30results about How to "Simplify identification steps" patented technology

The invention provides a method, device and apparatus for segmenting a person in a video, and a computer storage medium.,The method comprises the following steps of: acquiring a face image of a targetperson; obtaining all the individual images corresponding to each target person from the first video according to the face images of the target person; combining the individual images of the same target person are sequentially combined to obtain a second video corresponding to each target person. The invention can obtain the single-person video containing only the target person from the video containing a plurality of persons, thereby realizing the purposes of reducing the labor cost, simplifying the identification step, and improving the identification accuracy and the identification speed.

The invention discloses a multi-scale convolutional neural network based radio signal modulation identification method. The multi-scale convolutional neural network based radio signal modulation identification method comprises the steps of (1) generating a processed radio modulation signal; (2) generating a two-dimensional time-frequency diagram and performing Fourier transform on an instantaneouscorrelation function of the signal to obtain a Wigner-Ville time-frequency distribution diagram of the signals; (3) performing pre-processing on the time-frequency distribution diagram to generate atraining sample set and a test sample set; (4) building a multi-scale convolutional neural network module and training the model; and (5) testing the test set by utilizing the trained network model, calculating the correction rate, obtaining an identification accuracy rate and assessing the network performance. The multi-scale convolutional neural network based radio signal modulation identification method has the advantages of strong universality, no need for manual characteristic extraction and a plenty of priori knowledge, low complexity and accurate and stable classification results, and can be used in the field of signal classification identification technologies.

The invention discloses a method for identifying maize haploids. The method for identifying maize haploids provided by the invention comprises the following steps: (1) taking whether each corn kernel in a training set is a real haploid or a real diploid as a dependent variable, taking a near-infrared transmittance spectroscopy absorption value of each corn kernel as an independent variable, and establishing a haploid identification model by using a machine learning algorithm, wherein the training set is composed of a plurality of corn kernels, one part of the corn kernels are the real haploids, and the other part of the corn kernels are the real diploids; and (2) taking to-be-detected corn kernels, acquiring the near-infrared transmittance spectroscopy absorption values, inputting the absorption values into the haploid identification model established in the step (1), and outputting the result that the to-be-detected corn kernels are predicted haploids or predicted diploids by the model. The method provided by the invention can be used for automatic identification and has significances for promoting the maize haploid breeding technique engineering.

The invention relates to a method for detecting the quality of Zaoren Anshen Capsules based on thin-layer chromatography. A , salvianolic acid B, schisandrin A, schisandrin A and other seven main active ingredients were qualitatively detected, specifically including the preparation of the test solution, the preparation of the reference solution, the preparation of the reference medicinal material solution and TLC retrieval steps. By optimizing the preparation method and development conditions of the test sample for Zaoren Anshen Capsule TLC, the simultaneous detection of the whole medicinal flavor of the compound was realized on one TLC board, which simplified the identification steps, improved the identification efficiency, and promoted the development of Zaoren Anshen Capsules. overall quality control.

The invention discloses a pair of primers, a kit and an application thereof for identifying the mating type of the protoplast monokaryon of Piacrotomus ciliformis. The invention discloses two primer pairs, PMD_1 and PMD_2, for identifying the mating type of the protoplast monokaryon of the Piacrotomus grisa. In the present invention, the genome DNA of the protoplast monokaryon of the Piacrotomus gracilis to be identified is used as a template, amplified with PMD_1 and PMD_2 primers, and the mating type is determined according to the size of the detected PCR product. The method of the present invention is simple and easy, shortens the identification time and simplifies the identification steps, has high accuracy, and is applicable to the research of edible and medicinal fungi, especially the genetic breeding of Piacrotomus pilaris, genomics and other related fields.

The invention discloses an object recognition method and device, and relates to the technical field of image recognition. A specific embodiment of the method comprises the following steps: determining the position of a localizer in a given image, wherein the localizer is used for locating an object to be identified in the given image; according to the position of the localizer, segmenting an area corresponding to the object to be identified from the given image to obtain a slice image; and inputting the slice image into a classifier to obtain an identification result of the to-be-identified object. According to the embodiment, the object can be accurately recognized under the condition that the content of the object is diversified, fuzzy matching is not needed, and recognition steps are simplified.

