148results about How to "Resolve overlap" patented technology

The invention discloses an entity relationship joint extraction method. The method comprises the steps of performing data preprocessing on an input sentence; mapping each word in the input sentence into a corresponding word vector; inputting the obtained word vectors into an entity relationship joint extraction model based on a long and short-term memory network and a graph convolutional neural network for training; and adopting the trained LSTM-GCN model to carry out entity extraction and relationship extraction. According to the method, the sequence information and the region information ofthe input sentence can be captured at the same time through the LSTM and the GCN, each word can be better expressed, the performance of entity extraction and relationship extraction is improved, and the method has certain practicability.

A regional people stream density estimation method based on pixels and a support vector machine (VCM) is characterized by comprising the following steps: extracting a background from a video to be processed; performing background subtraction and binary calculations on a target image, counting the number of non-zero pixel values, and judging and performing background updating; computing a perspective correction coefficient with a perspective correction algorithm based on interpolation weight, and improving effect from perspective distortion to the algorithm; counting the number of the pixels after being corrected through the perspective correction coefficient; using a pixel-based method to estimate a people stream density when the number of the pixels is smaller than the number of normal people stream density pixels; and graying the target image when the number of the pixels is larger than the number of the normal people stream pixels, and extracting texture information of the target image and recognizing the target image by categories with the VCM so as to estimate a people stream density. The regional people stream density estimation method based on the pixels and the VCM can be used for people stream density monitoring in a region with a larger area, and has the advantages of being fast, convenient, accurate, reliable, and capable of timely providing a decision-making basis for an administrative department.

The present invention discloses a feedback transmission method and device. The method comprises the steps of using a base station to determine the feedback information corresponding to the uplink transmission of a terminal and / or the resources bearing the feedback information at least according to the identification information of the terminal; using the base station to transmit the feedback information to the terminal by the resources bearing the feedback information. The feedback transmission method and the feedback transmission device disclosed by the present invention are used to solve the problems caused by the feedback resource overlapping in the prior art after the same resources used by the terminal at the uplink autonomous transmission collide.

The invention relates to an artificial climate system, and particularly relates to an artificial climate system for a soil-vegetation-atmosphere continuous system model test, belonging to the field of environmental geotechnics. The artificial system can simulate the natural change of climate environments such as wind, rain, illumination, temperature, humidity and the like. The artificial climate system consists of an artificial climate box body, a box body liner, a light source, a rainmaking device, an intelligent control system, a multi-channel converter, an airflow circulation system, a rainwater circulation system, a temperature / humidity regulation system, a CO2 and O2 regulation system, a climate parameter sensor group and a digital imaging system. The system can realize automatic feedback regulation of the rainfall intensity, the illumination direction and intensity, the wind speed and direction and the temperature and humidity. Meanwhile, by adopting a top-rotatable rainfall illumination integrated device, the problem of position overlapping of the rainfall device and the illumination device is solved, and the uniformity of the illumination intensity and rainfall intensity in the horizontal plane is guaranteed.

The invention provides a method for synthesizing virtual sight points of a vehicle-mounted multi-lens camera looking-around system. The method comprises the following steps that S1, a left sight point reference image and a right sight point reference image are input in the looking-around system and disparity estimation is conducted so that a left sight point depth map and a right sight point depth map can be obtained; S2, according to the left sight point depth map and the right sight point depth map, a left sight point predicated depth map and a right sight point predicated depth map are generated respectively and correspondingly; S3, according to the left sight point reference image, the right sight point reference image, the corresponding left sight point predicated depth map and the corresponding right sight point predicated depth map, a left sight point predicated map and a right sight point predicated map are respectively generated; S4, a virtual sight point image is synthesized through the left sight point predicated map and the right sight point predicated map. According to the method for synthesizing the virtual sight points of the vehicle-mounted multi-lens camera looking-around system, the thought of a depth credible image is introduced, a shielding image is generated through the depth credible image, and the shielding image can clearly and accurately describe the information that interpolation information should come from which specific zone of which sight point when the virtual sight point image is synthesized. Theories and experiment results prove that a shielding zone can be effectively obtained through the method and the quality of the synthesized virtual sight point image is improved. However, according to existing methods, detection can not be conducted effectively and the virtual sight point image with the high quality can not be obtained.

