298 results about "External image" patented technology

The invention belongs to the technical field of medical image processing, and particularly relates to a medical vision-language pre-training system guided by an anatomical area. The system comprises an anatomical region and report clause alignment module, an internal representation learning reinforcement module and an external representation learning reinforcement module. The alignment module is composed of a target detector, a text parser and an automatic alignment normal form between the target detector and the text parser, and the model is endowed with fine-grained alignment capability through comparative learning; the internal characterization learning reinforcement module uses an image encoder and an image label identification decoder to associate image features with corresponding labels, and provides a function of open disease category detection; the external representation learning reinforcement module utilizes comparative learning of a soft label and a hard label to improve relevance between different image-report pairs. According to the method, the anatomical region is taken as the center, and meanwhile, the fine-grained semantic association of the image-report to the interior and the exterior is enhanced, so that the downstream task accuracy based on the current pre-training system can be remarkably improved.

The invention relates to an unmanned aerial vehicle control system and method for formation flight in a wind environment. The method comprises the following steps: generating a wind disturbance environment with a specified map scale and containing random obstacles; acquiring external image data and inertial measurement data of the unmanned aerial vehicle, and generating a grid map through point cloud extraction and conversion operation; the method comprises the following steps: modeling a spatial topological structure of an unmanned aerial vehicle cluster by adopting an undirected graph mode, and solving and generating a collision-free initial trajectory by adopting an unconstrained discrete optimization mode; optimizing the initial trajectory by adopting a task redistribution strategy, and re-planning a global target and a local target to obtain a re-planned optimized trajectory; and based on the optimized trajectory, solving by adopting a model prediction control mode to obtain an optimal control quantity for unmanned aerial vehicle bottom layer control so as to correspondingly control the motion state of each unmanned aerial vehicle and realize formation flight. Compared with the prior art, the unmanned aerial vehicle cluster can efficiently and accurately realize formation flight in an obstacle environment with wind influence.

The invention relates to the technical field of Internet of Things, in particular to an Internet of Things equipment operation and maintenance method based on large model fine tuning and multi-mode RAG, and a memory, and the method comprises the steps: respectively collecting external image information and data information for to-be-diagnosed equipment; generating question information according to the external image information and the data information, and inputting the question information into the Internet of Things large model to obtain feedback information fed back by the Internet of Things large model; and determining whether the to-be-diagnosed equipment is abnormal according to the feedback information. Aiming at the problems of low efficiency and inaccuracy of the manual inspection process of the Internet of Vehicles equipment in the prior art, external image information corresponding to equipment appearance inspection and data information acquired through channels such as a sensor and equipment metadata are introduced; and an Internet of Things large model optimized for the Internet of Things field is introduced to analyze the information, so that the Internet of Things large model effectively judges whether a fault exists or not, the inspection accuracy is improved, only appearance images need to be shot, and the efficiency is improved.

The invention provides an unmanned aerial vehicle commanding and dispatching system and an unmanned aerial vehicle combined inspection method and device. The system comprises a monitoring module, a situation overview module, a commanding and dispatching module, a combined flying module, a task module and an operation analysis module. The monitoring module is used for monitoring the frame outside the nest of the unmanned aerial vehicle and the frame of the unmanned aerial vehicle; the situation overview module is used for browsing a base station list, an unmanned aerial vehicle list, an inspection statistics list, an inspection record list and an operation analysis list to query corresponding information; the commanding and dispatching module is used for switching the base stations, acquiring real-time information corresponding to the base stations, and commanding and controlling the unmanned aerial vehicle based on the real-time information and a predetermined commanding strategy; the combined flying module is used for the plurality of base stations to perform combined inspection through a combined inspection route; the task module is used for displaying an inspection record list and an inspection plan list; and the job analysis module is used for displaying the task list and the result list. According to the mode, display and route planning of multi-base-station unmanned aerial vehicle combined inspection are realized.

The invention relates to the technical field of vending machine inventory management, in particular to a vending machine automatic inventory management method based on image recognition, and the method comprises the steps: collecting internal and external images of a vending machine in real time, extracting the internal images of the vending machine, recognizing and analyzing the types of commodities, and generating a commodity display list; extracting a selling number evaluation index and a selling speed evaluation index of various types of commodities corresponding to each monitoring time point, calculating a first selling evaluation value, extracting an external image of the vending machine, identifying and analyzing consumer behaviors, calculating a ratio of a new consumption value, an old consumption value and a target consumption number, determining a second selling evaluation value, and calculating a selling speed evaluation value; the first sales evaluation value and the second sales evaluation value are combined and calculated through normalization processing to obtain the total sales evaluation value of various commodities, the change trend of the total sales evaluation value is analyzed, the sales trend value of various commodities is determined, and inventory management parameters are adjusted accordingly, so that the real-time monitoring of the commodity inventory state is realized, and the commodity inventory management efficiency is improved. The operation efficiency and the profitability of the vending machine are improved.

