72results about How to "Improve speed and precision" patented technology

An object of this invention is to increase the precision and speed of autofocusing operation. To achieve this object, an image sensing apparatus includes a photographing optical system which forms an object image, the photographing optical system having a focusing optical system which adjusts the imaging position of the object image, an image sensing element which photoelectrically converts the object image formed by the photographing optical system to generate an electrical image signal, an in-focus position detection device which detects the in-focus position of the focusing optical system for forming the object image on the image sensing element, on the basis of the image signal generated by the image sensing element while the focusing optical system adjusts the imaging position of the object image, and a distance measurement device which has a plurality of light-receiving portions for receiving a beam from an object, and measures a distance to the object on the basis of output signals from the light-receiving portions.

The invention relates to a tracking method for the maximum power point, applied in an independent photovoltaic grid-connected power generation system. According to the tracking method, the maximum power point of a solar cell panel is tracked in a variable step size mode on the basis of the conductance increment method, so that the problem that speed and accuracy are ambivalent in a traditional conductance increment method is solved. According to the tracking method, the step size of a work point can be achieved in an automatic regulating mode according to characteristics of the solar cell panel, and therefore when the work point is far away from the maximum power point, the step size is larger, and tracking speed is increased; when the work point is close to the maximum power point, the step size is reduced so that the oscillation phenomenon nearby the maximum power point can be reduced, and therefore efficiency is improved. The step size correcting MPPT algorithm provided by the tracking method can solve the conflicting problem, and accuracy and speed of MPPT can be improved easily and effectively. The tracking method has the advantages of being accurate in control and high in response speed, and can be suitable for rapid reaction occasions where sunshine and temperature change suddenly, and the situation of continuous changes of the sunshine can be stably tracked.

A weapon sight of the kind comprising a lens having a partially reflecting surface, a light source spaced from the lens for emitting light towards said reflecting surface to produce a light spot by direct imaging of said light source on said surface to be superimposed on a target when sighting through the lens. For adjustment of the location of the light spot on the reflecting surface when observed from a defined fixed position, the sight includes a device for displacement of the light source, comprising a miniature drive motor or element operatively connected with the light source.

The invention discloses a multi-basin real-time intelligent water quality predication method and system. According to the multi-basin real-time intelligent water quality predication method and system, an NARX model is optimized through a genetic algorithm and a problem that earlier stage parameters of the NARX model are uncertain; a GA-NARX model is stored or called through a model storage module to predicate the multi-basin water quality condition in real time; essential data, recent data and history emergency data are performed on three-section training and accordingly an optimized GA-NARX model can basically comprise river pollution characteristics and the predication accuracy is improved; weather data replaces hydrological data and is performed on obfuscation processing and accordingly influences to the model from flow data missing are effectively solved; matching of similar pollution time sequence templates is performed through an improved DTW algorithm, a similar pollution process in the basin history is rapidly found out, experience is referred and learned, and sudden emergency accidents are accurately predicted. The multi-basin real-time intelligent water quality predication method and system can be widely applied to the water quality predication field.

The invention provides a method for processing DICOM (digital imaging and communications in medicine) medical images on the basis of formats. The method includes inputting a to-be-registered DICOM reference image and a to-be-registered test image, and converting the formats of the two images; extracting feature information from the images to form feature spaces; carrying out iterative optimization transformation in the feature spaces until requirements of preset threshold values of the similarity between the DICOM reference image and the test image are met by the transformation. The method has the advantages that the method is an improved method for registering the images, accordingly, the DICOM medical image registration speeds can be effectively increased, and the DICOM medical image registration precision can be effectively improved.

The invention provides a smart agriculture monitoring and management system, which comprises a cloud computing monitoring platform, a portable monitoring terminal, a pre-posed access unit, an image obtaining unit, a plurality of aggregation nodes, a plurality of automatic control nodes and a plurality of sensor nodes. The cloud computing monitoring platform has a function of multi-parameter combined treatment, and intelligent control speed and precision are improved; and besides, through the arrangement of the pre-posed access unit, the system not only realizes a conventional data forwarding function, but also carries out pretreatment on the received data, thereby reducing later data analysis complexity and delay and greatly reducing energy consumption of the whole sensor network.

The invention relates to a method and a device for identifying a ship brand character in a natural scene based on deep learning. The method comprises the steps of acquiring a video of incoming and outgoing ships at a port; decomposing the video into video streams composed of ship image frames, and classifying the video streams by use of a trained ship classification model to acquire a ship image in a specified type; carrying out ship brand detection on the ship image in the specified type by use of a trained ship brand detection module to acquire a ship brand detection result; and carrying out character identification on the ship brand detection result to obtain the ship brand character in the natural scene.