The invention relates to a method and a device for graphical display of network topology. The method comprises the following steps: determining the display center of a network topological graph; laying out core nodes and determining the sector angle of each core node; for son nodes which are not laid out, setting that each son node which is not laid out to equally divide the sector angle of each father node which is laid out; and executing the following steps: (1) for that the son node P contains superior son nodes or more than 1 primary son nodes, judging adjacent peer son nodes Q which are not laid out; (2) if the son nodes Q does contain superior son nodes and the primary son node number is 0, adding the angle occupied by the son nodes Q to the son node P; and if a son node R does contain superior son nodes and the number of the primary son node is 1, adding half of the angle occupied by the son node R to the son node P; and (3) laying out each son node which is not laid out on the central lines of the owned sectors. By using the method and the device, the problem of node overlapping in network visualization can be solved well.

An automatic counting and controlling device for rod material is composed of a china-type conveying platform, a sorting-queuing unit consisting of the barrier plates for sorting rods and barrier blocks for queuing location, and a counting unit consisting of counting wheel, rod driving slots, infrared photoelectric sensor and counting and controlling unit.

The invention discloses a method for semi-automatic marking of a three-dimensional (3D) model based on fuzzy K-nearest neighbor. The method comprises the following steps of: firstly establishing a 3D model training base; then extracting characteristic vectors of the input model to be marked, and matching similarity of the model with the model in the training base to find k neighbor models; then classifying by using a fuzzy K-nearest neighbor classifier to obtain a fuzzy classification result; and finally calculating classification uncertainty, and marking the models which are uncertain in classification by a feedback and retraining studying method. The marked models are added into the 3D model training base to further expand the model base. In the method for semi-automatic marking of the 3D model, the fuzzy classification result comprising more information is used; classification labels of the models with uncertain classification are determined by aid of feedback so as to solve overlapping problem of feature clustering edges; and a mechanism for training study is used to further automatically mark.

The invention provides a text-to-image generation method of an adaptive attribute and instance mask embedded graph. The method comprises the following steps: obtaining 64 * 64, 128 * 128 and 256 * 256semantic layouts by using a frame regression network according to an input text; according to the 64 * 64 semantic layout, the sentence embedding vector and the random noise, generating coarse-grained image potential features and a low-resolution 64 * 64 image in a low-resolution generator; forming a pixel-level feature vector in the first high-resolution generator according to the 128 * 128 semantic layout; generating a first fine-grained image potential feature and a high-resolution 128 * 128 image in a first high-resolution generator according to the 128 * 128 semantic layout, the coarse-grained image potential feature and the pixel-level feature vector; and generating a high-resolution 256 * 256 image in a second high-resolution generator according to the 256 * 256 semantic layout, the first fine-grained image potential feature and the pixel-level feature vector. By implementing the method, the resolution of the image is high, the shape constraint of the instance is met, and the attribute characteristics are consistent with the description.

The invention discloses aninkjet printing method, inkjet printing device, and inkjet printing apparatus. An image to be printed is converted into a plurality of sub-image matrices of equal size, based on the eclosion points set, a sub-image matrix is divide into three image matrix units, the first image matrix unit of the three image matrix units are combined with the mask image matrix to generate a second image matrix, the third image matrix unit and the complementary mask image matrix are combined to generate a third image matrix, so as to diffuse an error caused by the printing accuracy ofthe ink jet printing apparatus and the driving motor from an element in the first image matrix to each element in the sub-image matrix, and outputthe data of the plurality of the fourth image matrices to further spread the distance error of the two paper feeds onto the plurality of the fourth image matrices, thereby effectively solving the problem of overlapping or blank spacing of the printed images, and further ensuring the quality and quality of the printed images.

