28 results about "Image parsing" patented technology

The application belongs to the technical field of computer vision and intelligent control, and discloses a load device cooperative control system and method based on visual identification, which comprises a data acquisition module, an industrial camera is arranged in a work area, multi-view image acquisition of a load device and a work target is carried out, visual image data is acquired, time sequencing is carried out, and work visual data flow is formed; a pose analysis module is used for image analysis of the work visual data flow, extraction of load device structure characteristics, calculation of the spatial pose of each load device in combination with camera calibration parameters, determination of the work target position, generation of load device pose information; a shielding modeling module is used for equivalent of the geometric shape of each load device to a space bounding box changing with time according to the load device pose information, determination of the position and scale of the bounding box based on preset size parameters, formation of a space shielding body set, and improvement of the efficiency and safety of multi-load device cooperative work.

This invention relates to the field of flexible laser processing, specifically providing a collaborative control method, system, terminal, and medium for flexible laser processing based on image analysis and intelligent scheduling. The method includes: receiving a conventional task instruction containing known parameters and an image file to be processed; performing image processing and path planning on the image file to extract the process attribute parameters of the image processing task; merging the parameters of the two types of tasks to construct a multi-objective model with the optimization objectives of minimizing production cycle, equipment load imbalance, total energy consumption, and total tool changes; solving the model using an improved genetic algorithm to output the optimal task allocation and sorting scheme; and controlling the laser equipment to execute tasks according to the scheduling scheme, supporting automatic tool changing and real-time monitoring. This invention improves the intelligence level, production efficiency, and resource utilization of flexible laser processing systems.

The image analysis device (1) includes: an image holding unit (8) that holds an image; a learned model registration unit (10) configured to register a learned model created by machine learning; a learned model holding unit (12) that holds a learned model registered by the learned model registration unit (10); an algorithm holding unit (14) that holds a plurality of analysis algorithms for performing analysis processing of an image; a process creation unit (18) configured to create, for an analysis target image arbitrarily selected from the image held by the image holding unit (8), an analysis process for performing analysis of the analysis target image by combining a learned model arbitrarily selected from the learned models held by the learned model holding unit (10) and an analysis algorithm arbitrarily selected from the analysis algorithms held by the algorithm holding unit (14); and an analysis execution unit (20) configured to perform analysis of the analysis target image based on the analysis process created by the process creation unit (18).

This invention relates to the field of image data review technology, and in particular to a method and system for intelligent verification of fake images based on multimodal data. The method includes acquiring the image to be verified; using a multimodal distillation model to perform full-text recognition on the image to obtain text data; performing template recognition processing based on the text data, marking relevant images, and outputting template recognition results; using a large language fine-tuning model to classify the text data into image types, extract key information, and perform anomaly verification, outputting anomaly results; calculating a risk score and generating a risk analysis report; and using the risk analysis report to verify fake images. This invention's method, through modular design, forms a continuous automatic processing chain for image parsing, template recognition, information extraction, and rule verification, reducing manual switching between multiple systems and repetitive processing steps, thereby reducing the intensity of manual intervention and management costs.

This application discloses an asset identification method and terminal based on video spatial gridding. It uses a video spatial gridding mechanism to discretize the registered geographic video stream into grid units with geocoded mappings, and relies on multimodal feature extraction to interpret and separate asset categories and precise contours. This approach effectively solves the problems of spatial positioning errors and unstable recognition of dynamic objects across edges in traditional image analysis. The coupling between multimodal features enhances the anti-interference capability of the identification, and the coded integration realizes a structured association between asset attributes and spatial location. Ultimately, it not only significantly improves the completeness and consistency of asset extraction, eliminating boundary truncation and contour drift, but also provides digital records with semantic attributes and high-precision spatial coordinates for large-scale asset inventory and intelligent operation and maintenance monitoring.

The application discloses a kind of unified stream processing methods, systems, equipment, medium and program product of multimodal AI interaction content, comprising: receiving the request containing image and text sent by user, constructs request context;Image features are extracted;Image analysis is carried out and image description text is output;Intention recognition is carried out and intention is output;Multi-stage inference is carried out and thinking content segment is generated in stream;Buffering is carried out and thinking content segment is published;Text content segment is generated based on multi-stage inference;Buffering is carried out and text content segment is published;Recommendation trigger is carried out based on image description text and intention;Recommendation content segment;Insert recommended position, buffer and publish;It also includes process event management, publishes event notification when starting, any step fails and ends.The application is a kind of general method that can transmit different types of output in single ordered stream, can reduce analysis cost, improve interaction experience and expandability.

