36 results about "Defined Imaging" patented technology

The invention provides an image defogging enhancement and intelligent perception collaborative optimization method based on stream matching, and the method comprises the steps: obtaining a foggy input image, and constructing an ordinary differential equation for defining image transformation; the input image is input into a fog perception vector field, the fog perception vector field comprises an atmospheric scattering purifier and a defogging perception color lookup table, and the atmospheric scattering purifier is used for extracting multi-scale features from the foggy input image to generate clearer output; the defogging perception color lookup table adaptively adjusts the image color through a nonlinear color transformation mechanism; and solving the ordinary differential equation through an RK4 solver, iteratively updating the foggy input image at each time step, and ensuring that the input image can be stably converted into a clear image from an initial foggy state. The technical problems that in the prior art, the existing method is forced to reduce the model capacity due to the calculation efficiency constraint in the ultra-high-definition image defogging, so that the fog concentration estimation is incomplete, and the perception quality and the calculation efficiency cannot be considered at the same time are solved.

The invention discloses a high-performance video and UI (User Interface) display method based on a layered rendering architecture, which comprises the following steps of: defining image display layers and a priority sequence thereof, sequentially rendering each layer according to a sequence from low priority to high priority, and ensuring that the high priority layer can correctly cover the low priority layer; and controlling the display sequence of each layer by adopting a double-track system that a local display sequence is determined according to a z-index value in a lamination priority and the same priority, and synthesizing a final screen display image. The technical scheme of the invention has the main advantages that real-time rendering of a high-quality video stream is supported while UI responsiveness is kept, and the requirement of a high-performance scene is met; it is ensured that the UI layer is always kept above the video layer, or the stacking sequence is dynamically adjusted according to application requirements; and a hierarchical design principle is adopted, and loose coupling is realized among the modules, so that independent upgrading and maintenance are facilitated, and the expandability of the system is enhanced.

Surface determination systems, threat detection systems and medical treatment systems are described. According to one aspect, a surface determination system includes processing circuitry configured to access a three-dimensional complex-valued image volume of a target, define image locations of the image volume using first and second dimensions of the image volume, for each of the first locations, identify voxels along a third dimension of the image volume that correspond to the respective first location, for each of the first locations, select one voxel that corresponds to the first location as having an increased amplitude compared with other voxels that correspond to the first location, and for each selected voxel, use phase information to identify a second location in the third dimension for the selected voxel that corresponds to a location of a surface of the target and that is different than a third location of the respective voxel.

This invention discloses a navigation-following unmanned surface vessel (USV) formation control method based on active vision, comprising the following steps: constructing a kinematic model of the USV based on a geodetic coordinate system; establishing a gimbal monocular camera model for the gimbal camera mounted on the USV; defining servo tracking errors for the navigation-following USV formation system by combining the USV kinematic model and the gimbal monocular camera model; defining image coordinate errors based on the viewing angle constraints of the gimbal monocular camera; designing a motion controller and adaptive parameter update law for the USV using the Lyapunov synthesis method; and designing a gimbal camera steering controller and adaptive parameter update law using the Lyapunov synthesis method. This method ensures the smooth implementation of visual servo control of the USV by defining servo tracking errors and image coordinate errors, and then incorporating these two errors into the Lyapunov synthesis method to design the USV motion controller, the gimbal camera steering controller, and the adaptive parameter update law.

This invention discloses an image data filtering method and system based on an object detection learning algorithm, relating to the field of intelligent parking management. The method includes: using a modified object detection model training method and an initially labeled image dataset to obtain a basic object detection model; and on this basis, using a partitioned test set to define image-level uncertainty based on image content recognition accuracy information to obtain preliminary image filtering indicators; further, when the amount of image data filtered in one step is much greater than the expected amount of labeled data, a second image filtering is performed through image feature clustering to ensure that the filtered image content has a diverse distribution under the same data volume conditions; through several iterative model updates, the energy efficiency ratio of newly added labeled data to model accuracy can be ultimately improved.

The invention provides a method for determining the organic matter surface porosity of the total area occupied by organic matter pores, which comprises the following steps: carrying out an electron microscope scanning experiment on a shale sample to obtain a first scanning electron microscope image of the shale sample; extracting all organic matter regions in the first scanning electron microscope image of the shale sample by using preset image processing software to obtain the total area fraction of organic matters on the surface of the shale sample; obtaining the average surface porosity of organic matter pores from the first scanning electron microscope image of the shale sample according to a preset fifth rule; and according to the total area fraction of the organic matters on the surface of the shale sample and the average surface porosity of the organic matter pores, obtaining the organic matter surface porosity of the organic matter pores in the total area according to a preset formula. The organic matter surface porosity of the organic matter pores in the total area can characterize the characteristics of the shale surrounding pore structure, reflect the shale organic matter thermal evolution process, and provide a basis for evaluating shale gas reservoir and dessert section optimization.

