13results about How to "Accurate comparison" patented technology

The invention discloses a calligraphy stroke feature extraction method and system based on a fixed-size grid. The method relates to the technical field of stroke feature extraction, and comprises the following steps of calligraphy stroke image preprocessing and grid division, grid-based stroke space distribution feature extraction, size feature extraction, grid-based stroke direction feature extraction and feature fusion and output. The method comprises the following steps: S1, normalizing a binary image of calligraphy strokes to a fixed canvas and dividing the binary image into grids; s2, calculating the stroke density of each grid and constructing a density feature vector; S3, refining the image to obtain a skeleton, calculating a direction field and generating a direction feature vector; and finally, the two vectors are connected in series and fused, and a joint feature vector is output for a subsequent task, so that the calligraphy stroke feature extraction accuracy is improved, and the problem of low calligraphy stroke feature extraction accuracy caused by insufficient understanding of dynamic essence, complex form and high-level semantics of calligraphy art in the prior art is solved.

The application discloses a kind of fast moving face recognition method and system based on multi-frame image enhancement, comprising the following steps: through imaging device acquisition multi-frame image sequence, obtain multi-frame image after pre-processing;The multi-frame image after pre-processing is input into the improved MonST3R network, carries out spatial alignment and structure analysis, obtains the image after alignment and three-dimensional structure information;The above result is input into CRNet network, separates and enhances multi-scale feature, and obtains enhanced feature;Enhanced feature is input into space-time attention fusion module, and multi-frame feature representation is obtained by weighted fusion;Fusion feature is input into identification network, and dimension reduction is carried out, and face embedding feature vector is obtained;Similarity comparison is carried out between embedding feature and database template, and recognition result and confidence prompt are output.The application can realize efficient, accurate and resource-friendly face recognition under dynamic complex environment.

The invention discloses a substation constant value remote proofreading method based on a hybrid enhanced intelligent and elastic process. The method comprises the following steps: S1, three-stage screening and scene adaptation interrogation; s2, intelligently generating and optimizing a constant value template; s3, carrying out online closed-loop circulation and intelligent decision making; s4, performing semantic normalization on the heterogeneous parameters; s5, working condition sensing difference comparison and pseudo difference filtering are carried out; s6, performing elastic downloading execution and real-time monitoring; s7, performing feedback optimization; according to the method, a target is accurately locked through a three-stage screening structure, classification is performed through a CNN-LSTM model, an interrogation strategy is dynamically adjusted, a to-be-loaded setting value list is generated based on online closed-loop circulation and a hybrid enhanced intelligent framework, and space limitation is broken; heterogeneous parameter semantic normalization is realized through a BERT model, and comparison ambiguity is eliminated in combination with working condition sensing dynamic threshold and pseudo difference filtering; abnormality is handled based on a BPMN2.0 elastic engine, verification and feedback optimization are carried out after downloading, a complete closed loop is formed, and iteration upgrading of the system is promoted.

This invention discloses a quinoa SSR molecular marker primer set and its application in DNA fingerprinting, belonging to the field of molecular biology. This invention developed a quinoa SSR molecular marker primer set consisting of 25 pairs of SSR primers, the nucleotide sequences of which are shown in SEQ ID NO. 1-50. This invention also used this primer set to perform genetic diversity analysis and population genetic structure assessment on 170 quinoa germplasm resources, further screening out 18 core primer pairs, and using these core primers to construct a quinoa DNA fingerprint. The results show that the quinoa DNA fingerprint can effectively characterize quinoa varieties, simultaneously identify multiple quinoa varieties, and the identification method is convenient, rapid, accurate, and stable. This invention provides reliable theoretical and technical support for the precise identification and innovation of quinoa germplasm resources, the breeding of new quinoa varieties, and the protection of variety rights.