The invention relates to a novel method for identifying overvoltage inside a distribution network. An atomic decomposition algorithm optimized based on an imperial colonial competition algorithm is applied to the analysis of over-voltage signals, and the optimized algorithm can be used to extract the internal characteristics of signals quickly and effectively without being disturbed by noise. A three-phase characteristic atomic spectrum is constructed based on the atomic decomposition algorithm, waveform signals only containing time-domain information are converted into high-dimensional characteristics containing time-frequency information, and the time-frequency characteristics of overvoltage signals are described completely. The three-phase characteristic atomic spectrum is recognized directly by using a convolution neural network, the shortcomings of a shallow learning identification algorithm for the identification of high-dimensional characteristics are overcome, and the subjectivity and complexity in the process of characteristic reduction are avoided. Simulation waveforms and physical models verify that the distribution network overvoltage type identification method of the invention has high identification accuracy and strong adaptability.

The invention discloses a three-dimensional target detection method. The method comprises the following steps: extracting a candidate bounding box of a target on an original image by using a two-dimensional target detection algorithm; converting a depth image region corresponding to the candidate bounding box into a view cone point cloud; carrying out instance segmentation on the view frustum point cloud to obtain interested target point cloud; and regressing the three-dimensional bounding box of the target through the neural network. According to the method, a boundary frame is extracted through the two-dimensional target detection algorithm and then returned to the three-dimensional boundary, the target detection speed and precision can be improved, and the characteristics of the two-dimensional information and the characteristics of the three-dimensional information can be used for detecting the shielded target in the detection process.

A method for generating a land vegetation covering model in a three-dimensional visualized environment is characterized by comprising the following steps: acquiring remote sensing image data of a research region; preprocessing a remote sensing image; merging pixels with the similarity property to form a complete region, and dividing the remote sensing image; automatically extracting land vegetation covering information by a computer to generate a land vegetation covering graph; and carrying out fusion processing on digital elevation model (DEM) data and land vegetation covering graph data in the research region to generate the land vegetation covering model in the three-dimensional visualized environment. Advantages of an object-oriented classification technology and a technology of recognition by using vegetation indexes are integrated, the shortcoming that 'foreign matter co-spectrum' land covering types cannot be distinguished by the object-oriented classification technology independently is overcome, and meanwhile, the problems that classification results are finely broken and do not have clear geographic significance when spectral characteristics are used independently are solved. In addition, unique vegetation covering indexes are adopted, so that precision and speed of recognition of remote image vegetation are high.

The invention discloses a face attitude detection method. The face attitude detection method comprises the steps of extracting N face feature points from a face image according to a face detection algorithm; extracting key feature points from the N face feature points; and calculating face attitude information such as the rotating direction and angle of a face around a coordinate axis according tocoordinate values of the key feature points. The attitude information of the face in the face image is calculated according to the coordinate values of the face feature points, and real-time detection for the face attitude is achieved. Correspondingly, the invention further provides a calculating device and a computer readable storage medium.

The invention relates to a wide-dynamic-range device for measuring the extinction ratio of a star sensor light hood, relating to a technology for measuring the extinction ratio of the star sensor light hood, and used for solving the problems of being low in measurement precision and measurement speed in the existing method for measuring the extinction ratio of the light hood. A radiation emitting and shaping sub-system provided by the invention is used for providing a light source for measurement of the extinction ratio; a radiation data acquiring and processing sub-system comprises an integral sphere and a computer; the integral sphere is used for measuring energy at the inlet and the outlet of the light hood and sending a measurement result to the computer; a mechanical motion and control sub-system is used for bearing and driving the light hood, a light hood support structure and the integral sphere to rotate together; and an environmental parasitic light absorbing and inhibiting sub-system is a darkroom and used for absorbing background parasitic light. According to the invention, measurement with the illumination of 10 to 3-1051x can be realized; the integral sphere is used as the receiving device, so that the measurement error is less than 2%; automatic measurement is realized through software integration; the measurement speed is increased; and thus, the wide-dynamic-range device disclosed by the invention is applied to measuring the extinction ratio of the star sensor light hood.