The invention discloses a dual-energy under-sampling substance recognition system and a method based on straight path scanning. Low-energy projection data acquired by the whole layer of low-energy detector arrays in the straight path scanning process is used for carrying out CT reconstruction on the detected object; in a CT reconstruction image, a region image segmentation method is used for carrying out region segmentation and marking on a CT image according to attenuation coefficients; and meanwhile, projection data acquired by a few of high-energy detectors arranged behind a part of low-energy detectors is used for realizing dual-energy substance recognition imaging. Because straight paths are used and rotation is not needed, the mechanical design is simple, and the inspection clearance rate is high; and because only a small amount of the dual-energy detectors are used, the system cost is low.

The method comprises: the VPLS users in virtual private LAN network segment accesses public network through client edge (CE) device and provider edge (PE) device; the public network IP address is set in the interface of connecting CE and PE; a CE public gateway is connected to the interface of connecting CE and PE in order to make message transmission between VPLS user and public network. By the invention VPLS users can access public network.

The invention discloses an X-ray light field imaging and calibrating method based on a pinhole array. The X-ray light field imaging and calibrating method based on the pinhole array comprises the steps that (1) An X-ray light field imaging model based on the pinhole array is established and comprises an X-ray light source array, the pinhole array and an X-ray image sensor; (2) a linear exposure mapping image is established so that exposure of images of an X-ray light field can be normalized; (3) the pinhole center of each X-ray light field image is calibrated so that the central position of each pinhole image is extracted; (4) according to the center, obtained in the step (3), of each pinhole image, each X-ray light source image is extracted; (5) according to each X-ray light source image extracted in the step (4), the pixels at the same position in the pinhole images are extracted and reorganized into a frame of X-ray light source image; (6) the extracted X-ray light source images are converted into X-ray light source images which are orthogonally arranged. According to the X-ray light field imaging and calibrating method based on the pinhole array, the light field imaging principle of visible light is applied to X-ray imaging, and the requirement for the imaging quality of the X-ray images during clinical diagnosis and treatment is met.

The invention discloses a continuous multi-screen capturing method, which comprises the steps of (S1), calling a screen capturing command, completing capturing for the current screen, enabling the captured screen to act as an initial screenshot image; (S2), calling a screen scrolling command, and scrolling the current screen; (S3), capturing the scrolled screen to act as a scrolled screenshot image; (S4), detecting edges of the scrolled screenshot image and the initial screenshot image, and synthesizing the scrolled screenshot image and the initial screenshot image into a synthetic image in which only required content is reserved, and then enabling the synthetic image to act as a new initial screenshot image; (S5) repeating the step (S2) to the step (S4) until capturing and synthesis are completed. The continuous multi-screen capturing method solves a problem that disordering and overlapping exist when screen capturing is carried out on continuous multi-screen content in the prior art, the operation experience of users is improved, defects of an existing scheme are made up, the integrity, the uniformity and the operability of a final screenshot image are realized, and a requirement of the users for long screenshots is met.

The embodiment of the present invention provides a pedestrian detection method. The method comprises the steps of receiving a to-be-detected picture; adopting a neural network to determine the pedestrian information in the to-be-detected picture, wherein the pedestrian information comprises the pedestrian positions and the key point positions of the pedestrians; carrying out the postprocessing on the pedestrian information in the to-be-detected picture to obtain a pedestrian detection result. Therefore, according to the embodiment of the present invention, for the to-be-detected picture, the neural network can be adopted to obtain the pedestrian positions and the key point positions of the pedestrians, so that by introducing the key point information of the pedestrians, the shielding and overlap problem which is difficult to solve in the conventional pedestrian detection method is solved, and the detection accuracy is improved.

