80 results about "Manual interpretation" patented technology

The invention discloses an image-based railway contact net bird-nest abnormal condition detection method. According to the invention, an abnormal condition that there exists a bird nest in a railway contact net patrol image is automatically identified by a digital image processing method. The method provided by the invention comprises the following main steps: 1) image binaryzation of the railway contact net; 2) extraction of pillar main parts and fine lines of the contact net; 3) bird nest suspicious area location based on suspension point detection; 4) feature extraction of linear direction histogram and length histogram of a bird nest suspicious area; and 5) bird nest identification based on linear direction and length distribution characteristics. By the method provided by the invention, bird nest abnormal condition detection and identification of the railway contact net are carried out efficiently and automatically. The method has great safety significance as well as practical application value, is suitable for automatic inspection of high-speed rail and common railway contact nets, can effectively help an inspector rapidly find whether there exists a bird nest on a support of a contact net and troubleshoot potential safety hazard, is adopted to raise work efficiency of inspectors and avoid huge manpower consumption caused by manual interpretation.

The invention belongs to the field of hyperspectral remote sensing application, and particularly discloses a method for automatically identifying rock minerals based on spectral information. The method comprises the following steps: scanning rock minerals to be classified by using a hyperspectral scanner to obtain mineral spectral data; storing the obtained data in a database to be detected; extracting a part of data from the database; identifying mineral species information corresponding to the extracted spectral data according to features of the mineral spectral data by using manual interpretation, and storing the extracted spectral data and the corresponding mineral species information as a learning sample library; establishing an artificial intelligence learning system, and performinglearning and training by using the learning sample library; and detecting the spectral data in the database to be detected by using the artificial intelligence learning system after training and learning optimization to identify the mineral species information; and storing an identification result in an identification result database. By adoption of the method, the human resource consumption can be reduced, the working automation degree is improved, the spectral scanning data processing efficiency is improved, and the economic efficiency is improved.

The invention discloses an ore body modeling method, device and system based on borehole data and a storage medium. The method comprises the following steps of carrying out the discrete sampling on borehole data sampled by geological exploration to obtain discrete sampling points; extracting the sampling points of a sample section and converting into contact point constraints; extracting the sampling points of a non-sample section and converting into out-of-plane point constraints; constructing a discrete point distance field according to the contact point constraint and the out-of-plane pointconstraint; spatially interpolating the distance field of the discrete points to construct an implicit function distance field which is used to characterize the distribution of the distance field ofthe ore body boundary; sampling the distance field of the implicit function and calculating based on a reference isosurface extraction algorithm to obtain a three-dimensional model of the ore body. The ore body 3D modeling can be carried out directly based on the borehole data, which saves the troubles of manual interpretation of ore body profile and effectively improves the efficiency and qualityof ore body modeling.

The invention provides an MODIS remote sensing monitoring method of leucocyan in eutrophic lakes. The method comprises the following steps: images of eutrophic lakes are divided into phytoplankton bloom images and non-phytoplankton bloom images on the basis of MODIS RGB image manual interpretation and field measurement data analysis; with the combination of consecutive two years of satellite-ground synchronous data of the eutrophic lakes, an EOF decomposition-based leucocyan estimation algorithm is obtained; EOF algorithm-based leucocyan estimation results are obtained; and on the basis of the algorithm, a long-term sequence spatial and temporal distribution result of leucocyan in the eutrophic lakes is obtained. According to the method, pigment concentration is estimated after image classification. Thus, estimation result is more accurate, and a more comprehensive evaluation of water quality can be made. The method of the invention provides scientific supports for scientific decision-making of water resource management and water environment protection for water conservancy and environmental protection departments, etc.

