33results about How to "Not easy to be disturbed by noise" patented technology

The invention discloses a different-source image matching method based on refined feature optimization extraction, which mainly solves the problem of low matching precision in the prior art, and adopts the technical scheme that 1, gradient moduli and directions of an optical image and an SAR image are calculated by adopting different gradient operators respectively; 2, completing region-of-interest calibration by utilizing the specific property of the target in the image; 3, according to the result of the step 1, obtaining edge features of the interested areas of the two images, and obtaininga coarse matching affine transformation matrix between the two images through template matching; 4, obtaining a transformed optical image by using the coarse matching affine transformation matrix, andthen extracting refined features; 5, performing preliminary matching on the refined features, and eliminating singular point interference to obtain a fine matching affine transformation matrix; and 6, obtaining a final affine transformation relation matrix between the optical image and the SAR image according to the results of the steps 3 and 5. The method can achieve the precise registration ofa heterologous image, and can be used for the guidance of an aircraft.

The invention provides a capacitive stylus used for operations on a capacitive touch panel (200). The capacitive stylus comprises a transmitting unit (1) and a capacitive stylus control processing unit (2). The transmitting unit is used for transmitting high-voltage alternating current signals to the capacitive touch panel (200). The capacitive stylus control processing unit is used for generating high-voltage alternating current signals; the signals are transmitted to the capacitive panel (200) by the transmitting unit (1), so that the capacitive touch panel (200) can measure location of the capacitive stylus (100) according to the high-voltage alternating current signals. The capacitive stylus has the advantages that the high-voltage alternating current signals are transmitted to the capacitive touch panel to allow the capacitive touch panel to locate according to the high-voltage alternating current signals; the strength of the high-voltage alternating current signals is high, so that signal to noise ratio and locating precision can be increased.

The present invention provides a fingerprint light current detection unit, a fingerprint recognizer, a driving method and a display device. The fingerprint light current detection unit includes a conversion circuit and a detection circuit; the conversion circuit is connected with a fingerprint light current reading line and is used for converting the fingerprint light current of the fingerprint light current reading line into corresponding square wave signals; and the detection circuit is connected with the conversion circuit and is used for detecting the square wave signals and obtaining fingerprint light current information based on the frequency of the square wave signals. According to the fingerprint light current detection unit, the fingerprint recognizer, the driving method and the display device of the invention, the fingerprint light current of the fingerprint light current reading line is converted into corresponding square wave signals through the conversion circuit, namely, an electrical physical quantity-to-frequency detection conversion process of fingerprint light current detection can be realized; and a frequency detection mode is adopted, and therefore, signal strength is high, and the fingerprint light current detection unit has little possibility of being interfered by noises.

The invention discloses a mount element detection method based on color image segmentation and gradient projection positioning, and belongs to the technical field of automatic optical detection of mount elements. The mount element detection method based on color image segmentation and gradient projection positioning is characterized in that a designed image acquisition system is used to acquire color images; three-color (red, green and blue) images are converted into an HSI color model which is formed by three parameters (H (Hue), S (saturation) and I (Intensity)); as the HSI model directly uses the hue and the saturation having no relation with brightness when the HSI color model extracts the color information, the mount element detection method based on color image segmentation and gradient projection positioning is accurate and highly efficient in segmentation; one the above basis, positioning and detection of elements are carried out; positioning mainly uses the color and geometrical characteristics, and a positioning method based on gradient projection and a principal component characteristic value detection method are proposed; and the mount element detection method based oncolor image segmentation and gradient projection positioning has the advantages of high positioning accuracy and efficiency and accurate detection.