The present invention discloses a method and storage medium for estimating the absolute posture of a vehicle based on visual line features. The method comprises the following steps: S1, processing an image of the outside of a vehicle acquired by a vehicle camera and mapping the image onto a Gaussian sphere; S2, setting the total number of iterations of a RANSAC algorithm; S3, selecting any three points (P1, P2, P3) on the Gaussian sphere to construct a set of Manhattans; S4, calculating the number of inner points of the set of Manhattans according to an error formula of points and vectors; S5, repeating steps S3 and S4 until the total number of iterations set in step S2 is reached, selecting a set of Manhattans with the largest number of inner points, and calculating the camera posture R according to all the inner points and the formula cw ; S6, calculate vehicle posture R vw , to estimate the absolute posture of the vehicle. The present invention can accurately estimate the position so that the position and posture of the vehicle in the indoor environment positioning process are more accurate.

A medical robotic system and method of operating such comprises taking intraoperative external image data of a patient anatomy, and using that image data to generate a modeling adjustment for a control system of the medical robotic system (e.g., updating anatomic model and / or refining instrument registration), and / or adjust a procedure control aspect (e.g., regulating substance or therapy delivery, improving targeting, and / or tracking performance).

The invention provides a welding spot degradation state analysis method and system based on deep learning and a storage medium, and the method comprises the following steps: collecting equipment information and welding spot external images of various sample equipment, and constructing a welding spot external image set in combination with the equipment information and the welding spot external images; according to the welding spot position information, welding spot positions are processed through a welding spot slicing method, and welding spot internal images of all sample devices are collected; constructing a welding spot degradation state set according to the internal images of the welding spots; constructing a welding spot degradation classification model based on the neural network model; training a welding spot degradation classification model by using the welding spot external image set and the welding spot degradation state set; inputting the target welding spot image into the trained welding spot degradation classification model to obtain a welding spot state classification result output by the welding spot degradation classification model; and determining the welding spot degradation state of the to-be-detected welding spot according to the welding spot state classification result. The device has the advantages that the detection efficiency is high and the device cannot be damaged when the degradation of the welding spot of the device is detected.

Systems and methods for calibrating deformable sensors are disclosed. In one embodiment, a method of calibrating a deformable sensor includes capturing image data of the deformable sensor using an external image sensor, wherein the deformable sensor comprises a deformable membrane defining an enclosure that is configured to be filled with a medium. The method further includes comparing the image data of the deformable sensor to a metric. When the image data does not satisfy the metric, the method includes adjusting a pressure within the enclosure.

A display apparatus transmits a picture acquisition request for getting picture information to an external image apparatus connected through a predetermined interface to the display apparatus from the external image apparatus at predetermined intervals and gets a plurality of pieces of picture information from the external image apparatus to be displayed. The plurality of pictures may be switched at predetermined intervals, for example, to be displayed, so that the plurality of pictures may be displayed in a so-called slide show manner. A plurality of pictures for thumbnail may be produced from the plurality of pieces of picture information and be arranged together to be displayed in one picture screen of a display device.

An information processing apparatus includes a controller configured to acquire, when a suspension of operation occurs in an autonomous vehicle, positional information, vehicle information, a vehicle internal image, and a vehicle external image of the autonomous vehicle, and determine, in a case in which a suspension of operation occurs in a plurality of autonomous vehicles, an order of priority for a rescue team to go to the plurality of vehicles, based on the positional information, the vehicle information, the vehicle internal image, and the vehicle external image.

A vehicle data relation device includes an internal audio / image data analyzer, configured to receive first data representing at least one of audio from within the vehicle or an image from within the vehicle; identify within the first data second data representing an audio indicator or an image indicator, wherein the audio indicator is human speech associated with a significance of an object external to the vehicle, and wherein the image indicator is an action of a human within the vehicle associated with a significance of an object external to the vehicle; an external image analyzer, configured to receive third data representing an image of a vicinity external to the vehicle; identify within the third data an object corresponding to at least one of the audio indicator or the video indicator; and an object data generator, configured to generate data corresponding to the object.