This invention discloses an electrical performance testing device for flexible conductive sponge, belonging to the field of electrical performance testing technology. The feature construction module performs image analysis of basic physical property elements and conductive feature elements based on an electrical performance visual sample library, outputting pixel-level semantic segmentation results and corresponding proportion statistics for each basic physical property element. It also integrates the analysis results of the two types of elements to construct a visual feature expression system. The intelligent prediction module extracts representative image samples based on the electrical performance visual sample library, inputs the visual feature expression system of the image samples into a regression model, and the regression model outputs an electrical performance score, completing the evaluation of the electrical performance of the flexible conductive sponge. This testing device optimizes the entire process from visual data acquisition to intelligent electrical performance evaluation, improving detection efficiency and real-time performance.

The application relates to a chemical object structured recognition method and system based on image analysis, and relates to the technical field of chemical structure recognition. The method comprises the following steps: acquiring image data sets of chemical objects; constructing an image analysis model comprising a backbone network, an adapter hybrid encoder, a query selection module and a decoder; the backbone network is used for feature extraction to obtain backbone features; the adapter hybrid encoder is used for feature coding to obtain fusion coding features; the query selection module uses an uncertainty minimization query selection strategy to perform feature query to obtain an object query set; the decoder inputs a feedforward neural network after multi-layer iterative optimization of the object query set, and outputs a recognition result; and a chemical object image to be recognized is input into the trained image analysis model to obtain a recognition result. The application completes the recognition and structured reconstruction of multiple types of structure elements in a unified framework, and improves the recognition robustness in a complex image scene.

The application provides a glue dropping method and device of an electric core and an electronic equipment, which are applied to a glue dropping mechanism and include the following steps: a mapping relationship between an image coordinate system of an image sensing mechanism and a mechanical coordinate system of the glue dropping mechanism is established; a side skirt image of the electric core is captured by the image sensing mechanism, and a plurality of target skirt positioning points and image coordinates corresponding to each target skirt positioning point are obtained by analysis; for each target skirt positioning point, a mechanical coordinate corresponding to the target skirt positioning point is determined according to the image coordinate corresponding to the target skirt positioning point and the mapping relationship, so as to complete a point glue operation on the target skirt positioning point. According to the application, the track of the electric core arc side is recognized, and the mapping relationship between the coordinate systems is combined, so that the point glue head can complete the point glue operation according to the position information of each point on the electric core arc track, and the point glue precision and the point glue efficiency of the electric core are improved.

The present application provides an ultrasonic diagnostic apparatus and a diagnostic assistance method. An image analysis section (28) has a reliability calculation section (29) that calculates a reliability that indicates a likelihood that a lesion candidate is a lesion. A marker display control section (30) displays a marker that notifies of the lesion candidate on an ultrasonic image. The marker is continuously displayed in a continuous detection state in which the reliability continuously satisfies a candidate detection condition. At this time, from a time point at which the candidate detection condition is first satisfied to a passage of a shape-invariant period, the shape of the marker is fixed, and thereafter, the shape of the marker is changed in correspondence with a change in the reliability.

This invention discloses an intelligent construction and parameter optimization system for building BIM models based on AI image analysis. It performs semantic segmentation and component recognition on input building images through image analysis and component mapping, extracts the geometric and type features of each component region, and matches them with a pre-set BIM component family library to establish a mapping relationship between image components and parameterized components. Based on this mapping relationship, it calls the BIM software interface to automatically complete the component layout and generate a 3D model. It then performs quantity and multi-dimensional parameter analysis on the model, generates optimization strategies under constraints, and adjusts the model. For unmatched components, it extracts data and generates parameters to expand and update the component family. This invention improves the conversion efficiency from image to BIM model, enhances the collaborative capabilities of model construction and parameter analysis, and improves the adaptability and intelligence level of the component library.

This invention provides a trip planning method, electronic device, and storage medium based on a multimodal large model. The method includes: acquiring text and image content specified by a target user from a network platform; inputting the text and image content into a text and image parsing model to perform joint semantic understanding of the text and image information in the content, extracting structured trip data containing location information; inputting the structured trip data into a location verification model to standardize the location information by calling a map service interface, obtaining standard location data; inputting the standard location data into a path planning model to calculate the path between various locations by calling a navigation service interface, generating executable trip data containing a sequence of waypoints and navigation instructions; and returning the executable trip data to the target user. This invention can generate navigable trip planning data using text and image content from a network platform.

The application discloses a knitted structure intelligent generation method and system based on pattern features, and relates to the technical field of knitted structure generation.The method comprises the following steps: acquiring a pattern image corresponding to a to-be-generated knitted structure, and performing image analysis processing of the pattern image in the weaving direction; constructing a pattern feature field for describing the spatial distribution relationship of the pattern under the weaving semantics; identifying a feature area in the pattern feature that does not satisfy a preset knitted structure generation constraint, and establishing a corresponding structure constraint mark; constructing a knitted structure constraint space according to the pattern feature field and the structure constraint mark; performing knitted structure calculation, and establishing a target knitted structure.The application solves the technical problems that the existing technology has poor adaptability between pattern features and knitted structures, which leads to low structure generation efficiency and insufficient weaving feasibility of finished products, and achieves precise adaptation and intelligent calculation between pattern features and knitted structures, thereby improving the pattern restoration degree and weaving feasibility of the target knitted structure.