The invention relates to a camber admeasuring apparatus and a laser camber measuring method. After level gauge providing a level baseline, lighting target of data collector gathers its facula and sends the facula image to computer processing system. Based on the image, the computer calculates the vertical ordinate D of baseline on lighting target, data collector measures longitudinal separation H between measured point and itself simultaneously and sends it to the computer. Then the computer calculates longitudinal separation between measured point and level baseline. The invention adopts laser to measure intelligently, level gauge to provide baseline and high precision, data collector to complete laser ranging. Thus, camber admeasuring of lower flange can be completed on the ground. Computer deals with data directly to avoid manual interpretation, which achieves high accuracy, high efficiency and simple operation.

The invention relates to a measurement method for a time delay test of a quadrature phase shift keying modulator. A relation formula, meeting the requirements of time delay measurement, of input and output of the quadrature phase shift keying modulator is given, and a time delay measurement model is built through the principle of maximum likelihood estimation. Based on the above, a time delay test system formed based on a digital oscilloscope and a baseband signal and a measurement method for the time delay test are given, and the error components caused by the manual interpretation of the jump edges of input and output signals are eliminated. When a measurement main regulator digital oscilloscope carries out signal collection at a sampling rate of 25GS/s, the uncertainty of a time delay measurement result is not less than 100ps. The system is high in measurement accuracy, is simple and feasible in measurement method, is relatively low in requirements for hardware, is high in real-time response, and can be used for the link time delay calibration of a high-precision satellite measurement and control and navigation positioning system.

The invention discloses a splitting method of a CD image of a solid engine. The method comprises the following steps: a CD faultage image data of the solid engine is read in, an artifact which is formed outside the solid engine owning to the air agitation formed in the CD detection is eliminated; the image, the artifact of which is eliminated, filters and eliminates background noise by median filter, meanwhile, the edges of each component of the solid engine are sharpened in the CD image and a casing, a propelling agent and an astropyle of the solid engine is cut in the CD image; a defect is cut from the propelling agent. By adopting the method, an outer layer air ring artifact, the casing, the propelling agent, the astropyle and various defects of the solid engine in the CD image are cut automatically and accurately, thus solving the problem of low efficiency and poor accuracy in the CD detection image of the solid engine by the manual interpretation as well as being not limited by the eigenvalue of image information.

The invention discloses an effective feature extraction and tree species identification method oriented to a tree laser point cloud, which comprises the following steps of obtaining full-coverage point cloud data of a target tree; denoising the full coverage point cloud data of the target tree to remove anomalies; according to the point cloud data obtained after noise reduction, analyzing the treecharacteristics of three kinds of target trees respectively; extracting the optimal feature parameter group based on the relative clustering feature of trees; extracting the optimal feature parameters based on point cloud distribution features; extracting the optimal feature parameters based on apparent features of trees; combining and inputting the optimal feature parameters of three tree species to SVM classifier as variables to classify tree species. The invention achieves higher tree species classification accuracy, provides a powerful tool for obtaining more accurate forest tree speciesdistribution, reduces the high cost of field entity investigation, is time-consuming and laborious, and reduces the error caused by manual interpretation.

The invention provides a module lens test method and system. The module lens test method comprises that a capture unit captures images formed after lens resolution to form a resolution power picture; the capture unit carries out sampling on edges of a frame in the resolution power picture to form a plurality of sampling points; a processing unit carries out black and white contrast on the grayscale of the sampling points and forms a data fitting curve; and the processing unit carries out Fourier calculation on the data fitting curve and obtains a spatial frequency response value of a lens. The module lens test method and system realize fully-automatic module lens test; and since the processing unit carries out sampling and processing, that is, Fourier calculation, on the data fitting curve automatically, automatic processing of resolution power high-frequency test of the module lens can also be realized, and manual interpretation is not needed; and the module lens test method and system are suitable for full-range (low-frequency, intermediate-frequency and high-frequency ) resolution power test.