The application discloses an image classification method and device, a storage medium and a computer device. The method is applied to the computer device and comprises the following steps: in a self-defined image classification mode of an electronic album, in response to a triggering operation of a user on an update control of an image corresponding to a to-be-managed image category, updating the image in the to-be-managed image category, calling a training interface, obtaining a first image classification model corresponding to the self-defined image mode through the training interface, the model being pre-trained, and performing transfer training on all images of the to-be-managed image category according to the first image classification model, so as to obtain a first image classification model after transfer training. The classification of the updated image can be matched by using the first image classification model after transfer training, the classification accuracy is improved, finally, the to-be-classified images of the electronic album are classified by using the trained first image classification model, the image classification efficiency and accuracy are improved, and the image searching efficiency is further improved.

A head-mounted display system includes an emissive micro-display including an array of light emitters corresponding to pixels, where the array of light emitters is configured to output image light that defines images. The system also includes an aspheric lens configured to receive the image light from the emissive micro-display and correct the image light. The system also includes a first surface mirror configured to redirect the image light from the aspheric lens from a first direction to a second direction substantially orthogonal to the first direction. Moreover, the system includes a collimating optic configured to receive the image light from the first surface mirror and collimate the image light. In addition, the system includes an eyepiece configured to receive the image light from the collimating optic and to direct the image light to an eye of a user.

The decoding device includes circuitry and a memory connected to the circuitry. The circuitry acquires an image from a bitstream and a first parameter associated with the image. The first parameter includes modal information representing the image category, which indicates that the image category is at least one of the following: a visible light image, an infrared image, an ultraviolet image, an undefined image, or a pre-defined image category for extension.

A system (100), and method (200) thereof, is provided for identifying and forecasting animal activity. The method (200) may comprise receiving (202) the images (112) from the camera (110) to define image specific raw data (114); assimilating (204) the image specific raw data (114) with contemporaneous weather data (140); determining (206) a mean rate of animal sightings for each of a plurality of prior time intervals associated with the assimilated image specific raw data using a first analytics model; minimizing (208) a difference between observed sightings and predicted sightings based at least in part on outputs from the first analytics model using a second analytics model; and determining (212) an amount of likely animal sightings for each of a plurality of future time intervals based at least in part on the outputs from the first analytics model and outputs from the second analytics model using a third analytics model.

This invention discloses a method for inspecting the quality of bed making in hotels using a bed-making service robot. The method is integrated into the robot's control system and includes the following steps: determining the bed-making quality inspection target, defining image range information, range type, and inspection dimension; generating control commands for the robot's sensors based on the image range information; acquiring image group information returned by the robot sensors; loading an image recognition model, including a standard definition unit, a feature extraction unit, and a feature comparison unit; inputting the image group information into the image recognition model to obtain inspection results and correction strategies. According to this technical solution, when inspecting the quality of bed making services provided by the robot, clear inspection standards can be established, reducing reliance on manual labor, saving labor costs, improving the efficiency of hotel room preparation and control, ensuring greater accuracy and consistency in inspection, and providing readable evaluation results.

The present invention discloses a method for enhancing image watermarks based on a neural network structure. A neural network model for image watermark enhancement is constructed based on the neural network model ViT and pre-trained on an image dataset. The pre-trained model is then used to perform image watermark enhancement processing through a multi-step iterative method, generating deceptive adversarial noise on the image. This is used to enable the model to learn a pre-defined image watermark pattern, thereby making it difficult for the ViT network trained on the image after generating the adversarial noise to forget the pre-defined watermark pattern. The masked noise is then added to the image pixels, thereby obtaining an enhanced watermarked image. The technical solution provided by the present invention significantly improves the difficulty of removing watermarks using existing watermark removal technologies by reducing the difference in activation values ​​between clean samples and watermarked samples, and can be widely promoted and applied.

This application discloses an image retrieval method and apparatus, relating to the field of knowledge distillation technology. The method includes: using user-input query text as the retrieval source, and employing text semantic understanding and image feature matching algorithms to filter and output target images from a pre-defined image library that match the text content description, thereby completing cross-modal text-image retrieval. This solves the technical problems in related technologies, such as the difficulty in deeply mining core knowledge like fine-grained probability distribution and coarse-grained structural features of the attention mechanism during image retrieval, and the inability of the distillation process to cover all dimensions of knowledge at the attention level, resulting in insufficient integrity of knowledge transfer. The method achieves the technical effect of helping users quickly locate target images with concise text descriptions through deep correlation matching between text and images, effectively simplifying the image search operation process, and providing users with an efficient and convenient image retrieval and browsing experience.