The invention discloses a detecting device for output current of a voltage regulator. The voltage regulator comprises an upper bridge assembly, a lower bride assembly and an electrical inductance which are jointly coupled to a switching joint, and the other ends of the upper bridge assembly, the lower bride assembly and the electrical inductance are respectively coupled with an input end, a grounding electric potential and an output end; the voltage regulator also comprises a controller at least controlling the movement of either the upper bridge assembly or the lower bride assembly so as to generate output voltage at the output end. The detecting device for output current comprises an upper bridge current sensing circuit, a lower bridge current sensing circuit and a setting resistance; the detecting device of output current is characterized in that the upper bridge current sensing circuit senses the current of an upper bridge assembly to generate a first current; the lower bridge current sensing circuit senses the current of an lower bridge assembly to generate a second current; a totalling circuit adds the first current and the second current to generate a summation current; thesetting resistance detects the change of the summation current, and resistance change on the setting resistance is provided for the controller to carry out feedback control.

The invention puts forward a lithium battery capacity and service life prediction method based on a generalized degradation model and multiscale analysis. The method comprises the following steps of: firstly, carrying out phase space reconstruction on historical capacity degradation data, utilizing reconstructed data to fit the generalized degradation model, obtaining a model parameter, and fitting a root-mean-square error; secondly, in multiple scales, carrying out forward estimation on a future capacity, and obtaining a forward prediction result through weighted integration; thirdly, adding a prediction result into a known sequence, and carrying out forward iterated prediction; and finally, judging whether a multi-step prediction result achieves a failure threshold value or not, and calculating remaining service life. By use of the generalized degradation model which is put forward by the invention, a complex combined degradation rule can be fit, and the model has the advantages of high universality, wide applicable range and small fitting error. On the basis of phase space reconstruction, multiscale prediction is carried out, a prediction result is obtained through the weighted integration, and the prediction result is unlikely to interfere by noise and is high in accuracy.

The invention provides a facial expression analysis method based on a capsule network, and the method employs a multi-channel parallel technology, a Strict-Squash function, a Lane-Filter function, and DropCircuit technology to improve an original capsule network, avoids noise interference caused by the irrelevant features of facial expressions, reduces the workload of manual marking, and improves the accuracy of facial expression analysis. Therefore, sentiment analysis can be carried out on a face image with a complex background, and the expression recognition precision is improved. The method is innovated on the basis of the capsule network, can effectively learn the spatial relationship between objects, has a strong recognition capability for slightly deformed objects and pictures of the same object at different visual angles, is not liable to be interfered by noise caused by irrelevant features of facial expressions, and has a good recognition effect on the objects at different visual angles. The method can carry out correct sentiment analysis on a face image with a complex background, is high in expression recognition precision, is wide in application range, and is high in robustness.

The invention discloses a monophonic speaker separation model, a training method and a separation method. The monophonic speaker separation method comprises the following steps: acquiring audio containing a first speaker and a second speaker; segmenting the audio to obtain at least one segmented audio; inputting at least one part of segmented audio to the monophonic speaker separation model to obtain at least one part of first embedding corresponding to the at least one part of segmented audio; inputting a pre-recorded audio only comprising the second speaker to the monophonic speaker separation model, and obtaining a second embedding corresponding to the audio only comprising the second speaker; and judging whether the cosine similarity of the at least one part of first embedding and thesecond embedding is less than a preset threshold, if so, determining that the at least one part of segmented audio is a first speaker audio, and if not, determining that the at least one part of segmented audio is a second speaker audio.

The invention relates to a synchronous phasor calculation method and system for filtering noise. The method comprises the following steps: sampling a signal of a power system; windowing the sampling signal by adopting a two-week Hamming window; calculating a rotating phasor of the windowed signal relative to a rated frequency rotating coordinate system; constructing a unary quadratic equation by utilizing the three equally-spaced rotating phasors, and solving and obtaining a phasor z; obtaining a phase and a frequency of the signal and a conjugate phasor of the phasor z based on the phasor z;and correcting the rotating phasor based on the phase, the frequency and the conjugate phasor to obtain a synchronous phasor of the signal. Dynamic behaviors of a power system are considered, and themethod and system are not easily interfered by noise such as frequency offset and harmonic waves; the influence of noise on synchronous phasor measurement is effectively reduced; when the frequency ofthe system deviates from the rated frequency, the influence of spectrum leakage caused by fixed-interval sampling is reduced, and phasor and frequency measurement has high precision.