The invention discloses a focus processing method and device, and a terminal device. The focus processing method is applied to the terminal device with a master camera module and a secondary camera module. Each light filter unit in the secondary camera module corresponds to a combination pixel in a secondary image sensor. Each combination pixel comprises at least n adjacent pixels. The n is an even number greater than 2. The method comprises the steps of obtaining a first image collected by left n/2 pixels and a second image collected by right n/2 pixels of the same combination pixel in the secondary camera module; judging whether phase difference between the first image and the second image is in a preset range or not; and if the phase difference is not in the preset range, controlling a master motor in the master camera module to drive a master camera to move according to a preset strategy and the phase difference. The focus precision and speed are improved, the rapid focus cost is reduced, and the quality of the image photographed by the terminal device is improved.

The invention discloses a method for determining a structure tolerance of a sparsely-configured antenna array based on a radiation field and a scattering field. The method comprises the steps that a structure, electromagnetic working parameters and a sparsely-configured matrix of the sparsely-configured antenna array are determined, and an initial structure tolerance is provided; the structure tolerance is allocated as the position installation precision of array elements in the plane where a sparse array surface is located and the planeness of a normal array surface of the sparse array surface where the array elements are located; a random number of array element position installation errors and a random number of array surface planeness errors are acquired; a radiation field space phase difference of every two adjacent radiation units in the sparse array at a target is calculated to obtain an antenna radiation field oral surface phase difference; an antenna radiation field direction diagram is calculated; the deterioration degree of a phase antenna gain relative to design indexes is calculated; whether the design requirements are met or not is judged; an antenna scattering field oral surface phase difference and a scattering field direction diagram are calculated, and electrical property parameters of the scattering field are calculated; whether the design requirements are met or not is judged. According to the method, the processing and mounting difficulty can be reduced, the development cost can be reduced, and the development cycle can be shortened.

The embodiment of the invention provides an object selection method and system which help a user determine overall or local selection instructions of objects in overlapped objects as overlapped object groups to make selection instructions. Therefore, the user can select overlapped object groups with ease. Meanwhile, the object selection method and system are capable of displaying thumbnails of all objects in overlapped object groups to bring convenience for the user to make selections after the user chooses overlapped object groups. The object selection method and system have following beneficial effects: the user can indirectly select corresponding objects by selecting thumbnails so that the user does not need to select objects by dragging; rapid selections of objects can be made so that precision and speed of object selections are improved; the user feels comfortable with the object selection method and system; and meanwhile, the object selection method and system do not cause changes of all object positions so that display effect of all objects is ensured.

The invention relates to an image feature storing method, an image searching method and device based on compressive sensing. The image feature storing method includes the steps that an image to be processed is segmented into a plurality of subblocks; a plurality of image features of the subblocks are combined to form original signals of the images; the original signals are sequentially changed sparsely and projected to the same measurement matrix to obtain subblock measured values corresponding to the original signals. The image features are stored by using a subblock compressive sensing measuring method, needed storage quantity is small, and the storage can be achieved rapidly. On the other hand, difference between the image measured values and measured values of images to be searched for is used for sparsity judgement, so that people do not need to recover the original signals accurately and only need to extimate the sparsity of the difference, and image searching speed and accuracy are effectively improved.

The invention discloses an electric energy quality monitoring system for a high-voltage charging pile. The electric energy quality monitoring system includes an electric energy quality disturbance signal collection module, a denoising module, a characteristic extraction module, an electric energy quality disturbance type identification module, and an electric energy quality evaluation module, wherein the electric energy quality disturbance signal collection module is used for implementing collection of an electric energy quality disturbance signal in a target segment; the denoising module performs denoising processing on the electric energy quality disturbance signal based on a generalized morphological filter of an adaptive LMS algorithm of a mathematical morphological structure; the characteristic extraction module is used for extracting an energy distribution at different frequency bands of the electric energy quality disturbance signal and a fractal dimension of the electric energyquality disturbance signal; the electric energy quality disturbance type identification module implements evaluation of a quality disturbance type based on an online training support vector machine;and the electric energy evaluation module is used for completing a statistic of all present electric energy quality disturbance type parameters, and outputting a corresponding electric energy qualityevaluation result based on a statistical result. The electric energy quality monitoring system provided by the invention may implement effective monitoring on electric energy quality of the high-voltage charging pile, and may implement automatic optimized control on the electric energy quality of the high-voltage charging pile.