Adaptive acquisition control timing control for image and video acquisition and processing includes obtaining, by an auto-exposure compensation component of an image processing pipeline of an image capture apparatus, during a current sensor readout window corresponding to obtaining current input image data from an image sensor of the image capture apparatus, in accordance with defined image acquisition and processing timing data, current auto-exposure compensation data for processing subsequent input image data obtained from the image sensor during a subsequent sensor readout window sequentially immediately subsequent to the current sensor readout window, and outputting, to a second data storage unit instance of the image processing pipeline, by the auto-exposure compensation component, during the current sensor readout window, the current auto-exposure compensation data.

The present application relates to a kind of training method, equipment and medium for the multi-task prediction model of medical image, the method includes the following steps: obtaining multi-center, multiple devices, multiple tracer initial medical image and clinical risk factor feature, the initial medical image includes PET image and CT image;First equalization processing is carried out to initial medical image corresponding lesion segmentation gold standard, obtain equalized 3D image sequence;Resampling is carried out to the 3D image sequence, extract deep learning feature and pre-defined image group feature;Second equalization processing is carried out to the deep learning feature, and feature post-processing is carried out to the equalized feature;The deep learning feature, pre-defined image group feature and clinical risk factor feature obtained are spliced, and final feature is obtained;With final feature as input, multi-task prediction model is trained and obtained.Compared with prior art, the present application has the advantages of high medical image processing accuracy, suitable for multi-modal image and the like.

The invention relates to the technical field of transcranial magnetic stimulation medical image processing, and discloses a transcranial magnetic stimulation medical image processing system and method.The system comprises an image obtaining module used for collecting multi-modal medical image data and defining image partitions according to clinical diagnosis requirements; the feature analysis module is used for extracting stimulation associated features and constructing a feature mapping relation between the image subareas and the image positioning stimulation areas; the region fusion module is used for screening a core region and analyzing a parameter fusion matching degree to obtain parameter focusing fusion information; the mapping relation updating module is used for updating the feature mapping relation based on the parameter focusing fusion information; and the processing execution module is used for determining a processing target and a process according to the updated mapping relation and generating a processing execution library. The method corresponds to module steps of the system. According to the scheme, the precision and efficiency of transcranial magnetic stimulation medical image processing can be improved, and clinical diagnosis requirements are met.

The application provides a flow matching-based image defogging enhancement and intelligent perception collaborative optimization method, including the steps of: acquiring a foggy input image, constructing a common differential equation defining image transformation; inputting the input image into a fog perception vector field, the fog perception vector field including: an atmospheric scattering purifier and a defogging perception color lookup table, the atmospheric scattering purifier being used to extract multi-scale features from the foggy input image to generate a clearer output, and the defogging perception color lookup table being used to adaptively adjust the image color through a nonlinear color transformation mechanism; solving the common differential equation through an RK4 solver, and iteratively updating the foggy input image at each time step to ensure that the input image can be smoothly and stably transformed from the initial foggy state to a clear image. The application solves the technical problem that the existing method in the prior art is forced to reduce the model capacity due to the calculation efficiency constraint in the ultra-high-definition image defogging, resulting in incomplete fog concentration estimation and inability to balance the perception quality and calculation efficiency.

A method for comparing the utilization rate of UAV flight path images belongs to the field of remote sensing data processing technology. The method involves: labeling images acquired by the UAV to select target areas; classifying images based on these target areas; after filtering, images are divided into three categories: irrelevant images without target information, valid images containing target information, and invalid images where target information is occluded; defining image utilization evaluation indicators, including total utilization (TU), effective utilization (EU), and occlusion rate (OR); TU represents the image utilization rate without considering occlusion, and EU represents the image utilization rate considering occlusion; and calculating the image utilization rate. This invention can effectively quantify the utilization rate of images captured by UAVs, providing technical support for further flight path planning, improving modeling accuracy, and reducing operational costs.

A surgical system includes a handheld cable housing with a user input device positioned thereon. The surgical system includes an endoscope shaft with a distal end and a proximal end. The proximal end of the endoscope shaft is coupled to the handheld cable housing, where the distal end of the endoscope shaft comprises an image sensor. The endoscope shaft is rotatable relative to the handheld cable housing. One or more angular position sensors configured to measure an angular offset relative to a defined image horizon. Upon receiving angular offset data indicative of an angular offset relative to the defined image horizon, an image processor generates rotated image data based on the angular offset data and causes display of the rotated image data.