The invention relates to the field of aircraft external detection, and discloses an all-weather aircraft external defect intelligent detection system based on digital twinning, and the system comprises a multi-modal data collection module which is used for collecting aircraft external multi-modal data, and the multi-modal data comprises aircraft external image data, ultrasonic wave, X-ray and thermal infrared data; the fusion module is used for enabling the multi-modal data to form a unified fusion feature; the defect identification module is used for carrying out aircraft external defect identification based on a defect identification model and the fusion features; and the spatial positioning module is used for mapping the identification result. The unmanned aerial vehicle and the mechanical arm unit are combined, and external images, ultrasonic waves, X-rays and thermal infrared data of the aircraft are rapidly and efficiently collected. The unmanned aerial vehicle unit can cover surface images of a wide area, the mechanical arm unit detects external defects through ultrasonic and X-ray technologies, the defect recognition speed is increased, the system adapts to the complex environment, and the all-weather detection requirement is met.

The present invention relates to the field of automobile safety, and specifically discloses a method for eliminating the blind area of ​​the AB pillar of a vehicle. The present invention collects external images of the AB pillar of the vehicle, facial contour information and head posture information of the driver, determines the driver's line of sight direction through the facial contour information and head posture information of the driver, and then determines the feature vector of the external image of the AB pillar of the vehicle according to the driver's line of sight. The collected external image of the AB pillar of the vehicle is processed according to the feature vector to obtain the AB pillar blind area image corresponding to the driver's line of sight, and finally projects the AB pillar blind area image on the AB pillar of the vehicle to fill the AB pillar blind area field of vision that the driver cannot directly obtain, and helps the driver obtain more information about the surrounding area of ​​the vehicle, so as to make accurate driving operations.

The invention relates to the technical field of road condition acquisition, in particular to an intelligent footpath road condition information acquisition and management system. A rectangular groove is formed in the middle of an elastic plate, so that the end, close to a camera, of the elastic plate is of a frame-shaped structure, and when the end, close to the camera, of the elastic plate cleans the surface of the camera, the camera can collect external images all the time through the rectangular groove, a mounting groove is formed in the lower end of a rubber strip, and a cleaning column is arranged in the mounting groove; the length of the cleaning column is smaller than that of the mounting groove, rotating shafts at the two ends of the cleaning column penetrate through the two ends of the mounting groove and slide in the mounting groove in a reciprocating mode, and a spring is arranged between the rotating shaft at one end of the cleaning column and the mounting groove, so that when the end, close to the camera, of the elastic plate cleans the camera, the camera is cleaned more stably; firstly, dust impurities and intractable impurities on the surface of the camera are cleaned by the cleaning column, and then the cleaned area is scraped by the rubber strip.

The present invention proposes a fabric defect detection method based on multi-modal deep learning. First, a tactile sensor is placed onto the fabric surface with different defects to collect the fabric texture images, a camera is used to collect the corresponding fabric external images, and a fabric external image and a fabric texture image constitute a set of fabric detection data; then, a feature extraction network and a multi-modal fusion network are connected to establish a classification model based on multi-modal deep learning, which uses the fabric texture image and fabric external image in each set of collected fabric detection data as input, and the fabric defect as output; said classification model is trained using the collected fabric detection data; finally, the trained classification model is used to detect the fabric defect. The present invention employs vision-touch complementary information, which can greatly improve the accuracy and robustness of detection.

A medical robotic system and method of operating such comprises taking intraoperative external image data of a patient anatomy, and using that image data to generate a modeling adjustment for a control system of the medical robotic system (e.g., updating anatomic model and / or refining instrument registration), and / or adjust a procedure control aspect (e.g., regulating substance or therapy delivery, improving targeting, and / or tracking performance).

The invention discloses an incremental learning image processing system in the field of image processing, which comprises the following steps of: S1, inputting image information, namely inputting external image information into the system so as to set initial parameters for an incremental learning model in the system; s2, a preprocessing step is carried out through an incremental learning algorithm, the input image information can be marked through preprocessing of the image information, the content of the image information is classified and determined by an incremental learning algorithm model, information areas in the image are independently marked, relevance between the different information areas is set, and the content of the image information is determined; and in the preprocessing link, the color and saturation of the image can be adjusted, so that the partial image of the information area is more convenient to identify, and a plurality of learning logics can be mutually combined by adopting an incremental learning algorithm.