The invention relates to the field of graphene composite materials and discloses a nano silver loaded graphene antibacterial composite material and a preparation method thereof. A single-layer graphene oxide solution is obtained through a secondary ultrasonic dispersion technology and a discontinuous sheet layer is prevented; a carbon cake formed by mutually mixing a plurality of layers of graphene oxide sheets is innovatively subjected to secondary hydrothermal method reaction; green reduction agents including glucose and ascorbic acid and a green stabilizer soluble starch are adopted respectively and the hydrothermal method reaction is matched, so that the reduction rate is extremely improved and a sealing system with relatively high temperature and inner pressure is adopted; the recycling of dehydrated p conjugation is also promoted and the minimization of defects is facilitated; after heat treatment is carried out, the structure of graphene is easy to recover; the antibacterial activity of free Ag nanoparticles is remained in the composite material. The loading amount of silver particles in the nano silver loaded graphene antibacterial composite material reaches 58.0 percent and the antibacterial rate on escherichia coli reaches 99.9 percent.

The invention discloses a virtual viewpoint synthesizing method based on layered processing. Firstly preprocessing of depth correction and color correction is performed on depth maps of left and right viewpoints; then the image of a virtual viewpoint is generated by adopting layered 3D Warping; color images and depth images of left and right reference viewpoints generated by DIBR are integrated according to layered information, and then filtering processing is performed on the integrated depth maps; then the image of the virtual viewpoint is corrected according to the depth maps, and holes in the image are eliminated; and finally image interpolation and post processing are performed, a ghost phenomenon in the image is removed and finally a high-quality new viewpoint is generated. An intermediate viewpoint can be rapidly synthesized according to any two viewpoints, and picture quality of the newly synthesized viewpoint is relatively high.

A method for determining a critical petroleum inclusion reservoir forming moment based on a time-resolved fluorescence spectrum is disclosed. The method includes a step of building a laser induced time-resolved fluorescence spectrum measuring system; a step of data mining of 'micro-oil' and 'macroscopic-oil' multi-dimensional spectroscopic data based on time-resolved fluorescence; and a step of determining a critical oil and gas reservoir forming moment. A device for determining a critical petroleum inclusion reservoir forming moment based on a time-resolved fluorescence spectrum is also disclosed. The device includes the time-resolved fluorescence spectrum measuring system, a time delay system and a micro-imaging system. Based on three-dimension fluorescence spectrum information, information of fluorescence intensity changes of a plurality of polycyclic aromatic hydrocarbons along with time is increased to form a multi-dimensional fluorescence spectrum, a spectrum overlap problem can be effectively overcome, and component analysis of the polycyclic aromatic hydrocarbons in an inclusion can be achieved in a real sense, thus determining the critical petroleum inclusion reservoir forming moment and overcoming a problem that extraction of organic components in single petroleum inclusion is difficult and other problems.

The method comprises: 1) respectively connecting each service port of each network device into one path; connecting each path to get a test path; connecting the service ports at both ends of the test path; 2)configuring the IP address of two test port of the test apparatus; setting the source IP address and destination IP address of traffic on the test apparatus; according to the connection sequence, 3) one-to-one binding each service port of each network device with each VPN of relevant network device; according to the VPN type configured on each network device, making further configuration; starting the test apparatus to send the traffic to the make test.

The invention discloses a novel use of a porphyrin modified electrode. The porphyrin modified electrode is obtained by decorating porphyrin and alpha-Al2O3 onto a glassy carbon electrode, and the porphyrin modified electrode is used for simultaneously detecting ascorbic acid, dopamine and uric acid. Since a porphyrin ring forms a pi-pi conjugated structure with aromatic nucleus in the ascorbic acid, the dopamine and the uric acid, and pyridine N of the porphyrin ring can form a hydrogen bond with hydroxyl and amino in the ascorbic acid, the dopamine and the uric acid, so that the porphyrin modified electrode has excellent chemical catalytic effect on the three micromolecule substances, the micromolecule substances can be rapidly and accurately separated in the cyclic voltammetry, and the porphyrin modified electrode can be used for simultaneously detecting the ascorbic acid, the dopamine and the uric acid and can detect simply, sensitively and rapidly.