The invention discloses a vacuum electron beam welding method of a titanium steel composite billet using an edging assembled billet. According to the method, 1) by adopting the scheme that an edging design is used in the field of vacuum electron beam welding and edgings are sealed and welded in advance, the technical problem that a vacuum electron beam cannot be used for welding vertical seams issolved, and the accuracy and speed of processing assembled billets are improved; 2) for the first time, it is proposed that a titanium plate is sealed and welded in advance, so that the hidden dangerthat a parting agent may scatter and pollute a composite layer during rolling is solved; and the consumption of steel materials and production costs are greatly reduced by replacing a traditional catching groove process with the edging assembled billet process provided by the invention, flaw detection results of a slab after rolling are all qualified, and the performance is much higher than the user requirements, therefore, a strong technical support is provided for expanding the production of high-quality products.

The invention relates to an AR-based PCB assembling method and an auxiliary assembling device. The method comprises the steps of obtaining a first coordinate position of a user sight stop position on a PCB; comparing the first coordinate position with a PCB configuration effect electronic image input in advance, and determining whether the first coordinate position is a to-be-welded position or not; if the first coordinate position is the to-be-welded position, determining device information of the to-be-welded device at the position, and overlaying the device information at the position, so that a user conducts the assembling operation at the position according to the device information, and the PCB assembling speed and precision are improved.

The embodiment of the invention provides a simultaneous positioning mapping method and device. The method comprises: carrying out feature point extraction, matching and triangularization on a binocular image sequence, and obtaining three-dimensional road sign point cloud of each frame of binocular image; obtaining an affinity matrix according to the three-dimensional road sign point cloud of eachframe of binocular image; performing hierarchical clustering according to the affinity matrix, and determining a rigid body to which each road sign belongs; performing simultaneous positioning mappingaccording to the three-dimensional road sign point cloud of each frame of binocular image and the road sign included in each rigid body, wherein the affinity is the logarithm probability that the tworoad signs belong to the same rigid body. According to the simultaneous positioning mapping method and device provided by the embodiment of the invention, hierarchical clustering is carried out according to the affinity, the moving object in the scene can be detected more accurately, modeling and pose estimation are carried out on the motion, and the calculated amount is smaller, so that the precision and speed of simultaneous localization mapping can be improved.

The invention discloses an acquisition method and detection method of living body face detection head posture regression device, and the method comprises the steps: obtaining face data, and forming face three-dimensional model data with composite standardization; obtaining a training sample set; acquiring a corresponding Euler angle and a rotation matrix containing a corresponding relationship between each two-dimensional coordinate plane and the corresponding Euler angle by using the acquired two-dimensional coordinate plane set formed by the two-dimensional coordinates of the feature points;obtaining a head posture regression device. According to the invention, a high-precision head posture regression device is constructed. Corresponding earlier-stage processing is carried out on the human face two-dimensional model data, the human face three-dimensional model data and the iris data. The invention has the following beneficial effects: human faces are effectively recognized, therebymaking data more accurate and robust; and the success rate of face living body detection is improved.

The invention provides a double-impeller horizontal shaft wind turbine design optimization method. The method comprises the following steps: establishing a design optimization model of a DRWT; solvingthe design optimization model of the DRWT; predicting the aerodynamic performance of the DRWT in the process of solving the design optimization model of the DRWT by coupling AL theory and CFD. With the DRWT pitch angle combination, the interval between two stages of impellers, the size of the two stages of impellers and the rotation speed of the two stages of impellers as design variables, the method optimizes the design target with the maximum output power of the DRWT as the optimization design target, and the tower load as the constraint, the AL theory and CFD coupling numerical simulationmethod is proposed for the DRWT model to solve the detailed flow field between the two stage of impellers, so that it is convenient to optimize the design variables; and the method can forecast and evaluate the aerodynamic performance of a double-impeller wind turbine, and can improve the accuracy and velocity of the DRWT aerodynamic performance prediction substantially, thus improving the DRWT design efficiency and laying a foundation for the DRWT engineering design and engineering application.

The invention discloses a multi-level threshold image segmentation method, which comprises the following steps of: firstly, initializing a grey wolf group; calculating the fitness of each grey wolf individual, and determining the positions of an alpha wolf, a beta wolf and a delta wolf; and updating the positions of the alpha wolf, the beta wolf and the delta wolf according to the distances between the remaining grey wolf individuals and the alpha wolf, the beta wolf and the delta wolf, updating the positions of the remaining grey wolf individuals through a particle swarm algorithm, and finally obtaining an optimal threshold through multiple iterations to perform image segmentation.

