages of a training stage and a work stage. In the training stage, real and legal visit flow is adopted as training data to generate a standard array used for real-time detection and protection of the work stage. According to the invention, application layer DDOS equipment needs to be arranged in front of an application server in series, so that the visit flow can be filtered through a defense device before entering the server.

The invention introduces an abnormal power consumption detection method and device, equipment and a computer readable storage medium, and the method comprises the steps: carrying out the clustering and grouping of all to-be-detected users through a preset clustering algorithm according to the obtained power consumption data of all to-be-detected users; calculating an abnormal power utilization index of each to-be-detected user according to a clustering grouping result, and determining suspected abnormal power utilization users according to the abnormal power utilization index of each to-be-detected user; and determining a real abnormal power consumption user from the suspected abnormal power consumption users through a preset classifier. According to the invention, the workload of manual on-site troubleshooting of abnormal users is reduced, and the discrimination accuracy of abnormal electricity users is improved.

Disclosed are an energy spectrum modulation apparatus, a material discrimination method and a device thereof, as well as an image processing method are disclosed, which can discriminate the material in large- and medium-sized objects such as cargo containers, air cargo containers, etc. by using X-rays having different energy levels. The energy spectrum modulation apparatus comprises a first energy spectrum modulation part for modulating a first ray having a first energy spectrum, and a second energy spectrum modulation part coupled to the first energy spectrum modulation part and for modulating a second ray having a second energy spectrum different from the first energy spectrum. The present invention can be used in the non-opening inspection for large-sized container cargo at places such as Customs, ports and airports.

An image processing apparatus includes a document type automatic discrimination section that, based upon input image data read from a document, automatically discriminates a document type of the document. The document type automatic discrimination section, based upon plural types of parameters (quantity of halftone pixels, quantity of photographic-picture pixels, and quantity of high density pixels) obtained from plural types of characteristics (maximum density difference, total density busyness) extracted from the input image data and which parameters are used for discriminating a document type. A specific inkjet printed image whose output image data would not reach a standard level, if a process for a halftone reproduction region or a process for a continuous tone region were to be carried out on the input image data. This can improve the accuracy of document discrimination and the image quality of reproduced image.

The invention discloses a method for detecting and positioning of cancer regions with small samples or unbalanced samples. By combining the characteristics of the histopathological image containing cancer cells, the method enhances the data set by using a method of adding noise, rotating, increasing or reducing brightness, expands the data set, balances the specific gravity of the label type of the training set, and improves the training effect of the classifier. The method is characterized in that an Inception V2 network is used as a basis; aiming at the conditions of few samples and unbalanced samples, iterative training is carried out in combination with a small number of calibrated cancerous tissue image blocks and WSI pathological images; the method is advantaged in that WSI image-level pathological image cancer area detection and positioning can be accomplished with high accuracy; problems of over-fitting, local optimization and gradient disappearance of the over-deep neural network easily caused by too few labeled training samples are solved; and accuracy of the training result and usability of the network under the same level are improved.

The invention discloses a dynamic supervising method for gas-solid fluidized bed and the system. The method is: using wavelet analysis technology to analyze the pressure vibration signal, judging its flow state, displaying the flow state with capacitance chromatography imaging technology, measuring the aperture rate and its vibration of the fluidized bed, extracts the character value and carries on flow state judgment, using multi-sensor information melting technology to carry on flow state judgment, upgrading the accuracy of the judgment. The supervising system has a fluidized bed main body, under which there has a gas distribution board, gas chamber, on the external wall in the middle of the main body there arranges a capacitance chromatography imaging sensor, pressure sensor, the capacitance chromatography imaging sensor is connected to the tiny capacitance measuring model, data collecting system, and computer, the pressure sensor is connected to A / D converting card, and computer.

The invention provides a driving fatigue detection method and apparatus based on computer vision. The method includes the following steps: an on-board camera apparatus real-time acquiring an original video image of a driver; extracting each image frame containing a face part image of the driver from the original video image and each image frame containing the face part image serving as an object image frame; for the object image frame, positioning face part characteristic points so as to determine face part characteristic points, then on the basis of the obtained face part characteristic points, intercepting the object image frame and analyzing the object sub-images so as to determine whether the driver is driving at fatigue, wherein positioning face part characteristic points is conducted based on cascade regressions operation. According to the invention, the method and the apparatus have high accuracy and convenience.

The invention discloses a method and device for detecting a superposed stripe in a video image. The method comprises steps of: acquiring Y components in the current frame and the last frame of a video image in a YUV format, and calculating the difference data between the Y components of the two frames of the video image; smoothening the difference data to obtain a difference image; determining the global gradient and the local frequency domain of the difference image in order to determine whether the video image is provided with a superposed stripe. The device comprises a difference module, a smoothening module, a gradient determining module, and a frequency domain determining module. The method and the device shorten consumed time while improving determining accuracy so as to satisfy a real-time application requirement.

A coin denomination discriminating device for improving coin dispensing devices which incorporates components formed of non-magnetic material including a non-magnetic rotor having a coin receiving portion and a non-magnetic reference guide to guide coins from a storage coin holding bowl to a coin conveying device capable of selectively releasing coins of various denominations. Magnetic sensor units can be positioned along the coin movement path to provide characteristic signals representative of the coins independent of influence of metallic components in the transporting path. The non-magnetic members can be formed from wear resistant plastic material and relatively inexpensive magnetic sensors with coil and ferrite coils can be utilized adjacent the rotor.