The invention discloses a permanent magnet synchronous motor demagnetization fault diagnosis method and system based on a semi-supervised classifier. The method comprises the steps of collecting a magnetic flux leakage signal on the surface of a permanent magnet synchronous motor; performing symmetric dot matrix image generation on the magnetic flux leakage signal of the one-dimensional time domain to obtain a two-dimensional symmetric dot matrix image; inputting the symmetric dot matrix image into a wavelet scattering convolutional network for feature extraction to obtain a feature vector; and inputting the feature vector into a pre-trained semi-supervised depth rule classification diagnostor to obtain a demagnetization fault diagnosis result. According to the method, a one-dimensional time domain signal is expanded to a two-dimensional symmetric dot matrix image, complex signal processing is avoided, accurate fault diagnosis under small sample data can be realized based on the semi-supervised depth rule classifier, and the problem that actual equipment cannot be popularized and applied due to small data volume in fault diagnosis is solved.

The invention discloses a communication device which comprises a circuit board, a first metal shielding structure, a radio frequency circuit and a fundamental frequency circuit, wherein the circuit board comprises a circuit layer, an insulation layer and a grounding layer; the insulation layer is arranged below the circuit layer, the grounding layer is arranged below the insulation layer, and the grounding layer upwards passes through the insulation layer and is provided with an isolated body in an extending way so as to isolate a first circuit area from a second circuit area; the first circuit area and the grounding end of the second circuit area are electrically connected with the grounding layer, and a gap is arranged between the first circuit area and the isolated body; and the radio frequency circuit is arranged on the first circuit area, the foundational frequency circuit is arranged on the second circuit area, and the first metal shielding structure is arranged above the circuit board structure and electrically connected with the isolated body so that the first circuit area and the second circuit area are not easy to interfer by respective noise.

The invention discloses a rapid electrical fireproof current-limiting protection device which comprises a zero-cross detection module, a change rate normalization module, a microprocessor, a current sensor, a main control switch and a load circuit. According to the device, the load current is normalized, so that the value of the relative change rate drt of the real-time normalized current on which the microprocessor performs short-circuit judgment is always kept between 0 and 1, and the threshold table is the same no matter what value of the rated current of the load is, namely, when the rated current of the load is the value of 10A or 32A or 63A and the like, the value of the relative change rate drti of the standard normalized current is the same, and the value of the relative change rate drti of the standard normalized current is the same. Different current-limiting protection devices do not need to be replaced for different loads, and the device can adapt to most loads with rated current, especially load circuits below 64A.

The invention discloses a video target tracking method based on compressed regularization block difference. The method is realized by the steps that 1) a tracking area is selected; 2) a measurement matrix is initialized; 3) a target classifier is initialized; 4) a target classifier is updated; 5) a new video frame is input; 6) a rough candidate target set is generated; 7) CNBD characteristic vectors of all candidate targets in the rough candidate target set are calculated; 8) a rough tracking result is discriminated; 9) a fine candidate target set is generated; 10) CNBD characteristic vectorsof all candidate targets in the fine candidate target set are calculated; 11) a present-frame tracking result is discriminated; and 12) if the present frame is the last frame, tracking is completed, and otherwise, the step 4) is turned to. According to the method, differential characteristics of a compressed regularization block is used to describe the tracked or candidate targets, and the candidate target set is generated in a rough to fine sliding window manner.

The invention provides an acupuncture manipulation quantification method and system based on machine vision. The method comprises the following six modules: calibrating a binocular vision sensor, recording a process of an experimental needle, calculating space coordinates, analyzing corresponding manipulation operation, calculating physical parameters of the manipulation operation, and storing a result in a manipulation database. The binocular vision sensor is calibrated so as to obtain the conversion relation between image coordinates and space coordinates, the computer calculates the space coordinates of each frame of reflective target in the video, and all the space coordinates are analyzed so as to obtain manipulation operations in different time periods in the video. Physical parameters such as the amplitude and the angle of each manipulation operation are calculated and stored in a manipulation database, so that the quantification of the acupuncture manipulation is completed once.