36results about How to "Reduce matching calculations" patented technology

The invention relates to an image matching method, comprising the following steps: (1) collecting images to be matched by an image collector and extracting template images from an image processor; (2) reducing the images to be matched and template images by half in the directions of width and height; (3) repeating step (2) until the images to be matched and template images reach specified sizes and carrying out gray scale normalization cross correlation matching on the reduced images to obtain the optimal matching positions (x, y); (4) returning to the previous images which are reduced by half and carrying out further matching on 9 search positions within the range of ((2x-1)-(2x+1), (2y-1)-(2y+1)) to obtain new optimal matching positions (x, y); (5) repeating step (2) until returning to the primary image, namely the original image, and carry out final matching to obtain the optimal matching positions and matching values. The method features shorter matching time and smaller amount of calculation.

The invention discloses a speckle-based three-dimensional-face reconstruction method and device. The method and the device can reduce matching calculation quantity while image feature richness is improved, and realize synchronized improvement of three-dimensional-face reconstruction efficiency and reconstruction precision. The method comprises: inputting a plurality of speckle image pairs conforming to binocular stereo vision; carrying out polar line correction on the input speckle image pairs to enable corresponding matching points to be at the same horizontal lines; carrying out face detection on the speckle image pairs after polar line correction, and extracting region-of-interest image pairs of to-be-reconstructed faces; carrying out first matching processing on the region-of-interestimage pairs to obtain first parallax graphs; setting a matching point searching range according to parallax in the first parallax graphs, and carrying out second matching processing on the region-of-interest image pairs to obtain second parallax graphs; and generating corresponding three-dimensional point cloud data according to the second parallax graphs and camera calibration parameters to obtain the reconstructed three-dimensional faces.

The invention provides a rapid fuzzy matching algorithm for strings. According to the invention, firstly data preprocessing is performed on texts in a database to obtain a statistical model and an index is established via Hash. An input text is a shorter string. The algorithm traverses all Chinese characters therein, activates the positions of corresponding Chinese characters in a finite character complete set, and maps the activation state of the finite character complete set to each tag to filter tags. A few filtered tags are used for matching the texts and the DTW algorithm is used for approximate string matching. The algorithm also comprises the steps of performing scoring and sorting according to the result of the degree of approximation of matching and returning to a search result. Through the efficient tag filtering method, the calculation efficiency of the string matching algorithm is greatly increased; in a process of input text matching, a fuzzy matching effect is achieved and a good matching performance is guaranteed for fuzzy languages.

The invention provides a colorful animal image retrieval method based on content and a colorful animal image retrieval system based on content. The method comprises the following steps: step 1, inputting an inquiry image which serves as a retrieval object; step 2, extracting the color, grain and framework characteristic descriptions of interested content in the inquiry image based on the guidance of a characteristic template library, and performing normalization processing, wherein the characteristic descriptions, which are different in animal color, grain and framework and are formed on the basis of sample statistics, are stored in the characteristic template library; step 3, performing retrieval based on an R-tree index structure on the characteristic descriptions extracted in the step 2, performing search calculation on image characteristic indexes in a characteristic index library, and extracting a corresponding image from an image database according to the matching result. The method and the system can be used for overcoming the influence of illumination changes on image retrieval judgment to a certain extent, describing the angle and posture semantic information of an inquiry animal according to the framework characteristic analysis during retrieval, and generating a retrieval result consistent with the human visual perception.

The invention provides an indoor positioning method and device. The method comprises that a present fingerprint positioning point is determined by obtaining at least one grid point most similar to the fingerprint, and determining the fingerprint positioning point among the at least one grid point; a present inertial navigation vector of a terminal is obtained; and a present comprehensive positioning point is obtained based on the present inertial navigation vector of the terminal and the present fingerprint positioning point. Beacon nodes are ordered according to the intensity of receives signals during positioning calculation, fingerprint is coupled at sampling points near the N beacon nodes of the highest intensity, and the calculation amount of coupling is reduced. Preset inertial navigation is combined with the fingerprint positioning result in terminal equipment to improve the user experience.