The present invention discloses an X-ray security inspection method and device based on multi-camera fusion. When the inspected passenger places luggage on the X-ray security inspection instrument, the image of the person will be collected in real time, and then the luggage package will be scanned with X-rays. The background will bind the personnel information of this stage with the X-ray image of the luggage, and store the data while corresponding the X-ray visible image of the luggage scan to the passengers who have passed the security inspection one by one. When the background analysis server identifies contraband, the security inspectors can find the passengers carrying contraband or dangerous goods more quickly. In the present invention, on the basis of the traditional X-ray security inspection function, an automatic association function of people and bags is added, and the luggage is bound to its own personnel through image association technology, thereby helping security inspectors to implement security inspection work more quickly and effectively. With the address information of the data storage space and the external image information of the luggage as the association intermediary, the data information between the passenger and the luggage is associated and stored, which is convenient for tracing through images.

The present application relates to the technical field of electronic devices, and provides a terminal device, used for solving the problem of poor heat dissipation performance and safety of terminal devices. The terminal device provided by the present application comprises a display screen, a middle frame, and a back plate, the middle frame is connected between the display screen and the back plate, and the middle frame and the back plate define an accommodating cavity; the terminal device further comprises a cover plate and a lens assembly, the back plate has an outer plate surface facing away from the display screen, and the cover plate is arranged on the back plate and protrudes from the outer plate surface; the lens assembly is located in the accommodating cavity, the cover plate is provided with a light-transmitting hole, and the lens assembly is used for collecting an external image through the light-transmitting hole; the cover plate has a first outer surface, an included angle is formed between the first outer surface and the outer plate surface, the first outer surface is provided with a plurality of ventilation holes, and the mean ratio of the cross-sectional area to the perimeter of the ventilation holes is smaller than or equal to 0.25 mm. The first outer surface of the cover plate is provided with the plurality of ventilation holes for air circulation, and the cross-sectional dimensions of the ventilation holes are reasonably limited, so that the terminal device can achieve effective balance between heat dissipation performance and safety.

In a method for generating image data by means of a camera, which comprises an image sensor that has a plurality of light-sensitive sensor cells arranged in rows and columns, and for processing the image data by means of an external image processing device, the following steps are provided: providing camera specimen-specific data about determined properties of the camera that may deviate from camera model-generic properties of the camera; controlling the image sensor to generate respective image signal values in dependence on the intensity of the light incident on the sensor cells; reading out the image signal values and generating preliminary image data from the image signal values in the camera; transmitting the preliminary image data to the external image processing device together with metadata that comprise the camera specimen-specific data or at least allow an assignment of the camera specimen-specific data to the preliminary image data; and generating processed image data by means of the external image processing device by applying a correction dependent on the camera specimen-specific data to the preliminary image data.

The present invention discloses a vision-based method for estimating the spatial pose of a bionic flapping-wing UAV, which comprises the following steps: obtaining an image training data set; training the image training data set and annotation information using a deep learning detection algorithm; obtaining internal and external parameter information of an image acquisition device, and obtaining flight images of a bionic flapping-wing UAV; testing the flight images of the bionic flapping-wing UAV according to the deep learning detection algorithm to obtain test results of a key point image coordinate system and the wing profile state of the bionic flapping-wing UAV; measuring the three-dimensional coordinates and wing profile states of key points in the coordinate system of the bionic flapping-wing UAV under different wing profile states; and obtaining the spatial position and attitude results of the bionic flapping-wing UAV through calculation. The present invention defines wing profile states according to different flight attitudes, performs calculations after deep learning training and detection, and obtains the spatial position and attitude, thereby solving the problem that the spatial pose of a UAV is difficult to determine through external images due to its complex structure and diverse wing profile states.

The invention discloses a method and system for performing three-dimensional flaw detection on industrial products based on multi-source data, and belongs to the technical field of industrial product flaw detection, and the method comprises the following steps: determining the spatial resolution of a training sample and an ultrasonic flaw detection probe; segmenting the world coordinate system based on the spatial resolution to obtain a three-dimensional space containing a plurality of discrete units; configuring discretized point cloud data in the training sample in a three-dimensional space; assigning the discrete units with the point cloud data as a first type of units, and assigning the discrete units without the point cloud data as a second type of units; obtaining an internal flaw detection model based on the assigned discrete unit; acquiring an external flaw detection model through the external image data of the training sample; determining a to-be-detected workpiece, and inputting the internal point cloud data of the to-be-detected workpiece into the internal flaw detection model to obtain an internal flaw detection result; and inputting the external image data of the to-be-detected workpiece into the external flaw detection model to obtain an external flaw detection result.