The invention provides a method for extracting fracture structure information by means of image processing and priori knowledge. The method comprises the steps of (S1) obtaining a high-resolution remote-sensing image and basic geological data of an engineering area, (S2) carrying out edge enhancement on the image, (S3) extracting surface feature edge information by means of the edge detection algorithm, (S4) extracting linear information by means of the form matching algorithm, (S5) carrying out connection and removing on the linear information by means of the priori knowledge, and (S6) obtaining the fracture structure information through comprehensive analysis by means of a geographic information system. According to the method for extracting the fracture structure information by means of the image processing and the priori knowledge, the fracture structure information is extracted by means of the remote-sensing image processing technique; compared with a manual interpretation method, the method has the advantages that influence of experience and subjective factors of interpretation operators is greatly reduced, the accuracy of the extraction of the fracture structure information is improved by over 80%, and a novel technology is provided for better and faster extraction of the fracture structure information.

The invention provides a method for detecting a fluorescent brightening agent in paper. The method comprises the steps as follows: 1), sample pretreatment: preparing a standard fragment sample, and soaking the standard fragment sample by different solutions respectively after weighed; 2), adjusting the excitation wavelength of a fluorospectro photometer to be 345 nm, and detecting the sample; 3), weighing a defined quantity of a VBL fluorescent brightening agent to configure a standard liquid with appropriate concentration, scanning an excitation curve and an emission curve, and measuring a standard curve; and 4), measuring the content of the fluorescent brightening agent in different paper samples, and obtaining the content of the fluorescent brightening agent of the detected paper sample through comparison with the standard curve. Compared with a traditional detecting method, the method provided by the invention can detect relevant data rapidly and accurately, the problem of manual interpretation errors caused by manual operation is avoided, and the health and safety of relevant toilet paper products can be elevated accurately; and meanwhile, experimental instruments adopted by the invention are simple, the data repeatability is good, the reliable quality monitoring guarantee is provided, and accordingly the guaranteeing advantages on the product quality monitoring are achieved.

The invention provides an online and offline fused medical image quality interpretation self-adaption method and system, and a storage medium. According to the online and offline fused medical image quality interpretation method, ''online'' operation is reflected in that a doctor can acquire a medical image quality interpretation result in real time at the image acquisition end; ''off-line'' operation is reflected in correction of prediction results by an image storage end and a medical cloud server and retraining and upgrading processes of an AI model; and ''self-adaption'' is reflected in that a closed loop is formed in the whole online and offline processes, so that online real-time return and offline feedback correction are continuously iterated and optimized, and online and offline fusion is realized. The online and offline fused medical image quality interpretation method ensures the real-time performance of acquiring the quality interpretation result by the doctor, continuouslyimproves the generalization and accuracy degree of the model by retraining and upgrading the AI model, has great significance in clinical application, can greatly reduce the time cost of manual interpretation of doctors, improves the accuracy and efficiency of auxiliary diagnosis of intelligent medical image quality interpretation, and achieves early discovery and early modification of low-qualityimages to the maximum extent.

The invention provides a pavement crack disease detection method. The method comprises the following steps: laser generators emitting laser beams to lanes to form a series-connected bending marking line on the to-be-detected pavement; an image acquisition device, based on a preset collection period, acquiring a series-connected bending marking line image to obtain marking line images; carrying out high-pass filtering on the marking line images to obtain pulse elevation numerical images of the pavement crack disease; and processing the pulse elevation numerical images to obtain a result of the pavement crack disease. According to the pavement crack disease detection method, automatic and intelligent treatment of the pavement disease is realized, the problem that pseudo-cracks generated by water traces and oil traces, existing in an image photography method generally adopted at home and abroad, is solved, so that the automatic interpretation of complex grey-scale patterns is realized, relevant information of the pavement disease is directly obtained through processing, manual interpretation is not needed, and safety is guaranteed.