Embodiments of the invention provide methods for matching text information and pushing business objects, and devices for matching text information and pushing business objects. The method for matching text information comprises: obtaining a to-be-matched first text information set and a to-be-matched second text information set, the first text information set comprising first text information in limited number, the second text information set comprising second text information in limited number; and according to preset rules, inquiring one or more piece of information in the second text information in the limited number, the one or more piece of information being matched with each piece of information in the first text information in the limited number. An open extension thinking of directly finding extended words from the first text information is given up, and the method turns to a closed interval, to search one or more piece of information in the second text information in the limited number, the one or more piece of information being matched with each piece of information in the first text information in the limited number, so as to save unnecessary matching calculation amount, reduce system resource waste, and improve efficiency of matching calculation.

The invention belongs to the technical field of image detection and relates to a method for tracing a covariance matrix based on grayscale restraint, which comprises the following steps: selecting a rectangular area of a image as a target model to trace; calculating a grayscale value of each point of the target model; extracting the characteristic vector of each point of the target model; calculating the covariance matrix of the target model; after the length and the width of a subsequent frame are respectively amplified by taking the target model as a center, obtaining a tracing window, selecting a candidate target and checking whether the candidate target meets the grayscale restraint or not; and calculating a difference value of the covariance matrix of the candidate target which meetsthe grayscale restraint and the target model, wherein the candidate target area with the minimal difference value is the position of a traced target. The invention has the advantages of more accuratetarget positioning, higher tracing speed and higher instantaneity.

The invention discloses a method for splicing videos in real time and belongs to technical field of video image processing. The method computes the overlapped area of video frame images to be spliced by using a phase correlation method, improves a SURF (Speeded UP Robust Features) algorithm, simplifies the generation process of a characteristic point descriptor in the SURF algorithm, reduces the dimensionality of the descriptor, extracts the characteristic points of the overlapped area of the video frames by using the improved SURF algorithm. The invention provides an image registration method based on characteristic block matching to match the characteristic points, decrease calculated amount, increase computational efficiency, and quickly solve an image conversion model by using the characteristic block matching. Finally the invention provides a projection matrix updating method based on a correlation coefficient. The projection matrix updating method is used for updating a projection matrix, prevents erroneous splicing results, and fusing the video frame images so as to splice the videos in real time.

The embodiment of the invention provides an image retrieval method and device. According to the method, by obtaining an image to be retrieved, text information contained in the image to be retrieved is obtained accordingly; furthermore, the text information is utilized for conducting the retrieval so that a retrieval result matched with the image to be retrieved can be obtained. Therefore, by means of the technical scheme of the method and device, the complexity of image retrieval can be reduced, and the image retrieval efficiency can be raised.

The embodiment of the invention discloses a video data processing method and apparatus. The method comprises the following steps: performing frame rate change on source video data to obtain target video data, and obtaining SIFT features corresponding to each frame of data in the target video data; performing redundant processing and two-dimensional transformation processing on the SIFT features to obtain a video fingerprint feature vector corresponding to the target video data; searching pre-stored video data having the highest matching degree with the target video data from a preset video fingerprint feature matching library according to the video fingerprint feature vector corresponding to the target video data, and when the matching degree between the searched pre-stored video data and the target video data is greater than a preset matching degree threshold, determining the source video data as illegal video data. By adoption of the video data processing method and apparatus, the anti-noise ability of the video fingerprint can be improved.

The invention discloses a security policy matching method and device, network equipment and a storage medium, and the method comprises the steps: receiving a security policy matching request which carries the feature information of a target message; and determining target security policy information matched with the target message according to the security policy chain table and the feature information, wherein the security policy chain table comprises a transverse chain table and a longitudinal chain table cascaded with the transverse chain table, the transverse chain table comprises at leastone quintuple node, and the longitudinal chain comprises at least one policy information node. According to the method and the device, the plurality of linked lists can be generated by extracting thekeywords in the security policy, then association is performed according to the correlation among the linked lists, and finally the security policy is selected based on the linked lists, so the matching speed can be increased and the matching process of irrelevant policies can be skipped. And on the other hand, the matched result is copied to the standby linked list, so the standby linked list can be called again.

