40results about How to "Extract automatically" patented technology

The invention relates to a high resolution remote sensing image water body extraction method based on multi-feature combined treatment, comprising: performing radiation scaling on a remote sensing image, obtaining the ground feature apparent radiant quantity, and thereby performing threshold segmentation on pixel length characteristics to eliminate most non-water body pixels; performing threshold segmentation on the NDWI characteristics of water body pixels waiting for selection maintained in the last step, and obtaining a water body object waiting for selection; performing segmentation threshold again on the NDWI characteristics of pixels covered by a small area object waiting for selection, and further eliminating mixed background pixels; and finally using an area growth mode to perform water body edge optimization to obtain a final water body object mask. The method utilizes local structure length characteristics in water body extraction, embodies water body distribution characteristics, inhibits most shadow interference, avoids the problem that different characteristics can not be fully embodied caused by fusing characteristics in a direct classification mode, and automatically extracts water body with a fast running speed.

The invention provides a Deepdive-based field text knowledge extracting method comprising the steps of: (1) acquiring original texts required by a knowledge base construction system and performing pretreatment on the texts; (2) performing entity connection on the pre-treated texts, finding out target entities corresponding to a preset specific relation, generating entity-relation-entity triads and forming a candidate relation-entity pair set; (3) learning and labeling a plurality of candidate relation-entity pairs by using a weak supervising method and generating training samples of a Deepdive tool; (4) inputting the training samples into the Deepdive tool to train Deepdive, and outputting candidate relation-entity pairs with probability values greater than a threshold value to form an extracted knowledge base. The method can complete the work of construction of a field knowledge base, has great expandability and is of high practical value for utilization and extraction of unstructured data.

The invention discloses a GCAM-based high-resolution SAR image airport runway area automatic detection method and system. The GCAM-based high-resolution SAR image airport runway area automatic detection method comprises the steps of performing down-sampling on a high-resolution SAR image to generate a medium-resolution image; inputting the medium-resolution image into a geospatial context attention mechanism network GCAM to extract a runway area; and performing coordinate mapping on the extracted runway region to obtain an airport runway region detection result of the final high-resolution SARimage. Experiments show that compared with DeepLabV3 +, RefineNet and MDDA networks, the method is high in precision and short in consumed time, geographic space information of an SAR image airport can be fully learned, and high-precision, rapid and automatic extraction of a high-resolution SAR image airport runway area can be achieved.

The invention discloses a corpus keyword automatic extraction algorithm based on data mining. The corpus keyword automatic extraction algorithm comprises the following steps of obtaining a to-be-processed text; performing word segmentation processing on the obtained text; performing part-of-speech tagging and meaning tagging on words in the word segmentation set; performing word removal processingon the word segmentation set; counting co-occurrence information of the word frequency and the word pair; setting a word frequency threshold, and judging whether the word frequency of the words in the vocabulary set is greater than the word frequency threshold or not; generating a candidate keyword set; obtaining position information of each candidate keyword in the text; calculating a weight value of each candidate keyword in the text; setting a weight value threshold value, and judging whether a calculation result of each candidate keyword is greater than the weight value threshold value ornot; generating a set of keywords. According to the method, a corpus keyword extraction algorithm is optimized, the operation is simple and convenient, the keywords can be automatically extracted from the corpus, time and labor are saved, and the keyword extraction accuracy is remarkably improved.

The invention provides an element extraction method, device and equipment and a storage medium, and the method comprises the steps: obtaining a target text and an element label defined based on an element extraction demand; extracting target element contents corresponding to the element tags from the target text; extracting an object reference word matched with the target element content from thetarget text to serve as a target object reference word corresponding to the element label; and determining a triple consisting of the element label, the target element content corresponding to the element label and the target object anaphora as an element extraction result of the target text. According to the element extraction method provided by the invention, the elements can be automatically, efficiently and accurately extracted from the target text.

The invention relates to an MRI image heart structure segmentation method based on a multi-path convolutional neural network. The MRI image heart structure segmentation method comprises the steps: collecting cardiac movie MRI training data of a normal person and a heart patient; enabling a doctor with rich experience to manually mark the heart structure in the training data to serve as a heart segmentation marking result; training the heart region extraction model based on the training data, so as to enable the heart region extraction model to accurately extract the heart region; training theheart segmentation network according to the heart region extracted from the training data to segment each structure of the heart; and measuring the segmentation performance of the constructed heart segmentation network by taking the standard segmentation labeling result as the standard. The MRI image heart structure segmentation method adopts the heart region extraction model based on the generative adversarial network to extract the heart, so that the accuracy of heart region extraction is improved. Meanwhile, the MRI image heart structure segmentation method utilizes contextual information between adjacent layers through the multi-path convolutional neural network, so that the segmentation precision and accuracy are improved.