The invention discloses an intelligent interpretation system and method for broken PCCP (prestressed concrete cylinder pipe) steel wires. The system comprises a vehicle, wherein a receiving probe, a distance measuring wheel and a transmitting probe are arranged on the surface of the vehicle, and a main control computer and a detection host computer electrically connected with the main control computer are arranged in the vehicle; the receiving probe, the distance measuring wheel and the transmitting probe are electrically connected with the detection host computer in the vehicle. According tothe intelligent interpretation system and method for the broken PCCP wires, intelligent interpretation of the number and position of the broken PCCP steel wires can be realized, the broken wires are positioned accurately, the number of the broken PCCP steel wires can be accurately calculated by a fitting expression, and the method has the advantages of accurate and objective calculation of the number of the broken wires, high efficiency and high adaptability compared with manual interpretation .

The invention provides a method for quickly obtaining radar basic reflectivity three-dimensional contour data. The method for quickly obtaining the radar basic reflectivity three-dimensional contour data constructs a triangular mesh by directly using original data points detected by the radar, thereby quickly obtaining the radar basic reflectivity three-dimensional contour data with vector features. The method for quickly obtaining the radar basic reflectivity three-dimensional contour data is beneficial to realize global sharing of radar product data, provides effective decision support for disaster reduction and disaster prevention of meteorological disasters, and has good social and economic effects; meanwhile, the data acquisition has high efficiency and high precision, and is convenient for manual interpretation analysis, which can shorten the response time of relevant departments to disasters.

The invention relates to an intelligent interpretation and evaluation method for logging an oil-gas-water layer. According to the method, logging parameters are scientifically classified on the basisof six classification models of Bayes, decision trees, neural networks, support vector machines, clustering analysis and correlation analysis in data mining Meanwhile, the time sequence database similarity search method based on the time sequence mode is used for quickly determining the gas logging curve, and compared with a manual interpretation method, the time sequence database similarity search method based on the time sequence mode is faster and more accurate and has obvious advantages.

At present, the selection of an effective data segment of a pneumatic thermal test mainly depends on the manual interpretation and marking, the workload is large, and the time and labor are consumed.The invention provides a method for automatically selecting data segments. The method comprises the following specific steps of step 1, collecting the heat measurement test data; step 2, classifying the heat measurement test data through the CNN, and outputting the effective data; step 3, intercepting data by a sliding window; 4, judging data by using a CNN network; 5, outputting the correct probability of the effective data segment; and 6, determining an initial labeling position. The method provided by the invention has the adaptive capacity, can effectively improve the extraction efficiencyof the heat measurement test data, and has the very important significance for the design of a heat-proof structure and the flight safety of a hypersonic flight vehicle.

The invention discloses a scene-based image self-adaptive nonuniformity correction method, which comprises the steps of automatically ranking DN values of an image according to different pixels, determining at least one relatively dark uniform region and at least one bright uniform region, calculating to obtain a correction coefficient of each pixel according to an algorithm, and completing nonuniformity correction of the image. The scene-based image self-adaptive nonuniformity correction method is suitable for nonuniformity correction of massive images with great correction difficulty and severe nonuniformity of complex scenes, dark objects, infrared image and the like, significantly simplifies the complexity and reduces the cost of correction, is based on scenes but breaks through the limitations of the scenes, has good universality and adaptivity, does not need manual interpretation, is fully automatic in calibration, has fast calculation speed and good correction effect, improves imaging quality of the image, and lays a foundation for subsequent image analysis and application.

The invention discloses a rapid monitoring method of large-scale land desertification and aims to develop a rapider and more accurate large-scale desertification monitoring method based on easily available data. The method comprises steps as follows: a research time domain is selected, and the period is divided; a research zone is selected, and changes in a town, a water area, a forest and a farmland in six periods are obtained by use of LUCC data; the average value of NDVI data of recent three years is defined as NDVI data of each period; grassland information and native desert information are acquired through NDVI with a decision tree classification method; a non-desertification region and a native desert region are removed from the research area, so that the desertification region is obtained; 15-degree correction of a slope is performed through auxiliary DEM data. According to the method, easily obtained data are adopted, heavy workload of manual interpretation is avoided, a scientific basis is provided for treatment of desertified land, and the method has great practical significance in guiding desertification control and desertification decision.