This application provides a method for extracting images from an Excel file and an electronic device. The method, applied to an electronic device, includes: acquiring an Excel file and extracting all original images from the Excel file; parsing all original images in the Excel file and determining the position information of each original image, the position information describing the display position of the original image in the Excel file; and identifying abnormal original images based on the coordinate position information of each original image, the abnormal original images including original images appearing repeatedly in different cells and multiple original images existing in the same cell. According to the method of this application, extracting original images from an Excel file avoids image distortion and increases the accuracy of image extraction from Excel files.

The application provides a method and device for processing images of a drone formation, a readable storage medium and a user terminal. The method comprises: acquiring an image of a drone formation taken by a user terminal; analyzing the image of the drone formation and matching the analysis result with a preset drone formation library; and triggering a preset processing action according to the matching result. The method provided by the application can enhance the interaction between a user and a drone flight show and provide better user experience.

The application discloses an image information extraction method and device, and relates to the technical field of image processing. A specific embodiment of the method comprises the following steps: obtaining text information in an image through an image analysis interface to create a context; obtaining extraction configuration through a configuration engine to store in the context; finding a field corresponding to the extraction configuration from parameters to store as an output result; performing result verification to verify whether key information in the output result is empty, and if the key information is empty, providing an alarm. The embodiment can realize a configurable image information extraction scheme, thereby greatly improving development efficiency.

Provide a system. 【Solution means】 Means for inputting voice, Means for converting the voice into text data, Means for performing natural language generation based on the text data to generate diary data, Means for inputting an image, analyzing the image, and identifying a specific event, Means for analyzing the emotion based on the voice and adding emotion information to the diary data, Means for a mobile body to patrol within a home and record voice and images in real time, Means for transmitting the generated diary data to the cloud, Means for making the diary data accessible from a remote location, A system including the above.

The application belongs to the technical field of computer vision and intelligent control, and discloses a load device cooperative control system and method based on visual identification, which comprises a data acquisition module, an industrial camera is arranged in a work area, multi-view image acquisition of a load device and a work target is carried out, visual image data is acquired, time sequencing is carried out, and work visual data flow is formed; a pose analysis module is used for image analysis of the work visual data flow, extraction of load device structure characteristics, calculation of the spatial pose of each load device in combination with camera calibration parameters, determination of the work target position, generation of load device pose information; a shielding modeling module is used for equivalent of the geometric shape of each load device to a space bounding box changing with time according to the load device pose information, determination of the position and scale of the bounding box based on preset size parameters, formation of a space shielding body set, and improvement of the efficiency and safety of multi-load device cooperative work.

This invention discloses a document parsing method, system, device, and medium, relating to the field of document image parsing technology. The method includes the following steps: obtaining an initial representation of a multi-page document image; performing layout parsing on each page of the document image based on the initial representation, and during layout parsing, using text region detection, table region detection, and image region detection to obtain the position, content, and category of element information within each page of the document image, converting the position, content, and category of the element information into vector representations, reconstructing each page of the document image using vector representations, obtaining the cross-page matching relationship of each reconstructed page of the document image, and integrating the document images based on the cross-page matching relationship to obtain the document image parsing result. This invention alleviates the problem that existing technologies cannot effectively parse multi-column document images and multi-chart document images by reconstructing each page of the document image using vector representations.

The application relates to a steel bar detection method and equipment for a fabricated shear wall edge component, and the method comprises the following steps: adopting an industrial camera to shoot steel bars of the fabricated shear wall edge component from multiple angles, obtaining shooting images, labeling steel bar regions and categories, establishing a target detection data set, training a target detection algorithm, and obtaining a steel bar recognition model; when on-site detection is performed, an auxiliary calibration plate is vertically erected at a central layer position of the fabricated shear wall edge component, a camera module is adopted to shoot and obtain detection images; sub-pixel corner point coordinates and pixel precision of the auxiliary calibration plate are analyzed, a standardized perspective transformation matrix is calculated, and the shooting angle of the images is corrected; geometric size and non-geometric size information is obtained by adopting the steel bar recognition model, and the information is compared with design information of the component to detect whether the component meets the specification. Compared with the prior art, the application has the advantages of automatic and accurate steel bar detection, high automation degree, simple operation and the like.