In a method for generating image data by means of a camera comprising an image sensor having a plurality of photosensitive sensor units arranged in rows and columns and for processing the image data by means of an external image processing device, the following steps are provided: providing camera sample-specific data relating to determined characteristics of the camera, these characteristics can be different from camera model general characteristics of the camera; manipulating the image sensor to generate a corresponding image signal value as a function of the intensity of light incident on the sensor unit; reading the image signal values in the camera and generating preliminary image data from the image signal values; transmitting the preliminary image data and metadata to an external image processing device, wherein the metadata comprises camera sample specific data or at least allows distribution of the camera sample specific data to the preliminary image data; and generating processed image data by means of the external image processing device by applying a correction dependent on the camera sample-specific data to the preliminary image data.

The invention provides a defect detection system and a defect detection method for a vertical furnace, which can remarkably improve the detection efficiency, effectively reduce or avoid the problems of missing detection, false detection and consistency caused by human factors, thoroughly eliminate occupational health risks in a severe operation environment, and improve the detection accuracy. Therefore, mechanization and automation of semiconductor vertical furnace defect detection are realized. The defect detection system comprises a walking unit configured to walk along a preset track; the rack is arranged on the walking unit; the supporting rod is arranged on the rack and extends in the first direction; the driving unit is arranged on the rack and connected with the supporting rod, and the driving unit is configured to be capable of driving the supporting rod to move in the first direction and driving the supporting rod to rotate around the axis of the supporting rod; and the image acquisition unit is arranged at the first end of the supporting rod and is used for acquiring a furnace body internal image and a furnace body external image of the vertical furnace.

In embodiments, an electronic device is provided. The electronic device may comprise: a camera; a display; at least one sensor; a memory for storing instructions; and at least one processor. The instructions, when executed by the at least one processor, may cause the electronic device to: obtain an external image of the electronic device through the camera while displaying a virtual image of a virtual environment through the display; display, through the display, at least a part of the external image in a virtual area according to first depth information in the virtual environment, while displaying the virtual image; obtain a distance from an external object included in at least the part of the external image through the at least one sensor; and display a virtual object corresponding to the external object through the display while displaying the virtual image and at least the part of the external image, on the basis that a position corresponding to the distance from the external object is within a space according to second depth information.

The invention belongs to the technical field of AR (Augmented Reality) equipment, particularly relates to AR glasses supporting dynamic expansion image acquisition equipment and an implementation method thereof, and aims to solve the problem that the existing Android system AR glasses are poor in compatibility when being externally connected with image acquisition equipment and need to depend on an exclusive SDK (Software Development Kit). An Android operating system is built in the AR glasses, and a Camera service module, a driving application program and a video application program are configured in the system. The Camera service module provides a registration interface and a distribution identifier for the external image acquisition equipment, uniformly manages the built-in and registered external image acquisition equipment and generates an equipment identifier list; the driving application program adapts to the registration interface through protocol conversion, and registration of the external image acquisition equipment is completed; and the video application program obtains the identifier list, and calls the corresponding equipment through the system universal interface based on the selected target identifier. According to the invention, the dynamic expansion of the external equipment is realized, the application development cost is reduced, and the adaptation flexibility and use experience of the AR glasses are improved.

The present invention discloses an image processing system for incremental learning in the field of image processing, comprising the following steps: S1 is an image information input step, in which initial parameters are set for an incremental learning model in the system by inputting external image information into the system; S2 is a preprocessing step using an incremental learning algorithm, in which the input image information can be marked by preprocessing the image information, the incremental learning algorithm model classifies and determines the major content categories of the image information, and separately marks the information areas in the image, and sets the correlation between different information areas, and the preprocessing link can adjust the image color and saturation, thereby making the image part of the information area easier to identify, and the use of an incremental learning algorithm can combine multiple learning logics.

The invention discloses a hyperspectral image classification method based on nonlinear weighted graph convolution and a classification model training method, and belongs to the technical field of hyperspectral remote sensing image classification. The objective of the invention is to solve the problem of poor classification performance of a model caused by insufficient mining of high-dimensional information of a hyperspectral image in an existing hyperspectral image classification method. According to the method, a hyperspectral image is clustered into superpixels, a superpixel feature matrix is generated by utilizing internal image convolution, weights are determined based on superpixel nodes and similarity between adjacent superpixel nodes, and weights of adjacent edges of the superpixels are combined into an adjacent matrix of the whole hyperspectral image; performing local information aggregation on the superpixel feature matrix according to an adjacent matrix by external image convolution, and generating a depth feature matrix based on image convolution processing so as to realize classification; in a classification model training process, a curvature regular strategy and an intrinsic dimension regular strategy are utilized to constrain a nonlinear weighted graph convolution layer, so that manifold local information measurement is more accurate, and interference of redundant information on the model is reduced.