47results about How to "Fast and accurate assessment" patented technology

The invention provides a nondestructive detection system of sheet metal by ultrasonic planar guided-wave. An ultrasonic transceiver of the system is connected with an encoder and a sensor. The encoder and the sensor are connected with a controller. The sensor is coupled to a steel plate to be detected through machine oil used as a couplant. Electrical signal transmission process of the system is as follows: a computer inputs excitation parameters to control the ultrasonic transceiver to apply an excitation signal onto the sensor while positional information acquired by the encoder and an echo signal acquired by the sensor are transmitted to the computer through the ultrasonic transceiver. Meanwhile, the invention provides a nondestructive detection system of sheet metal by ultrasonic planar guided-wave and a detection method thereof. The invention has the following beneficial effect: by the characteristic that propagation distance of ultrasonic Lamb wave is long, detection range is large and large-area long-distance detection of a plate can be realized only from one detecting point, continuous acquisition of test data can be carried out, and fast effective large-area detection of a sheet metal structure can be achieved. The ultrasonic Lamb wave detection process is flexible and convenient. Quantitative assessment is achieved; detection efficiency is high; and it is easy to achieve automation.

The invention provides a house type assessment method and apparatus. The house type assessment method comprises the steps of receiving a target house type image; identifying house characteristics in the target house type image to obtain identification results of the house characteristics of the target house type image; obtaining scores corresponding to the identification results of the house characteristics in a preset characteristic score table, wherein the characteristic score table stores one-to-one correspondence relationships between the identification results of all types of the house characteristics and different scores; and, according to the identification results of the house characteristics and the scores corresponding to the identification results of the house characteristics, generating a house type assessment file for the target house type image. The advantages and disadvantages of a house type can be automatically extracted; the advance process is quick and reliable; andthe house type assessment accuracy is effectively ensured.

The present invention relates generally to methods and materials pertaining to assays, for example immunoassays, for biomarkers in body fluids e.g. blood. The invention also relates to diagnostic or screening methods for infections, and methods of differentiating between infectious and non-infectious conditions in mammals, particularly equines, for monitoring response to anti-infective/antibiotic therapy. The invention further relates to a test fluid collection system adapted to permit dilution and analysis of the collected test fluid. The invention further relates to monitoring exertional rhabdomyolysis in equines, and assay devices for all these things.

The invention discloses a method for estimating the overall load response of long-span bridges under the action of stochastic traffic flow, which comprises the following steps: S1, collecting the traffic flow Q of a road surface and the proportion P of a freight car to obtain the traffic speed V; S2, collecting the traffic flow Q of the road surface and the proportion P of the freight car; S2, calculating the traffic density K according to the collected traffic flow Q, the freight car proportion P and the traffic speed V according to the function relationship Q=f (K) between the traffic flow Qand the traffic density K determined by the flow-density-speed model; 3, calculating the car density K1 and the freight car density K2 according to the traffic density K and the freight car proportion P; S4, calculating the car load concentration degree q1 according to the car density K1; S5, calculating the load concentration degree q2 of the freight car according to the freight car density K2;S6, calculating the uniformly distributed load set q of the bridge accord to the load set degree q1 of the car and the load set degree q2 of the freight car. Comparing the deck uniform load concentration q with the design load concentration qk of the vehicle load model, this method provides a method to estimate the overall load response of long-span bridges quickly by traffic flow macroscopic parameters.

The invention relates to an assessment method and system for electric transmission and transformation equipment states. The assessment method comprises the steps of collecting site data of N equipment state characteristics of the electric transmission and transformation equipment; separately calculating the cumulative probability of the N equipment state characteristics according to the site data and a corresponding cumulative distribution function obtained by pre-modeling; calculating a score-reducing value of the N equipment state characteristics according to the cumulative probability and a characteristic degradation score-reducing model corresponding to the cumulative distribution function; separately setting weights for the N equipment state characteristics; and obtaining a total score-reducing value of the electric transmission and transformation equipment according to the score-reducing value and the weights of the N equipment state characteristics, and assessing the states of the electric transmission and transformation equipment according to the total score-reducing value of the electric transmission and transformation equipment. The assessment method is simple, free of demand on the data size, capable of rapidly and accurately assessing the healthy condition of the electric transmission and transformation equipment, and relatively high in practicability.