The invention discloses a white blood cell extraction device. The white blood cell extraction device successively includes five parts of an upper cover plate, an extraction module, a storage module, alysis module and a lower cover plate from top to bottom, the upper cover plate is provided with a white blood cell channel, a lysed red blood cell channel and a normal saline channel, the white bloodcell channel and the lysed red blood cell channel are respectively provided with an outlet, and the normal saline channel is provided with an inlet; and the extraction module is provided with a helical extraction channel, a white blood cell channel and a lysed red blood cell channel, the storage module is provided with a reaction channel, the lysis module is provided with a mixing channel, and the lower cover plate is provided with a blood channel and a red blood cell lysate channel. According to the white blood cell extraction device, high-activity white blood cells can be automatically, rapidly and steadily extracted from undiluted blood, lysed red blood cells are removed at the same time, and a quick and effective way is provided for application such as clinical diagnosis of subsequentdiseases.

The invention discloses an information extraction method and device and electronic equipment. The method comprises the steps of obtaining a semantic vector of a to-be-extracted text, and predicting a main body element in the to-be-extracted text and first position information of the main body element based on the semantic vector; obtaining a first semantic vector of the main body element from the semantic vector based on the first position information of the main body element; predicting an object element in the text to be extracted and second position information of the object element based on the first semantic vector and the semantic vector of the main body element; according to the first position information and the second position information, extracting a text fragment from the to-be-extracted text to generate structured information of the to-be-extracted text. Therefore, according to the first position information and the second position information, the information can be automatically extracted, manual intervention is not needed, the labor cost is saved, the extraction time consumption is shortened, the accuracy of an information extraction result is ensured, and the efficiency, accuracy and reliability in the information extraction process are improved.

The invention discloses a method for dynamically extracting relationships between entities and attributes. The method comprises the following steps of: obtaining text data; dynamically extracting features of entities and attributes from the text data on the basis of a dynamic entity-attribute relationship library and a training model. Moreover, the invention provides a server and a readable storage medium. By adoption of the method, the dynamic entity-attribute relationship library and the training model are adopted, and the features of the entities and the attributes can be automatically extracted from the text data.

The invention discloses a disparity map DBSCAN clustering-based region of interest extraction method, which comprises the following steps of 1, performing matching calculation on a binocular view to obtain a disparity image; 2, deforming the parallax image into an image matrix with one row and N column vectors, wherein N is the product of the width W and the height H of the image; 3, taking the absolute value distance of the parallax as a DBSCAN sample distance, and carrying out the clustering analysis of the image matrix through employing a DBSCAN clustering algorithm; and 4, mapping a clustering result back to a certain path in the original binocular view, and obtaining a region of interest according to the clustering result. The method does not need manual interaction and sample set training, can automatically extract the region of interest at each depth, is not affected by a complex background, and facilitates the further analysis and processing of the image.

The invention discloses a method, a system and a device for head image processing, and a storage medium. The method comprises: judging the type of each cross section image according to the cross section image sequence of the head and an eye socket classification algorithm. According to the technical scheme, the eye socket part image sequence is extracted from the classified cross section image sequence, the technical problem that in the prior art, the processing efficiency of the head image is low is solved, manual operation is not needed, the eye socket part image sequence is automatically and accurately extracted from the cross section image sequence of the head, and the processing efficiency of the head image is improved.

The invention relates to the technical field of artificial intelligence, and discloses a voice extraction method for a target speaker, a device and equipment, and a medium, and the method comprises the steps: determining a plurality of first to-be-extracted voice data segments through employing a preset segmentation method according to first to-be-processed voice data in a first direction; performing segmentation extraction on second to-be-processed voice data in a second direction according to the plurality of first to-be-extracted voice data segments to obtain a plurality of second to-be-extracted voice data segments; performing data extraction on the plurality of first to-be-extracted voice data segments and the plurality of second to-be-extracted voice data segments at the same time to obtain a plurality of to-be-extracted voice data segment pairs; and inputting each to-be-extracted voice data segment pair into the single speaker voice extraction model for voice extraction to obtain a plurality of target speaker voice data segments, and splicing the target speaker voice data segments according to a time sequence to obtain target voice data of the target speaker. Therefore, the cost of service quality evaluation is reduced, and the comprehensiveness of service quality evaluation is improved.

The invention discloses a rotary structure of a nucleic acid extraction device, relates to the technical field of nucleic acid extraction, and comprises three layers of structures: an upper-layer structure, a middle-layer structure and a lower-layer structure, the upper-layer structure and the middle-layer structure are rotatably connected through a shaft, an elastic element is arranged outside the shaft, the upper-layer structure is provided with a plurality of annularly distributed upper layer liquid storage chambers, the upper-layer liquid storage chambers surround a sample chamber, the bottom of each upper-layer liquid storage chamber is communicated with the bottom of the sample chamber, and the sample chamber extends outwards to form a hole communicated with a reaction chamber; the reaction chamber is located in the middle-layer structure, a groove matched with the shaft is formed in the top of the middle-layer structure, the middle-layer structure is provided with a first channel and a second channel, and the first channel, the second channel and the upper-layer liquid storage chamber are located on the same circle; and the lower-layer structure and the middle-layer structure are coaxially and fixedly connected. Compared with the prior art, the rotary structure of a nucleic acid extraction device has the following beneficial effects: (1) nucleic acid is automatically extracted, which liberates labors, and is convenient and fast; (2) the extraction result is stable; (3) little pollution and high extraction purity are achieved; and (4) the kit is small and exquisite, convenient to carry and operate and wide in application range.