The invention discloses a floods identification method for different drainage basins. The method comprises the following steps: collecting flow sequence data and preprocessing the flow sequence data; initializing the preprocessed data to generate a mean value and a variance; performing high and low point extraction on the initialized data; updating an and sigma n based on the current traffic sequence data; performing staggered verification on the obtained 'high' point and 'low' point sequence, if at least one 'high' point is contained between every two adjacent 'low' points, generating a wave crest, otherwise, abandoning the 'low' point; finally discriminating the generated wave crest sequence, and determining the rising point and the water recession point of the flood in the session according to the wave crest; the problems that in the prior art, during session flood recognition, auxiliary intervention of more human experience knowledge is needed for peak extraction, or the working mechanism of the method excessively depends on preset fixed parameters, the effectiveness is poor, and the like are solved.

The invention discloses a safety protection method for the personnel around a foundation pit. The method comprises steps that S1, a safety warning area is set at the area around the foundation pit; S2, cameras are placed around the foundation pit, and the cameras are guaranteed to capture photos including both the foundation pit and the alert area; S3, all the cameras mounted around the foundationpit are connected with a remote control center, and the monitoring area is adjusted by the remote control center according to the surrounding conditions of the foundation pit, and the foundation pitarea is delimited; and S4, the photos captured by the cameras are utilized by the remote control center, through the pedestrian detection algorithm, the personnel around the foundation pit are detected, the relative position between the detected surrounding personnel and the safety warning area is further obtained, whether issuing a warning message is needed is determined, for the pedestrian detection algorithm, the probability that a small quantity of sample classes is acquired through model network calculation is enhanced. The method is advantaged in that through the principle, accuracy andthe speed of detection on the personnel around the foundation pit are improved, the personnel around the foundation pit is rapidly and timely prompted, and safety of the personnel around the foundation pit is guaranteed with high reliability.

The invention discloses a computer numerical control constant force spring machine which is characterized by comprising a base, a forming mechanism, a feeding mechanism, a blanking mechanism and a PC control system, wherein the forming mechanism is used for carrying out stretch bending forming on the belting, and installed on the base; the feeding mechanism is used for conveying the belting and installed at the output end of the forming mechanism; the blanking mechanism is used for machining the butt end of the spring and installed at the output end of the feeding mechanism; and the PC control system is used for controlling the operation of the whole machine. The PC platform of the PC control system is a human-machine interaction interface, the program is compiled according to the product requirements, and finally, the signal is output by the control card to a servo driver; the servo driver controls an AC servomotor; and the AC servomotor drives the feeding mechanism and the forming mechanism, thereby achieving the goals of numerical control, automatic on-line size adjustment and continuous production of constant force springs. The invention can obviously enhance the production speed and precision of the constant force springs, reduce the labor intensity of the workers and enhance the production efficiency of the constant force springs.

The invention relates to the technical field of lithium batteries, in particular to a lithium ion battery thermal runaway early warning method based on deep learning. The method comprises the steps of obtaining an original data set of a lithium battery, obtaining fusion features of external environment data and text data, extracting and selecting related health indexes, obtaining time feature vectors of key features, obtaining input of a graph convolutional network, and obtaining spatial features of battery aging data. According to the method, a lithium ion battery thermal runaway early warning model is established, and battery safety early warning evaluation is performed, so that high-precision battery thermal runaway early warning is realized, the speed and precision of lithium ion battery thermal runaway early warning are improved, the intelligent progress of new energy storage safety is promoted, and the application of the fourth industrial revolution in energy and energy storage industries is accelerated.

The invention, which belongs to the technical field of uranium isotope centrifugation separation, particularly relates to an output signal processing device of a uranium enrichment gas centrifuge rotating-speed sensor. The output signal processing device comprises a channel switch, a filter circuit, a shaping circuit and an output circuit. An original rotating speed signal is gated by the channelswitch to carry out time division processing; the filter circuit carries out filtering to remove high and low frequency clutter interferences; the shaping circuit carries out amplification and shapingand an effective signal is transformed into a standard square wave signal accurately; the output circuit carries out inverse transformation to eliminate accompanied clutters, so that the square wavequality is improved and an ideal square wave signal is outputted. According to the invention, the same kind of device that is introduced from Russia and includes discrete components is replaced completely; the accuracy and speed of the centrifugal machine rotating speed and friction power consumption measurement can be improved obviously; the accuracy and reliability of the detection are improved;the output signal processing device has the great economic and social benefits; and the nuclear fuel production line construction projects continue to increase with continuous expanding of the nuclear fuel centrifugal separation production scale, so that the massive economic benefits are created.