The application provides a fire extinguisher operation auxiliary method and system based on real-time voice guidance and optical indication, relates to the technical field of intelligent emergency assistance, and collects real-time voice, environmental light images and user position data; voice is compressed to generate a voice stream, image analysis identifies obstacle distribution, and position filtering obtains accurate coordinates. Based on the coordinates, optical equipment is controlled to form a continuous light path on the building surface, which points to the nearest fire extinguisher. At the same time, the emergency degree in the voice is identified to adaptively adjust the voice guidance intensity, and the position sequence is updated; in combination with the position and the obstacle, the user movement trend is predicted, personalized operation instructions are generated, and the voice output is realized. In addition, the visibility of the light path is adjusted through the change of the environmental light brightness, and the instructions and the light path direction are corrected according to the user voice feedback, so that real-time voice guidance and optical indication can be realized, and an intelligent and safe guidance path directly leading to the fire extinguisher is constructed for the user.

Provide a system. 【Solution means】 Means for photographing a plant using an image acquisition device, Means for recognizing the type and health condition of a plant using an image analysis device, Means for generating a care plan for the plant based on the identified plant information, Means for presenting the generated care plan to the user, Means for receiving and analyzing plant environmental sensor data using a wireless communication device, Control means for performing optimal automatic care for the plant based on the environmental data, Means for checking the health condition of the plant in real time and receiving notifications using smart devices, A system including the above.

The optical information reading method shortens the time from successfully capturing an image from which the optical information of the target can be read to outputting the reading result when parsing the captured image and reading the optical information contained in the image. In the case where the imaging unit periodically captures images and the reading routine parses the images and reads the optical information contained in the images (S12), the target determination process (220) estimates the amount of movement of the imaging unit in each frame based on images captured within a certain time range (S14). Furthermore, when a time corresponding to the certain time range has elapsed since the start of imaging of the image being parsed, the reading routine (SB, SC) is restarted to parse images captured after the image being parsed and to read the optical information contained in those images. If, based on the estimated amount of movement, it is determined that the target for imaging has been determined ("Yes" in S15), the restart is not performed, and image parsing continues.

Techniques for using deep learning to identify malicious image files are disclosed. A sample set comprising a plurality of image files is received. A first image file included in the sample set is processed, at least in part by using an image parser to extract a first set of sections of the first image file. The first set of sections includes at least one normal section. A second image file included in the sample is processed, at least in part by using the image parser to extract a second set of sections of the second image file. The second set of sections includes at least one abnormal section. A model is trained using at least a portion of the first set of sections and the second set of sections. The trained model is provided as output and is usable by a security system to determine a likelihood that a target file is malicious.

The invention relates to the technical field of image processing, and discloses an image processing method, device, equipment and medium, and the method comprises the steps: obtaining an image access request initiated by a client, and processing the request feature information carried in the image access request to obtain a request feature set; wherein the request feature set comprises equipment feature information and network environment information; performing analysis modeling on the image analysis capability of the client based on the request feature set to obtain a capability judgment result, and inputting the capability judgment result into an image delivery decision model as a format selection basis; performing matching processing on the original image through an image delivery decision model to obtain target image format data corresponding to the capability judgment result; and packaging the target image format data into an image response result, and returning the image response result to the client. The method and the device can be applied to financial science and technology or medical care service program systems, and can realize finer processing of image access and delivery processes.

The application relates to the field of electric power, and discloses a substation control device terminal strip drawing paper identification method, device and equipment, which is used for converting a terminal strip drawing paper into a standardized intermediate file. The method comprises the following steps: acquiring an electronic image of the terminal strip drawing paper, and identifying spatial coordinates of connection points, spatial coordinates of connection lines, line types, line widths and spatial coordinates of short-circuit symbols; a plurality of character sequences and corresponding minimum bounding box coordinates are identified and extracted; a semantic binding relationship between the connection points and the character sequences is established based on the spatial coordinates, the spatial adjacent relationship of the connection points is determined in combination with a terminal strip arrangement rule, and a short-circuit relationship is identified in combination with the adjacent relationship and the spatial coordinates of the short-circuit symbols; the spatial coordinates of the connection lines, the line types and the line widths, the spatial coordinates of the connection points, the semantic binding relationship and the short-circuit relationship are integrated, and a topological relationship atlas is constructed; and elements and loops in the atlas are classified based on an electric power industry coding rule library, and are converted into the standardized intermediate file in combination with engineering modeling rules.

This invention relates to the field of pollution detection technology, specifically to a method and apparatus for detecting pollution in golf clubs based on AI image recognition technology. The method includes: preprocessing acquired dual-environment images; performing regional correlation analysis based on the topological characteristics of the golf club structure to extract dual-environment spatial feature images; processing the dual-environment spatial feature images; decomposing the processed images to obtain multiple modal feature components; and inputting each modal feature component into a preset static image parsing channel to obtain dual-environment structured feature data. This invention generates personalized cleaning strategies based on the determined actual pollution point information, including matching suitable cleaning media and cleaning methods, thereby ensuring the targeted and effective nature of cleaning operations and reducing the impact on non-polluted areas.