The invention relates to an artificial intelligence technology, and discloses a road matching method, which comprises the steps of segmenting GPS data to obtain segmented GPS data; encoding the segmented GPS data to obtain encoded GPS data; carrying out associated segmentation on the map data by utilizing the coded GPS data to obtain segmented map data; performing first screening analysis calculation by utilizing the segmented map data and the segmented GPS data to obtain a target longitude and latitude point set and a candidate starting point set; performing second screening analysis calculation by utilizing the segmented map data and the target longitude and latitude point set to obtain a candidate end point set; and performing calculation according to the candidate starting point set and the candidate terminal point set to obtain a target path. The invention also relates to a blockchain technology, and the segmented GPS data can be stored in the blockchain. The invention further provides a road matching device, electronic equipment and a computer storage medium. According to the method, computing resources required by road matching are saved.

The invention discloses a rapid pose estimation algorithm for a dot-line feature visual SLAM system. The algorithm comprises the following steps: 1, carrying out the extraction of dot-line features ofsub-threads and the grid distribution of the features, and enabling the line features to be distributed to a pre-defined image frame grid according to the midpoint position; 2, carrying out sub-thread projection matching on the map points and the map line midpoints, and rejecting line features with mismatched line features; and 3, performing a pose optimization algorithm of map points and map line midpoints based on the step 2. The pose estimation algorithm provided by the invention is an algorithm used when a dot-line feature SLAM system works normally, and original complex four-degree-of-freedom line features are simplified by introducing point information in the line features, so that the efficiency of the dot-line feature pose estimation problem is greatly improved.

The invention relates to the field of road pavement detection, in particular to a smart city road pavement detection method and system based on computer vision. The method comprises the steps of acquiring a reference image and a comparison image of a to-be-detected pavement area, and pre-processing the reference image and the comparison image; performing edge detection on the reference image to obtain each edge region of the reference image; and based on each edge region of the reference image, matching the pixel points of the reference image and the comparison image, and determining the flatness condition of the to-be-detected pavement according to the matching result of the pixel points. According to the invention, the real-time detection of the flatness of the road pavement is realized, the accuracy of road detection is improved, and the time used for road pavement detection is shortened at the same time.

The invention provides a visual positioning navigation method, a visual positioning navigation device and a terminal. The visual positioning navigation method comprises the steps of: establishing a critical path based on critical locations and route information between the critical locations, and acquiring critical location contour features associated with the critical location; acquiring surrounding map information of surrounding locations of the critical path, wherein the surrounding map information comprises surrounding location contour features and displacement information between the critical locations; and determining the critical location adjacent to a current location based on critical location contour features when performing positioning navigation, and determining position information of a current location according to the surrounding map information and the critical locations. The visual positioning navigation method, the visual positioning navigation device and the terminalcan perform pre-learn training of manual intervention of the critical locations and paths, effectively reduce the global graphic feature matching calculation amount, improve the recognition efficiency and reliability, and provide users with precise and low-cost indoor positioning navigation experience.

The invention discloses a P2P energy transaction method based on a block chain and a multi-objective decision, and the method comprises the following steps: S1, enabling an electricity selling / purchasing user to submit electricity selling / purchasing information and an electricity selling / purchasing matching requirement on the block chain, generating a record, and saving the record by chains; S2, after a transaction matching settlement contract in the period is triggered, reading the latest electricity selling record from the blockchain and matching transactions S3, generating a transaction pre-matching record and performing uplink storage; S4, enabling an electricity selling user to acquire the transaction pre-matching record from the blockchain and authorize a transaction pre-matching record, and generating and saving an electric power transaction matching record by chains; and S5, enabling an electricity purchasing user to acquire the transaction pre-matching record and an electricity transaction matching record with the electricity selling user private key signature from the block chain, complete transaction account transfer, generate an account transfer record, and upload and save the account transfer record by chains. According to the matching method based on the multi-objective decision, an accurate utility function calculation method is provided, and a corresponding multi-attribute combination method is designed according to the relationship among the attributes, so that matching satisfaction of a transaction subject can be improved.

The invention belongs to the technical field of machine vision template matching online positioning, and particularly relates to a multi-template matching real-time positioning method based on improved 2D-ICP. Firstly, an industrial camera is installed and adjusted on a machining platform, a template workpiece is placed, a template image is collected, and a template database is made off line and stored; secondly, collecting images in real time, performing rough matching on the to-be-searched images by adopting a multi-template matching method and OpenMP parallel programming, and then performing accurate matching through a 2D-ICP algorithm of an improved matching calculation function to obtain pose information of target sub-pixel precision; calculating the actual coordinate and angle of the target object on the processing platform according to the actual installation pose and pixel equivalent of the camera; according to the method, the matching process of any plane workpiece template can be realized, the workpiece pose can be determined, and the method has the advantages of rapidness, accuracy and stability.