A tax-involved risk assessment method for a real-estate project comprises the following steps: according to the type of a real-estate project to be analyzed, selecting a similar project formation sample set from an existing real-estate cost information base; calculating an average value and a corresponding interval for each cost subject according to cost data in the sample set; comparing the cost subject with the corresponding interval, determining the cost subject as a low risk subject if the subject falls within the interval, and determining the cost subject as a high risk subject if the subject falls outside the interval; and if the local government formulates a cost limitation value and presets it to a forced value, once the cost subject exceeds the preset value, adjusting the risk value to a higher value. A risk quantification method analyzes risks and associates the analyzed risks in combination with the interval, conducts calculation by weight value, assigns values to high and low risks, and compares the risks with a plurality of preset risks areas to obtain an assessment result indicating risk levels.

The invention discloses a ceiling earthquake damage identification method based on a convolutional neural network, belongs to the field of computer vision, and aims to solve the problem of relativelylow efficiency of earthquake damage degree investigation by professionals on site at present. The identification method comprises the steps of 1, performing normalization processing on ceiling earthquake damage pictures; 2, performing rotation and mirroring processing on the target sample; 3, loading a pre-trained AlexNet model, modifying the last full connection layer of the model, and performingback propagation by adopting a cross entropy loss function and an Adam adaptive optimization function to obtain an initially-trained AlexNet model; 4, continuing to train the model, and adjusting thelearning rate, batchsize and the like of the neural network; and 5, evaluating the test sample by using the trained AlexNet model. According to the method, the use function of the post-earthquake building can be accurately, quickly and timely evaluated only by means of a certain hardware module and in combination with the trained model, and the model accuracy is high.

The invention discloses a neural network model pruning method and system based on adaptive batch standardization. Randomly sampling the number of floating points to serve as the pruning rate of each layer, generating pruning rate vectors (r1, r2,..., rL) as pruning strategies under the limitation of preset calculation resources, and pruning the model based on the pruning strategies to form a pruning model candidate set; updating statistical parameters of batch standardization layers of the pruning models in the candidate set by using a self-adaptive batch standardization method; evaluating andobtaining the classification accuracy of the neural network model with the updated statistical parameters, and finely adjusting the model with the highest classification accuracy on the training setuntil convergence to serve as a final pruning model. According to the method, the candidate sub-networks are quickly and accurately evaluated by adjusting the batch standardization layer, and the parameters of the final pruning network are obtained by finely adjusting the winning pruning strategy in the quick evaluation method, so that huge time consumption required for finely adjusting all pruning networks is avoided, and meanwhile, the accuracy rate also has the advantage.

The invention discloses a human body posture evaluation method and device, computer equipment and a storage medium, and the method comprises the steps: obtaining a to-be-evaluated image which comprises a user posture; inputting the to-be-evaluated image into a preset human body posture estimation network to obtain human body key point data, the human body key point data comprising key point coordinates and a human body surrounding frame; zooming the to-be-evaluated image through the human body surrounding frame, and performing coordinate transformation on the key point coordinates according tothe zoomed to-be-evaluated image to obtain to-be-evaluated coordinates; acquiring standard comparison data, wherein the standard comparison data comprises standard coordinates and confidence corresponding to each standard coordinate; calculating the similarity between the to-be-assessed coordinates and the standard coordinates through the confidence coefficient to obtain assessment information ofthe to-be-assessed image; by evaluating and analyzing the similarity between each to-be-evaluated coordinate and the standard coordinate in the to-be-evaluated image, the human body posture of the user can be quickly and accurately evaluated.