A quality monitoring device, method, and additive manufacturing equipment for laser sintering process are disclosed. The device includes: a host computer that generates scanning data from the STL file of the part to be printed and sends it to a scanning control unit; a scanning control unit that generates laser switch command signals and laser power command signals based on the scanning data and sends them to the laser, and generates motion control commands and sends them to the scanning unit; and a timing acquisition, analysis, and logic judgment unit that acquires the actual output power signal from the laser in real time, as well as the laser switch command signals and laser power command signals sent to the laser by the scanning control unit, and determines whether the laser is working normally by comparing the actual output power signal, the laser switch command signal, and the laser power command signal, and alarms when the laser is malfunctioning. This invention provides more precise monitoring, recording laser power data down to each scanning line, providing detailed data for additive manufacturing equipment.

The invention relates to the technical field of underground engineering grouting, in particular to a flowing water grouting plugging performance quantitative evaluation method based on micro-fluidic fluorescence imaging, which comprises the following steps: calibrating grout with different known concentrations by using a fluorescent tracer agent to establish a quantitative relation model; carrying out a flowing water grouting test in the micro-fluidic chip so as to capture a fluorescence image sequence in the slurry migration and plugging process; inverting the fluorescence image sequence into a slurry concentration spatio-temporal evolution field based on a quantitative relation model; identifying the slurry-flowing water interface in the spatio-temporal evolution field of the slurry concentration, and tracking the dynamic evolution process of the slurry-flowing water interface to calculate the formation time of the slurry-flowing water interface and the spatio-temporal evolution law of the slurry concentration; and calculating a plugging strength index according to the slurry concentration spatio-temporal evolution field and the slurry-flowing water interface forming time, and evaluating the plugging efficiency of the grouting material by using the plugging strength index. Therefore, the problems that in an existing flowing water grouting technology, the plugging process is invisible, and key performance is difficult to quantify are solved.

The invention provides a test system and equipment for a driving controller, which are applied to the field of electrical equipment detection and control, and the system comprises a first test module, a second test module and a control module, the first test module is connected with an external power supply and is connected with a power supply test pin and a communication test pin of the driving controller; the second test module is connected with a function test pin of the driving controller; the control module is connected with the first test module and the second test module, and is configured to obtain first electrical parameter information of the power supply test pin, communication information of the communication test pin, second electrical parameter information analyzed through the communication information and third electrical parameter information of a working mode of the driving controller; and the control module is used for performing a simulation test on the driving controller based on the acquired first, second and third electrical parameter information and the communication information. The test system provided by the invention is used for detecting the performance of the vehicle driving controller, and can realize the full-process simulation test of the driving controller.

This utility model discloses a multi-level regional monitoring device for water quality in water utilization, including a detection end and a sampling end. The detection end includes a sinking plate, a cover, a horizontal shaft, a drive motor, and a water receiving assembly on the top of the sinking plate. The sinking plate is rotatably connected to the horizontal shaft via bearings, and steel wire ropes are installed at both ends of the horizontal shaft. The sampling end includes a bottom plate, a central chamber, and a rope connecting seat. The rope connecting seat is correspondingly connected to the steel wire rope. The central chamber has a top seat on top, a floating assembly on the outside, a water inlet, and an equipment cavity at the bottom. An arc-shaped guide edge and an arc-shaped guide plate are provided on one side of the equipment cavity. The central chamber contains a pump drive source, a pump assembly, a linkage frame, and a sealing plate. The sealing plate is located between the arc-shaped guide plate and the arc-shaped guide edge. This utility model has the advantage of being able to monitor water quality at different depths within a water area, providing a more comprehensive and accurate assessment of water quality.

The invention discloses a multilayer large-scale constellation vulnerability method and system based on a global efficiency index, and mainly solves the problems that the prior art is difficult to adapt to a large-scale satellite constellation scene and cannot reflect the differentiated influence of damage on various topological connection relationships. According to the implementation scheme, satellite layered numbering is completed based on multi-layer constellation parameters, an intra-layer / inter-layer link establishment rule and a distortion factor set are synchronously initialized, and a basic constellation topology adjacency matrix is constructed; judging damage state conditions of an intra-layer link, a satellite and an inter-layer link, and respectively loading lossless and lossy topological information into a matrix; calculating the global efficiency of the corresponding scene according to the two types of matrixes, constructing a vulnerability quantitative model according to the global efficiency attenuation rate, and completing the vulnerability evaluation of the topology; and traversing topological configurations corresponding to different distortion factors, and screening out any optimal topology of which the vulnerability of other configurations is not less than the current value by comparing vulnerability numerical values. According to the method, the evaluation accuracy and adaptability of a large-scale constellation are improved, and topological vulnerability evaluation and robustness design of a multi-layer satellite constellation can be effectively supported.