Embodiments of the invention provide a target detection method, a device, equipment and a readable storage medium. The method comprises the steps of obtaining center point data of a to-be-detected target in a pixel coordinate system and point cloud data of the to-be-detected target in a camera coordinate system; obtaining conversion data of the central point data in a camera coordinate system; determining a point set associated with the center point data under a camera coordinate system according to the conversion data and the center point data; generating a corresponding three-dimensional anchor frame by taking each point in the point set as a central point; and performing target detection on the point cloud data falling in the three-dimensional anchor point frame. According to the methodprovided by the embodiment of the invention, the generation of the three-dimensional anchor point boxes only in part of the camera space is limited, so that the number of the generated three-dimensional anchor point boxes is reduced, the matching calculation amount of the later anchor point boxes and the real target detection box is reduced, the matching speed is increased, and the target detection efficiency is improved.

The invention discloses a data-processing method and an electronic device. The data-processing method comprises steps of: acquiring a first image and a second image of a target object by using a first image acquisition unit and a second image acquisition unit respectively; dividing the first image into N subareas, wherein N is greater than or equal to 2; acquiring the pixel at the mass center of each subarea of the first image, matching the pixels at the mass centers of the subareas of the first image with pixels within a first predetermined area of the second image to obtain an operation result; according to the operation result, using the pixel satisfying a predetermined condition as a pixel matching the pixel at the mass center so as to obtain a matching result.

The present invention provides a visual positioning and navigation method, device, and terminal, wherein the method includes: establishing a key path based on key points and route information between key points, and obtaining key point contour features associated with key points; obtaining key paths The surrounding map information of the surrounding locations, including: the outline features of the surrounding locations, and the displacement information between the key locations; during positioning and navigation, the key locations adjacent to the current location are determined based on the outline features of the key locations, and the key locations are determined according to the surrounding map information and key locations. Location determines the location information for the current location. The method, device and terminal of the present invention can carry out manual intervention pre-learning training on key locations and paths, effectively reduce the amount of global graphic feature matching calculations, improve recognition efficiency and reliability, and provide users with accurate and low-cost indoor positioning navigation experience.

The invention provides a rapid fuzzy matching algorithm for strings. According to the invention, firstly data preprocessing is performed on texts in a database to obtain a statistical model and an index is established via Hash. An input text is a shorter string. The algorithm traverses all Chinese characters therein, activates the positions of corresponding Chinese characters in a finite character complete set, and maps the activation state of the finite character complete set to each tag to filter tags. A few filtered tags are used for matching the texts and the DTW algorithm is used for approximate string matching. The algorithm also comprises the steps of performing scoring and sorting according to the result of the degree of approximation of matching and returning to a search result. Through the efficient tag filtering method, the calculation efficiency of the string matching algorithm is greatly increased; in a process of input text matching, a fuzzy matching effect is achieved and a good matching performance is guaranteed for fuzzy languages.

The invention discloses a speckle-based three-dimensional human face reconstruction method and equipment, which can increase the richness of image features and at the same time reduce the amount of matching calculations, and achieve synchronous improvement in three-dimensional human face reconstruction efficiency and reconstruction accuracy. The method includes: inputting a plurality of speckle image pairs conforming to binocular stereo vision; performing epipolar correction on the input speckle image pairs so that corresponding matching points are on the same horizontal line; Face detection, extracting the image pair of the area of ​​interest of the face to be reconstructed; performing the first matching process on the image pair of the area of ​​interest to obtain the first disparity map; setting the matching point search range according to the disparity in the first disparity map, The region image is subjected to a second matching process to obtain a second disparity map; corresponding 3D point cloud data is generated according to the second disparity map and camera calibration parameters to obtain a reconstructed 3D face.

The present invention provides a content-based color animal image retrieval method and system. The method includes the following steps: Step 1, inputting a query image as a retrieval target; The color, texture and skeleton feature description of the content, and perform normalization processing, wherein the aforementioned feature template library stores the feature descriptions of different animal colors, textures and skeletons formed based on sample statistics; and step 3, extracting from step 2 The feature description searches based on the R-tree index structure, searches and calculates with the image feature index in the feature index library, and extracts the corresponding picture from the image database according to the matching result. Utilizing the present invention can overcome the influence of illumination changes on image retrieval judgment to a certain extent, and describe the angle and posture semantic information of query animals through skeleton feature analysis during retrieval, and generate retrieval results that are consistent with human visual perception .