Improved systems and methods for improving lean muscle mass monitoring of patients and for identifying and treating at least one of Cachexia or Sarcopenia are disclosed. The system may include a dynamometer provided to patients, the dynamometer being configured to perform remote monitoring. The dynamometer may be used and data may be collected by patients without the supervision of a healthcare professional. Using the collected data, lean muscle mass may be tracked over time, to provide fast and more accurate assessment of conditions. Using such an assessment, early treatment options can be targeted to patients to avoid progression into various disease states, for examples, more severe states of at least one of Cachexia or Sarcopenia.

The invention provides a method for accurately and quickly checking carbon emission of the same series of products. The method is characterized by comprising the following checking steps of firstly, analyzing a production process of products A and then carrying out modular decomposition on the carbon emissions of all production processes of the products A to obtain a plurality of modular unit carbon emissions with simple structures, clear boundaries and accurate and controllable data according to a checking purpose and a theory of modularization of carbon emission; and transferring the unit carbon emissions of the products A or partially modifying the unit carbon emissions according to the characteristics of the production processes of products B of the same series, and then quickly checking the carbon emissions of the products B of the same series through modular combination. By adopting the method for accurately and quickly checking carbon emissions of the same series of products, the complicated, changeable and mutually-entangled carbon emission influence factors and the skeleton in the production process are clearly sorted to check the carbon emissions of the products. The complexity for checking the carbon emissions of the products is reduced; the data reuse rate and the checking working efficiency are improved; and the time and the economic cost are saved.

The invention discloses a residual life evaluation method for a low-voltage and multi-core cable. The method comprises the steps of during the simulation test on the operation of a low-voltage and multi-core cable sample made of the insulating material, providing an Alzheimer's life evaluation equation on the actual working temperature of the low-voltage and multi-core cable; measuring the insulation value of the insulating material of the low-voltage and multi-core cable to obtain an insulation resistance of the cable at the end of the life thereof based on a cable residual life equation of the cable insulation value; calculating the service life of the cable based on the Alzheimer's life evaluation equation and the cable residual life equation; verifying the result of the cable residual life equation according to the judgment result of the Alzheimer's life evaluation equation; and estimating the residual life of the on-site low-voltage and multi-core cable according to the insulation value of the low-voltage and multi-core cable during the on-site project implementation process based on the cable residual life equation. Based on the method of the invention, the residual life of the on-site multi-core cable can be quickly and accurately evaluated. Therefore, the method provides an ample foundation for the maintenance and the replacement of the multi-core cable. As a result, the manpower, material resources and financial resources are greatly saved.

A lane recognition algorithm evaluation method, a computer device and a storage medium of the invention are provided. The method comprises: identifying a lane in an original image through a lane recognition algorithm; performing perspective transformation on the original image to convert the original image into a bird's-eye view domain image; acquiring a lane line equation along a first lane edgeline on both sides of an identified lane in a bird's eye view domain image; in the plane coordinate system where the bird's eye view domain image is located, the lane line equation representing the second lane edge line is obtained by fitting the coordinate values of the markers arranged at intervals on the second lane edge lines on both sides of the real lane. Obtaining a lane line equation of asecond lane edge line on both sides of a real lane corresponding to the lane in a plane coordinate system where a bird's eye view domain image is located; calculating a difference amount between the identified lane and the real lane according to the lane line equations of the first lane edge line and the second lane edge line for evaluating the lane recognition algorithm to obtain an evaluation result; The labels required for the present application are small in workload, rapid and accurate in evaluation.

The invention provides a method for establishing an analysis module for assessing neurological functions. The method comprises the following steps: acquiring culture cells which include a plurality of fluorescent labels by virtue of a micro-fluorescent imaging system so as to conduct image analysis, wherein the culture cells include nerve cells and non-nerve cells; in accordance with area size of cell nucleuses and fluorescence intensity, selecting the nerve cells showing axon and dendrite fluorescent labels, eliminating non-nerve cells, and calculating the area sizes of the nerve cell bodies as well as axon and dendrite lengths and the number of branches, so that axon and dendrite growth situations of the nerve cells are determined; and in accordance with axon and dendrite fluorescent labeling ranges, calculating the number of synaptic puncta fluorescent labels on the nerve cells, so that the growth situation of the nerve cells is determined, and the neurological functions are assessed. The method disclosed by the invention can further serve as a drug screening platform, so as to rapidly detect whether drugs have a protective or neurotoxic effect on the nerve cells or not.