The invention belongs to the technical field of natural language processing, and particularly relates to an entity relation joint extraction method and system based on relation guidance, sentences in a target text are encoded, and sentence vector representation in the target text is obtained; for the sentence vector representation, a relation extraction module is used to extract a relation type contained in the target text; the extracted relation type is used as priori knowledge to be fused with word vector representation in a target text sentence, and an entity recognition module is used for recognizing an entity corresponding to the extracted relation type in the target text. Attention to irrelevant entities can be reduced, extraction of redundant entities is avoided, then entity pairs corresponding to multiple identified relation types are identified, the problem of entity overlapping is solved, all entity relation triples contained in sentences are extracted finally, entity relation identification accuracy is improved, and actual scene application is facilitated.

The invention discloses a map data display method and apparatus, which relates to the field of data processing and is invented for radically solving the problem of overlapping of data display regions. The method comprises the steps of determining a local region needed to be amplified in a map graph in response to an amplification operation instruction, wherein the map graph consists of a plurality of terrain graphs, a data display region is correspondingly displayed on each terrain graph, and the local region at least covers the data display regions of the two terrain graphs; searching for the data display region with the highest overlapping degree in the local region to obtain the target data display region; calculating an amplification multiple of the local region according to region parameters of the target data display region; and performing amplification display on the local region according to the amplification multiple, thereby enabling the data display regions in the local region not to be overlapped mutually. The method and the apparatus are mainly applied to a client side.

The invention discloses a distance measuring system, which comprises an emitter, a collector and a processing circuit, wherein the emitter and the collector are arranged along a baseline, the emittercomprises a light source array composed of a plurality of light sources, the light source array comprises a plurality of sub light source arrays, the sub light source arrays are configured to be started one by one in the baseline direction to emit spot light beams, the collector comprises a pixel array composed of a plurality of pixels, the pixel array comprises a plurality of sub-pixel arrays, the plurality of sub-pixel arrays are configured to collect photons in a light beam reflected back by the spot light beam through a to-be-measured target object and form a photon signal, the processingcircuit comprises a plurality of sub-processing circuits, and the sub-processing circuits are connected with the sub-pixel arrays in a one-to-one correspondence mode so as to control pixels in the sub-pixel arrays to be started, collect photons in the reflected light beams and calculate flight time. According to the invention, the spatial resolution is improved, the problem of superpixel overlapping is solved, and the distance measurement precision is improved; and the process cost and the complexity are effectively reduced.

The invention discloses a separation mechanism, comprising a coin channel and a separator. The separator is arranged in the coin channel and can move up and down. A gap is reserved between the separator and a bottom plate of the coin channel. The sum of the displacement height of the separator and the height of the gap is not less than the thickness of the thickest coin. A frictional force between the bottom plate of the coin channel and a coin is larger than a frictional force between the coin and the separator, and a frictional force between the coin and the separator is larger than a frictional force between coins. The invention also discloses a coin sorting machine, comprising the separation mechanism, a pre-separation mechanism and a sorting mechanism. The pre-separation mechanism comprises a first coin outlet. The first coin outlet adapts to the height of the single thickest coin. The coin channel comprises a coin inlet and a second coin outlet. The first coin outlet is communicated with the coin inlet of the coin channel. The second coin outlet adapts to the sorting mechanism. According to the invention, the coin overlapping is effectively solved.

The invention discloses an assistant spooling gear of a mooring winch. A leading screw (4) and a polished rod shaft (9) are arranged between two lateral bases; one end of the leading screw (4) is provided with a big gear wheel (7); a center hole of the big gear wheel (7) is provided with a pin shaft; the pin shaft is provided with a curved handle (5) in a dismountable mode; the outer side of the big gear wheel (7) is provided with a small gear wheel (6) in a matching mode; a center hole of the small gear wheel (6) is fixedly provided with a second pin shaft; the leading screw (4) is provided with a roller wheel rack which rotates along with the leading screw to realize reciprocation, and the upper part of the roller wheel rack is erected with a vertical roller wheel (1) wit a longitudinal axial direction; and the roller wheel rack is sleeved on the polished rod shaft (9). The assistant spooling gear is mainly used in the processes of shipping and ocean engineering mooring and anchoring winch mooring rope recycling, and mooring ropes are arranged in order on a rope roll through manpower assistance. The assistant spooling gear can effectively solve the spooling problems of extrusion, overlapping, knocking over and the like during recovering the mooring ropes because the arrangement of the mooring winch cannot automatically realize neat keeping of the mooring ropes due to limited deck area.