The present application relates to the energy power industry equipment material supply chain data management technology, it discloses a kind of construction method of energy power industry equipment material price database, solve the problem of energy power industry equipment material price database construction, equipment material type is miscellaneous, there is no unified standard, price data source is single and credibility is low, and price adjustment lacks scientific quantization basis.The scheme in the present application, first, equipment material type parameter is standardized definition and modularization split, determine the price adjustment coefficient of basic type parameter, fixed description standard and additional function module, and scientifically set the density of basic type parameter;Second, collect historical contract / price inquiry data, regularly price inquiry data, project special price inquiry data three kinds of multi-source price data, according to time distance and credibility weight and after pre-processing, generate standard price;Finally, in combination with project region, project stage and other factors, through weight assignment and algorithm calculation, generate the price adjustment coefficient of actual price relative to standard price.

This utility model belongs to the field of ball joint wear detection technology, specifically disclosing an automatic ball joint wear detection mechanism, including: a support platform, a turntable, a standard ball joint component, and an image acquisition camera. The turntable is rotatably mounted on the support platform, the standard ball joint component is longitudinally mounted on the upper surface of the support platform, and the image acquisition camera is located on the outer side of the support platform. A clamping groove is formed on the right side of the upper surface of the turntable, and a sliding clamping block is slidably connected in the clamping groove. An adjusting screw is screwed to the right side wall of the turntable. This solution uses image acquisition to compare the acquired image with the ball joint to be detected through subsequent intelligent recognition and comparison, determining whether the ball joint to be detected has defects. Based on image acquisition and judgment, the wear condition of the ball joint to be detected can be quickly and accurately compared and obtained. Through rotation, the ball joint to be detected can be inspected from all directions.

The utility model discloses a kind of hematuria collection colorimetric devices, including box, the top of box is recessed with multiple placing grooves, test tube is equipped in placing groove, the top of one end surface of box is provided with multiple viewing ports, viewing port is respectively slidably connected with movable plate, the middle part of movable plate is provided with observation hole;The bottom of one end surface of box is fixedly connected with support slide rail, support slide rail is slidably connected with slide, the bottom one end of slide is fixedly connected with support plate, the bottom other end of slide is fixedly installed with clamping assembly. By clamping assembly to hematuria colorimetric card is clamped and fixed, slide cooperates support slide rail, the flexible movement and accurate positioning of hematuria colorimetric card can be realized, it is convenient for medical staff to adjust the position of colorimetric card according to actual demand, to compare the color of hematuria specimen and colorimetric card more accurately;This design effectively avoids the inconvenience of observation or comparison error possibly caused by the fixed position of colorimetric card, improves the accuracy of hematuria degree judgment.

The utility model relates to chain detection equipment capable of automatically checking defects, which comprises a working table, a camera shooting assembly is arranged on the top surface of the working table, two supports which are opposite left and right are arranged on the top surface of the working table, rotating frames are arranged on the opposite sides of the two supports, and the rotating frames are rotatably connected with the supports through transverse first rotating shafts. First motors for driving the rotating frames to rotate are arranged on the opposite sides of the two supports, chain wheels are arranged on the opposite sides of the two rotating frames, the horizontal heights of the two chain wheels are the same, the chain wheels are rotationally connected with the rotating frames through vertical second rotating shafts, and a second motor for driving the corresponding chain wheel to rotate is arranged on one rotating frame. The vertical projection of a camera of the camera shooting assembly is located between the two chain wheels. According to the utility model, the first motor drives the rotating frame to rotate, so that the chain can be turned over, the other chain plate group corresponds to the camera to shoot, the chain does not need to be turned over manually, the use is more convenient, and the detection efficiency is greatly improved.