The invention discloses a river capture automatic identification method, which comprises the following specific steps of (1) calculating the river reach flow direction feature points and the river reach flow directions based on a research area vector water system graph; (2) forming a confluence pair by the river reaches according to a confluence relationship; (3) based on an included angle betweenthe branch flow direction and the main flow direction, screening a confluence pair approximating to a right angle; (4) calculating an included angle between branches, and screening a confluence pairwith the included angle between the branches being a flat angle and the branches being collinear; and (5) screening a shorter confluence pair with the branch inflow of 0 to generate a river capture seizure layer. Compared with the prior art, the method has the advantages that the automation degree is high, and the excessive misjudgment during manual interpretation can be effectively avoided.

The invention provides a method, device and system for detecting and analyzing inherited metabolic diseases of newborns. An interpretation mode which combines a computer system, a disease database anda doctor is originally created; the tandem mass spectrometry detection result of the inherited metabolic diseases of newborns can be interpreted more quickly and accurately, and an accurate conclusion can be given; the comprehensive datamation of the whole working process is realized; therefore, the interpretation process can be traced back more easily and can be repeatedly compared and researched in the later period, and the situation that retrospective analysis cannot be carried out due to forgetting in the later period of manual interpretation can be greatly avoided. The system adopts a man-machine cooperation mode to realize higher degree of automation, the investment of labor cost is greatly reduced, and with the increase of data volume and the continuous optimization of the system,the tandem mass spectrometry detection result interpretation of the inherited metabolic diseases of newborns is even possible to realize the artificial intelligence interpretation of a computer.

The invention discloses a method and a device for improving image brightness based on background modeling under a low-illuminance scene. According to the method and the device which are provided by the invention, background modeling (such as Gaussian distribution) is carried out on a high-illuminance scene (such as daytime) and motion detection background modeling is carried out on the low-illuminance scene (such as night) respectively, a night background is corrected by the obtained daytime background, and after a motion target of the night scene is extracted, the motion target is fused in the corrected night background to complete a video image under the correction night scene. By adopting the method and the device, details, which cannot be captured at night by a video surveillance camera, can be recovered, so that the brightness of a night picture monitored by a video is remarkably improved and the detail distinguishing degree is increased so as to facilitate manual interpretation.

The invention discloses a natural resource data fusion method, electronic equipment and a computer storage medium. The method comprises the following steps: performing overall graph overlay analysis by using selected fusion data and reference data to obtain an early-stage analysis result; deleting the reference data which are not superposed with the fused data; performing adjacent pattern spot grouping fusion on the reference data according to grouping fields; on the basis of a set superposition cutting condition, cutting the reference data by using the fusion data to obtain result data, and performing fine-crushing merging and attribute processing in the cutting process; performing batch judgment on the result data based on interpretation factor fields and some presettable conditions to determine interpreted pattern spots; or determining interpreted pattern spots based on manual interpretation; and carrying out batch fusion and quality inspection on the interpreted pattern spots, and storing/exporting qualified result data. According to the method, the data overlapping and cutting part is automatically processed by setting parameters/conditions, so that loopholes caused by manual work are avoided, and the working efficiency is improved.

The invention relates to a camouflage effect evaluation method based on kernel density estimation. The method comprises the following steps of automatically obtaining a background sub-region of a target size; calculating the eigenvectors of images and calculating the distance of samples by a similarity measure method; calculating the parameters of the kernel density estimation model and constructing the background feature distribution model; calculating the matching probability of the target features in the background feature distribution model and further calculating the recognition probability of the features; comparing the recognition probability of the features before and after camouflage to evaluate the camouflage effect. Compared with the current camouflage evaluation method, the method of the present invention has the advantages of not relying on manual interpretation, requiring less sample data, and being applicable to the current arbitrary identification features and the like.