The invention discloses an intelligence use value evaluation method. According to the invention, the indirect action process of the information use value is described by using observation, judgment, decision and attack (OODA) loops; the dynamic uncertainty of the information action process is learned and quantified through a statistical relationship learning framework Markov logic network; therefore, rapid and accurate evaluation of the information use value is realized, and the problem that dynamic and rapid evaluation of existing information use value evaluation is difficult to perform underthe condition of big data is effectively solved.

The invention provides a postoperative tumor evaluation method, a device and a computer storage medium. The postoperative tumor evaluation method mainly comprises the steps of obtaining a preoperative image and a plurality of thermal ablation images of a target tumor; according to preset treatment parameters and the thermal ablation images, obtaining thermal dose maps corresponding to the thermal ablation images; obtaining a plurality of target feature vectors according to the thermal dose maps, the preoperative image, the thermal ablation images and the preset treatment parameters; and obtaining a postoperative evaluation result of the target tumor according to each target feature vector. Therefore, the postoperative tumor recurrence risk can be quickly and accurately predicted, so that a clinician can be assisted to more accurately master the condition of a patient.

The invention relates to an evaluation method for monitoring the degradation degree of alpine grassland by using soil arthropods. The method comprises the following steps: (1) collecting information of alpine grassland soil arthropods and target backgrounds; (2) calculating the individual number and class group number difference degree of large and small and medium-sized soil arthropods in alpine grassland with different degradation degrees, and compiling an alpine meadow degradation degree evaluation table according to the individual number and class group number of the soil arthropods; (3) establishing an evaluation sample plot; (4) collecting large and small and medium-sized soil arthropod samples in the evaluation sample plot according to collection specifications, determining family or objective information according to the retrieval graph, counting the number of individuals and the number of classes according to the family or objective, and determining the difference degree of the large and small and medium-sized soil arthropods under the grazing forbidding and grazing conditions through statistical analysis; (5) determining the alpine grassland degradation degree according to the degradation degree evaluation table. The method is simple, accurate and low in cost.

The invention discloses an online learning diagnosis method and device based on artificial intelligence, and the method comprises the steps of collecting the test paper and homework data of a student through an OCR technology, taking the data as a first training set, training, and obtaining a first neural network model; taking the test paper or the homework, the total score of each corresponding knowledge point and the teaching suggestion as a second training set, and training to obtain a second neural network model; obtaining the answer content of the to-be-analyzed test paper or homework through the handwriting of an electronic pen, inputting the to-be-analyzed test paper or homework into the first neural network model in combination with the to-be-analyzed test paper or homework, and obtaining the total score of each knowledge point corresponding to the to-be-analyzed test paper or homework; inputting the total score of each knowledge point corresponding to the to-be-analyzed test paper or homework into the second neural network model to obtain the teaching suggestions of the to-be-analyzed test paper or homework. According to the present invention, the real mastering condition of students on each knowledge point can be quickly and accurately evaluated, and proper teaching suggestions are given, so that the time and energy of teachers are saved, and the students can be helped to quickly check the leakage and fill the vacancy, and the score is improved.

The invention discloses a communication impairment assessment method and equipment. The communication damage assessment method comprises the following steps: acquiring position information and coverage area information of a plurality of fault signal stations in a geographic area; calculating a signal damage range based on the position information and the coverage area information; and evaluating the communication damage degree of the geographic area based on the signal damage range. According to the method, the area of the communication damaged area in the area can be quickly and accurately evaluated by fully utilizing the geographic position distribution of the signal stations and combining the connected subgraph and the polygon convex hull algorithm, so that a more accurate judgment premise is provided for subsequent emergency response.