A method for classifying plants for agricultural purposes. The method includes: detecting a field section having plants using an optical detection unit inclined toward the field section to obtain image information having a rear image edge having an image width and a front image edge having a larger image width; classifying the plants in the image information into first and second plant classes using an algorithm to treat the plants of the second plant class with a liquid spray agent, in accordance with the classification. The step of classifying comprises a step of reclassifying, in which classified plants are reclassified from the second plant class to the first plant class using the algorithm if they extend in a rear image section of the image information defined by a defined image separation line from a lateral image edge of the image information maximally up to a defined reclassification boundary line.

The present invention refers to an image acquisition method, comprising, in particular in a repeating, continuous, conditional closed-loop and / or conditional feedback manner, processes of capturing one or plural images of or from a sample carrier supported on a microscopy arrangement by using an optical engine (20), analyzing one or plural of already captured images and deriving imaging status data therefrom by using one or plural image analyzer units (30), comparing said derived imaging status data with predefined imaging target data and obtaining a comparison result, based on the comparison result, deriving setting data and / or a setting command, in particular designed for setting the optical engine (20), and depending on the comparison result and / or based on the setting data and / or the setting command conditionally setting the optical engine (20) and / or its operation for further images to be captured. Also, an image acquisition system (1) as suggested, which is configured to capture one or plural images of or from a sample carrier supported in or on a microscopy arrangement and to perform, use and / or to be used in the image acquisition method.

A method and system for traffic event recognition based on a large model using video is disclosed. The method includes: training a pre-defined attention-based neural network model using historical traffic videos as input data and corresponding classification information as output data; training a pre-defined image-text encoder using keyframe data from historical traffic videos as input data and corresponding encoded data as output data; fusing and encoding the classification information and encoded data in a projection layer to obtain fused data; training a pre-defined large language model using the fused data as input data; and inputting real-time traffic videos into the combined model to obtain fine-grained classification information corresponding to the real-time traffic videos. This method achieves finer-grained traffic event recognition, better meeting current traffic management needs.

The present application relates to the technical field of digital watermarking, and proposes a traceability watermarking method and system resistant to partial screen shooting. The watermarking generation method resistant to partial screen shooting comprises the following steps: converting watermark information into a watermark bit matrix; replacing 0 and 1 in the watermark bit matrix by defined image blocks and through Kronecker product, to obtain a watermark image matrix; periodically replacing a bit block without a bit in the watermark image matrix by a positioning block, to form a hybrid watermark matrix, and redundantly copying the hybrid watermark matrix to obtain a redundant watermark matrix; mapping the bit block to a red channel on a white background, rendering the positioning block to a green channel, keeping a blue channel white, and generating a visual watermark image with the ability to resist local shooting interference. W M x N WM WM t 0 、t 1 WM1 WM1 t m WM2 WM3 RGB t 0 、t 1 t m W rgb The present application can effectively deal with watermark extraction under the condition that screen content is not complete and lacks synchronization basis, and ensure high accuracy of watermark extraction.​​​​​​​​​​​​​

The invention discloses an industrial quality inspection confrontation attack defense method and device, electronic equipment, a medium and a program product, and relates to solving the problem of how to effectively defend confrontation attacks at low cost. The anti-attack defense method for industrial quality inspection in the embodiment of the invention comprises the following steps: collecting image data of an industrial product through a visual sensor; defining the image attributes and features related to the defects of the industrial product or material as labels; performing machine learning by using a combination of the image data and the labels as training data, and constructing a quality inspection model for detecting and classifying defects of industrial products or materials; generating an adversarial sample by generating an adversarial network, generating a disturbance corresponding to the image data using the image data, and superimposing the disturbance on the image data; mixing the training data with the adversarial samples to generate an adversarial training set; and performing adversarial training on the quality inspection model by using the adversarial training set.

The invention discloses an image fusion enhancement method and system based on an FPGA. The method comprises the following steps: acquiring image data comprising a CLAHE enhanced image of a first channel and an original image of a second channel; performing two-dimensional mean filtering processing on the original image to generate a mean image; defining image blocks; calculating the offset position of the pixel in the image block to which the pixel belongs; calculating the distance from the pixel to the center of the image block to which the pixel belongs according to the offset position of the pixel in the image block to which the pixel belongs; loading and switching a weighting function curve by utilizing a programmable lookup table; obtaining a weight value according to the weight function curve and the distance from the pixel to the center of the image block to which the pixel belongs; and performing weighted fusion on the CLAHE enhanced image pixel value and the mean value image pixel value according to the weight value, and outputting a weighted fusion pixel value and an effective signal. According to the invention, flexible switching of image fusion strategies during operation can be realized, and the smoothness and adaptive ability of image enhancement are improved, so that the technical problem that a fixed fusion weight in a traditional hardware structure is not suitable for multiple scenes is solved.