The embodiment of the present invention discloses a cable partial discharge signal identification method and device based on S-transformation, which is used to solve the problem that the waveform time difference method is usually used in the current location of the partial discharge signal source of the power cable, and the identification accuracy is low and the required identification time is too long. Long technical question. The method in the embodiment of the present invention includes: performing S-transformation on the acquired partial discharge signal from a known source to obtain a complex time-frequency matrix; performing modulus on the complex time-frequency matrix to obtain a modulus matrix, and performing singular value decomposition on the modulus matrix to obtain The singular value sequence of the modulus matrix; the singular value sequence is divided into at least two intervals, the ratio of the Shannon entropy of the singular value in each interval to the Shannon entropy of the singular value sequence is calculated, and the partial discharge signal eigenvector is established with the ratio Partial discharge signal feature sample library and build a support vector machine model; input the feature vector of the partial discharge signal to be identified into the support vector machine model to obtain the source of the partial discharge signal to be identified.

The invention discloses a method and device for realizing telephone transfer according to language. The method comprises steps: the call center server transfers the called party to the first seat phone, and records the conversation process within a period of time; adopts the first The first recording file is identified and judged by the language recognition method; the first recording file is recognized and judged by the second language recognition method; after the caller completes the call, the caller is prompted and receives the feedback information of the caller; the caller hangs up, The call center server records the calling number and the corresponding agent phone number. The beneficial effect of implementing the method and device of the present invention is that it can be automatically transferred to the service group of the same language according to the language used by the customer. At the same time, the double language identification method is adopted, which not only ensures the accuracy of the calling language identification, but also simplifies the common The step of language recognition saves system resources of the call center.

The present invention relates to a method for quickly identifying Chaxiong based on an electronic nose, comprising: (1) powdering a certain amount of Chaxiong, Chuanxiong, and a mixture of Chaxiong and Chuanxiong mixed in gradient proportions to obtain multiple samples; (2) opening Electronic nose instrument, the electronic nose contains multiple sensors; set the parameters of the electronic nose to the gas flow rate of 0.5-0.7L / min, the cleaning time of 160-200s, the detection time of 100-150s, and start the sample injection measurement at 15-30°C. Obtain the data of each sample in each sensor of the electronic nose; (3) Find out a group of data with the highest significance corresponding to different samples in different sensors, and use the sensor corresponding to this group of data as the subsequent electronic nose detection sensor; The test sample is powdered and the detection data of the test sample in the sensor is determined according to the conditions of step (2); (4) According to the electronic nose detection data and a set of data with the highest significance as the characteristic reference value, identify the test sample . The invention has simple operation, simplifies identification steps and shortens identification time.

The present invention provides a method, device, device and computer storage medium for identifying abnormal traffic lights in the technical field of automatic driving. The state sequence is matched in the preset abnormal features, and when the matching result is obtained, it is determined that the traffic signal is an abnormal traffic signal; the position information of the abnormal traffic signal is obtained and recorded. The invention can simplify the identification steps of abnormal traffic lights and improve the identification efficiency of abnormal traffic lights.

The invention discloses an identification method for an 8PSK (8 Phase Shift Keying) signal and a PI / 4-DQPSK (PI / 4-Differential Quadrature Phase Shift Keying) signal. The identification method comprises the following steps: extracting an instantaneous phase from a down-converted I / Q signal; calculating phase difference of adjacent code elements based on the instantaneous phase; calculating center normalized instantaneous phase difference based on the phase difference of the adjacent code elements; calculating a parameter based on the center normalized instantaneous phase difference; calculating absolute value standard derivation of a nonlinear component of the parameter; identifying the signal according to the absolute value standard derivation, wherein the signal is the PI / 4-DQPSK signal when the absolute value standard derivation is equal to 0 and is the 8PSK signal when the absolute value standard derivation is greater than zero. According to the technical scheme, the 8PSK signal and the PI / 4-DQPSK signal can be easily, conveniently and effectively identified.