The invention discloses a high-spectral-imaging-based field pest monitoring and trapping device. The device comprises a pest killing device, a sampling device and a terminal machine; the sampling device is a hyper-spectrometer camera; the device also comprises a pest discharge device, a conveying device and a power device; the pest discharge device comprises a housing whose top part is provided with a pest inlet connected to the pest killing device and bottom is provided with a pest outlet; a hollow mandrel which extends into the housing; a pest discharge wheel which is rotationally installed on the hollow mandrel and whose periphery is provided with pest discharge holes which are communicated with the inner chamber of the pest discharge wheel, and the inner chamber of the pest discharge wheel is communicated with a negative device through the hollow mandrel; a absorption stop valve which is fixedly mounted on the hollow mandrel and can seal pest discharge holes that are going to discharge pest. The device can solve the problems of inaccurate counting and variety identification due to overlapped pest bodies; by means of the device, not only the number of pest can be counted, but also the varieties, sex and instar of the pest can be identified, which can predict and discriminate field pest situation more accurately.

The invention discloses an LTCC (Low Temperature Co-Fired Ceramic) miniaturization microwave passive device. The LTCC miniaturization microwave passive device comprises dielectric layers and metal layers which are arranged between the dielectric layers; the dielectric layers comprise a first micro-strip dielectric layer, a second micro-strip dielectric layer, a third micro-strip dielectric layer, a first strip line dielectric layer, a second strip line dielectric layer, a third strip line dielectric layer, a fourth strip line dielectric layer, a fifth strip line dielectric layer and a sixth strip line dielectric layer which are arranged from top to bottom in turn; the metal layers comprise a micro-strip line, a first grounding plane, a strip line and a second grounding plane; the micro-strip line is arranged on the first micro-strip dielectric layer; the first grounding plane is arranged between the third micro-strip dielectric layer and the first strip line dielectric layer; the strip line is arranged between the third strip line dielectric layer and the fourth strip line dielectric layer; the second grounding plane is arranged below the sixth strip line dielectric layer. According to the LTCC miniaturization microwave passive device, the multilayer wiring scheme is adopted and accordingly the circuit size and weight is reduced under the same working frequency, the problem of transmission line overlapping is effectively solved, parasitic parameters are reduced, the structure is simple, and the performance is good.

Provided is a crop-weed identification and positioning method based on three-dimensional flight time method imaging. When plant seedling stalk characteristics are obvious, a three-dimensional flight time camera is adopted to obtain side view three-dimensional images of plant seedlings; and when plant seedling leaf blade characteristics are obvious, the three-dimensional flight time camera is adopted to obtain overlook three-dimensional images of the plant seedlings. Then three-dimensional soil partition pretreatment is respectively performed on the side view three-dimensional images and the overlook three-dimensional images, a Hough conversion algorithm and a tube characteristic measuring algorithm are adopted to extract stalk characteristics for the plant seedlings with the obvious stalk characteristics; for the plant seedlings with the obvious leaves, a three-dimensional image local area increase algorithm or a continuous method for inspecting local surface normal direction changes is adopted to extract leaf blade proceeding characteristics; weeds are identified and positioned on the basis of characteristic extraction, and rapid and automatic weeding operation can be performed by matching with a weeding device; simultaneously plant stalk and seedling center positions undergo three-dimensional reconstruction and positioning, accurate positioning of the crop plant seedlings is performed, crop growth data such as planting distances and seeding numbers of crops are obtained automatically, and dynamic automatic accurate management of crop growth is facilitated.