The invention discloses an unmanned ship formation control method based on an active vision piloting-following method. The method comprises the following steps: constructing an unmanned ship kinematics model based on a geodetic coordinate system; establishing a cradle head monocular camera model for a cradle head camera carried on the unmanned ship; defining a servo tracking error for the piloting-following method unmanned ship formation system by combining the unmanned ship kinematics model and the holder monocular camera model; defining an image coordinate error based on the view angle constraint of the cradle head monocular camera; designing an unmanned ship motion controller and an adaptive parameter updating law through a Lyapunov synthesis method; and designing a pan-tilt camera steering controller and a self-adaptive parameter updating law through a Lyapunov comprehensive method. According to the method, a servo tracking error and an image coordinate error are defined, the two errors are introduced into a Lyapunov synthesis method to design an unmanned ship motion controller and a pan-tilt camera steering controller, and an adaptive parameter updating law is adopted, so that the smooth operation of visual servo control of the unmanned ship is ensured.

The application relates to the technical field of transcranial magnetic stimulation medical image processing, and discloses a transcranial magnetic stimulation medical image processing system and method. The system comprises an image acquisition module, which is used for collecting multi-modal medical image data and defining image partitions according to clinical diagnosis requirements; a feature analysis module, which is used for extracting stimulation-related features and constructing a feature mapping relationship between the image partitions and image positioning stimulation areas; a region fusion module, which is used for screening core regions and analyzing parameter fusion matching degrees to obtain parameter focus fusion information; a mapping relationship updating module, which is used for updating the feature mapping relationship based on the parameter focus fusion information; and a processing execution module, which is used for determining processing targets and processes and generating a processing execution library according to the updated mapping relationship. The method corresponds to the steps of the modules of the system. The scheme can improve the accuracy and efficiency of transcranial magnetic stimulation medical image processing and meet the clinical diagnosis requirements.

The invention relates to the technical field of digital watermarking, and provides a traceability watermarking method and system capable of resisting partial screen shooting. The watermark generation method for resisting partial screen shooting comprises the following steps: converting watermark information W into an M * N watermark bit matrix WM; respectively replacing 0 and 1 in the watermark bit matrix WM with defined image blocks t0 and t1 through a Kronecker product to obtain a watermark image matrix WM1; in the watermark image matrix WM1, bit blocks without bits are periodically replaced with positioning blocks tm to form a mixed watermark matrix WM2, and the mixed watermark matrix WM2 is subjected to redundant copying to form a redundant watermark matrix WM3; and on a white RGB background, mapping the bit blocks t0 and t1 to a red channel, rendering the positioning block tm into a green channel, keeping the blue channel white, and generating a visual watermark image Wrgb with local shooting interference resistance. The method can effectively cope with watermark extraction under the conditions that shot or captured screen content is incomplete and synchronous basis is lacked, and ensures high accuracy of watermark extraction.

The embodiments relate to a method and a technical equipment for implementing the method. The method comprises receiving encoded omnidirectional video content for displaying, wherein the encoded omnidirectional video content comprises a viewport region, a margin region outside the viewport region and a background region (801); defining image regions to be included in the margin region, wherein the definition is based on a margin speed threshold (802); determining a head speed and a direction of motion (803); determining when the head speed is greater than or equal to the margin speed threshold, and as a response to the determining, adding negative margins in one or more direction with respect to the viewport (804); downloading the viewport with a first quality; and downloading the negative margins at a quality lower than the first quality (805).

The invention belongs to the technical field of unmanned aerial vehicle visual servo control, and particularly discloses an unmanned aerial vehicle self-adaptive visual servo control method based on second-order cone constraint, and the method comprises the steps: building a four-rotor kinetic model, and building an inertial coordinate system and a vehicle body coordinate system; a virtual image plane thought is introduced, a visual servo model in a virtual image plane is deduced, and an image point dynamic model which does not change along with the posture is obtained; defining image moment characteristics and deducing a decoupling image moment characteristic dynamic model; the difference between the image features and the expected image features is used as input, the target is restrained to be within the view range of the camera through a second-order cone rule, and the restrained features are converged to the expected position with the shortest path; and finally, selecting a reference servo gain value according to the size of the image feature error, and adjusting the adaptive servo gain according to the size of the image feature error. The method can enable the control system to converge rapidly and stably, and effectively guarantees that the image features are always in the field of view.