The invention provides a checkerboard angular point automatic extraction method, system and device and a medium. The checkerboard angular point automatic extraction method comprises the following steps that a checkerboard image is obtained, binarization processing is conducted on the checkerboard image, a checkerboard binarization image is generated, and a plurality of contours are extracted from the checkerboard binarization image; grid contours are screened out from the multiple contours, and then the center point of each grid contour is determined; according to the center points of the four adjacent grid contours, coarse angular points of each grid area are determined; and extracting a sub-pixel of each coarse corner point, and determining a target corner point of each grid region according to the sub-pixel of each coarse corner point. According to the invention, the plurality of contours are extracted from the checkerboard binary image, the lattice contours are screened out from the plurality of contours, and then the central point of each lattice contour is determined, so that the coarse angular point of each lattice area can be determined, and then the target angular point is determined, and therefore, the angular points of the checkerboard are automatically extracted, and the accuracy of the checkerboard angular point extraction is improved. And the number of the checkerboards in the checkerboards does not need to be known.

The invention relates to a method and device for extracting official document elements based on OCR character recognition, belongs to the technical field of intelligent file processing, and solves theproblems that an existing method wastes manpower and time and is low in efficiency. The method comprises the following steps: scanning a paper official document file containing official document element information to obtain an electronic official document file; dynamically generating an algorithm selection box based on the electronic official document file, and obtaining a corresponding algorithm based on the algorithm selection box; acquiring official document element information in the electronic official document file based on an algorithm; and storing the obtained official document element information to a corresponding field position of an official document element form in a business information processing system. The invention is simple and feasible, is easy to implement, achievesthe quick extraction of the official document elements, saves the manpower and cost, and improves the efficiency.

The invention discloses a dry land soil leaching solution collecting device based on agricultural technology research, and relates to the field of soil detection.The dry land soil leaching solution collecting device comprises a base and ground anchors, the ground anchors are in threaded connection with the bottom of the base, a stand column is welded to the upper surface of the base, guide grooves are symmetrically formed in the two sides of the stand column, and sliding seats are slidably arranged on the surfaces of the guide grooves; a fixing plate is welded to one end of the sliding seat, a hydraulic jack is movably installed on one side of the fixing plate, a collecting box is arranged at one end of a piston rod of the hydraulic jack, a limiting frame is arranged at the top of an inner cavity of the collecting box, and a filtering plate is arranged on the surface of the limiting frame. The device solves the problems that when existing equipment collects a leaching solution, excavation needs to be conducted at a collection position, then the collection equipment is buried in soil, when the collection method is used, an original soil layer above the collection equipment is damaged, and therefore the research accuracy of the leaching solution is affected.

Equipment for picking electronic component (91) from dissected wafer (9) includes an optical device (8) which is allowed for screening the image of the first side of the dissected wafer (9); and a lighting unit (3, 40) for lighting the second side of the dissected wafer (9) facing the first side. A method for picking the electronic component (91) includes screening the first side of the electronic component (9) and lighting the second side of the electronic component (9) facing the first side. The back light irradiation of the electronic component (91) to be picked allows its boundary and its position to be determined by the optical device accurately.

The invention discloses a communication abnormity discovery method and device based on baseline behavior description, and the method comprises the steps of carrying out the data cleaning and preprocessing of to-be-measured traffic data, and obtaining a traffic matrix y; and constructing an auto-encoder by using a CapsNet model, obtaining a reconstructed traffic matrix of the traffic matrix y based on the auto-encoder, and obtaining a communication anomaly discovery result according to the traffic matrix y and the reconstructed traffic matrix. According to the method, the features can be automatically extracted from the input data, the specific positions between the traffic features and the arrangement sequence between the data can be used as learning features, and the direction of the vector can represent attributes such as the size and the relative position of the feature value, so that normal behaviors can be identified, known abnormal behaviors can be detected, and novel unknown network attacks can be found.

The invention discloses a building vector extraction model based on deep learning and an extraction method thereof, and the method comprises the steps: S1, receiving a remote sensing image, carrying out the feature extraction of a plurality of stages of the remote sensing image, and obtaining the extraction features of a plurality of scales; s2, performing feature optimization on the extracted features of each scale to obtain optimized features of a plurality of scales; s3, performing feature fusion on the optimized features of the multiple scales to obtain fused features; s4, performing feature recovery and category judgment on the fusion features to obtain a building preliminary extraction result; and S5, post-processing the preliminary building extraction result to obtain a final building vector result. According to the scheme, the feature extraction advantages under multiple scales are fused, the feature extraction capability is effectively improved, the information extraction capability of the network on the original image is enhanced, the extraction result is extracted according to the fused features, post-processing is performed on the extraction result, a relatively regular building vector result can be directly obtained, and subsequent editing and application are greatly facilitated.