The invention discloses a method and system for obtaining image information. The method includes the steps that S1, an image sensor is controlled to expose a whole exposure area, a total graph search is carried out on exposed image data to obtain a target point, a window range containing the target point is set, and a window position is recorded; S2, the image sensor is controlled to expose the window position area recorded in the exposure area to extract the exposed image data; S3, feature extraction is carried out on the data extracted in the S2, whether the target point exists or not is calculated in a matching mode, if the target point does not exist, jumping to the S1 is performed, and if the target point exists, the S4 is performed; S4, whether the target point and a window meet a preset positional relationship or not is judged, if the target point and the window do not meet the preset positional relationship, the window position is adjusted according to a preset adjustment strategy, moreover, a window position record is updated, and returning to the S2 is performed; when a suspending signal is detected, the process is finished. By the mode that target predication and the image sensor are combined to perform partial exposure to obtain the data in the prediction range, data obtaining quantity and image matching calculated quantity are reduced, and speed and accuracy for tracking a target are improved.

The invention relates to the field of road pavement detection, in particular to a computer vision-based smart city road pavement detection method and system. Obtain the reference image and the comparison image of the road surface area to be detected, and preprocess them; perform edge detection on the reference image to obtain each edge area of ​​the reference image; based on each edge area of ​​the reference image, pixel Points are matched, and the smoothness of the road surface to be detected can be determined according to the matching results of the pixel points. The invention realizes the real-time detection of the smoothness of the road surface, improves the accuracy of the road detection, and shortens the time used for the road surface detection.

The invention relates to the field of artificial intelligence, and discloses a neural network-based online appointment registration method and device, and the method comprises the steps: receiving an online registration request inputted by a user, and collecting the registration information of the online registration request, the registration information comprising user information, illness state information and the registration time; performing information authentication on the user information; when the information authentication succeeds, acquiring a face image of the user, and performing face authentication on the face image; when the face authentication is successful, performing feature extraction on the illness state information to obtain feature illness state information; and according to the registration time, matching a correlation degree between the characteristic illness state information and medical personnel in a medical library, and selecting the medical personnel of which the correlation degree is greater than a preset threshold value as a target registration object of the online registration request. In addition, the invention also provides a neural network-based online appointment registration device, electronic equipment and a storage medium. According to the invention, the information consistency of online appointment registration can be ensured, the reliability of online appointment registration is improved, and the phenomenon of medical resource waste is avoided.

The embodiment of the invention discloses a video data processing method and apparatus. The method comprises the following steps: performing frame rate change on source video data to obtain target video data, and obtaining SIFT features corresponding to each frame of data in the target video data; performing redundant processing and two-dimensional transformation processing on the SIFT features to obtain a video fingerprint feature vector corresponding to the target video data; searching pre-stored video data having the highest matching degree with the target video data from a preset video fingerprint feature matching library according to the video fingerprint feature vector corresponding to the target video data, and when the matching degree between the searched pre-stored video data and the target video data is greater than a preset matching degree threshold, determining the source video data as illegal video data. By adoption of the video data processing method and apparatus, the anti-noise ability of the video fingerprint can be improved.

The invention discloses a distance measurement method based on image matching, a storage medium and electronic equipment. The method comprises the following steps: S1, collecting an image of a to-be-measured target; s2, performing three-dimensional correction on the acquired image according to the three-dimensional correction matrix; s3, extracting edge features of the image after three-dimensional correction; s4, performing stereo matching by using a semi-global stereo matching algorithm to obtain a sparse disparity map; s5, adaptive sampling is carried out on the sparse disparity map, and a dense disparity map is recovered; and S6, solving the distance between the detection point and the to-be-detected target according to the dense disparity map. According to the scheme, the edge features of the to-be-detected target are extracted and then stereo matching is performed, so that the matching calculation amount is greatly reduced, and the real-time performance of stereo matching is improved. And meanwhile, the sparse disparity map is recovered into a dense disparity map by adopting a self-adaptive sampling algorithm, so that the integrity and the accuracy of distance measurement are ensured.