29 results about "Multi-image" patented technology

The present application belongs to the technical field of intelligent prediction of geotechnical engineering parameters, and discloses a dynamic multi-modal loess physical and mechanical parameter prediction method based on physical constraints: obtaining and preprocessing the microscopic images and multi-modal data such as penetration resistance of loess samples, and extracting image and numerical features respectively; fusing multi-image features through self-attention and cross-attention guided by numerical features, and then generating joint deep representation through a dynamic gated multi-expert module, and outputting parameter prediction values by a shared multi-layer perception prediction head; adopting a composite loss function based on physical constraints to train the model in a phased strategy, and after deployment, the model is used for new sample prediction. The present application realizes deep adaptive fusion of multi-source information and controllable embedding of physical laws, and improves the precision, generalization ability and physical rationality of loess physical and mechanical parameter prediction.

This invention discloses a single-target tracking method and apparatus that incorporates existence determination through multi-image input and historical trajectory enhancement. A network model for single-target tracking is constructed, comprising a modified micro-sliding converter, a bounding box embedding module, a multi-layer encoder module, a multi-layer decoder module, and a multi-layer perceptron. The network model is trained in two stages by inputting consecutive frames of images with known target bounding boxes into the single-target tracking network model. Then, the trained network model is used to process consecutive frames of images to be tracked, where only the first frame has a known target bounding box, and the remaining frames have unknown target bounding boxes. This invention can fuse target appearance information and target motion trajectory information through multi-image input and multi-bounding box input, making the tracking results more robust and accurate. The model effectively improves the speed and accuracy of target tracking, can determine the existence of the target, and has wide applications.

A system and method for inspection maintenance and / or diagnosis of a variety of workpieces is provided. The system serves workers working on a workpiece, inspectors who are distal from the workers and / or can be used for remote training or for advanced diagnosis and / or repair. The system preferably includes a template of a set of one or more predefined required images of a workpiece required by an inspector to perform their inspection or diagnosis. The set of predefined required images is provided to the worker. The worker captures the images with an appropriate workpiece data capture device and provides them to the inspector for review. The inspector examines the provided images and either approves the workpiece based on their content, requests additional images for further examination and / or provides annotations and other information to the worker to address identified issues. The system maintains a database of all images and information.

The application discloses a performance equalization compression method of satellite images. The method equalizes the excellent compression performance of images and the poor compression performance of images from the perspective of a multi-image system by arranging the images, overcomes the influence of the barrel effect on the practicability of image compression, overcomes the defect that the compression quality of a plurality of images compressed by a JPEG2000 standard compression algorithm is too different, and improves the performance of the standard compression algorithm without increasing complexity. The method has the characteristics of easy software and hardware upgrading and realization, has practical value in a satellite image compression transmission system, and can also play a role in any other multi-source standard compression transmission occasion.

Systems and techniques are described for imaging. For instance a process can include obtaining a first image captured using a first portion of an image sensor. The process can further include obtaining a second image captured using a second portion of the image sensor, wherein the second portion is different from the first portion, determining an image capture setting based on the second image, and applying the image capture setting to the first image.

The application discloses a kind of multi aerial photograph image encryption methods based on cross-image collaborative scrambling and coupling diffusion, comprising the following steps: based on SHA-512 algorithm, respectively for the image in image set generates unique hash sequence value;Chaotic sequence is generated based on 2D-SCHCS chaotic system based on control parameter and initial value;Adaptive block decomposition is carried out, and the index array generated;Optimized Markov chain Monte Carlo is used, and multi-image block hybrid scrambling is carried out, and is decomposed into image-level unit array;Realize image pixel-level reconstruction by metadata-driven reverse reconstruction;Chaotic sequence is dynamically distributed to RGB channel according to channel parity check segmentation mechanism, and pixel-level gradient sensitive disturbance and direction sensitive cyclic shift are executed, to realize single-channel pixel scrambling;Based on the image after single-channel pixel scrambling, dynamic XOR stream is generated using chaotic sequence nonlinear combination, and the coupling diffusion of multi-image pixel value is completed, and finally the ciphertext image set is output.

The present application relates to a kind of ovarian cancer classification staging system and method, belong to medical image processing technical field.The problems of insufficient multi-image processing capability, poor multi-modal compatibility, fixed staging logic, insufficient explainability and clinical adaptability of existing ovarian cancer classification staging technology are solved.The ovarian cancer classification staging system of the present application includes data input module, pre-processing module, feature extraction module, dynamic aggregation module, double-branch reasoning module and result output module.The present application also provides ovarian cancer classification staging method.The ovarian cancer classification staging system and method simultaneously support CT and MRI input, and have high clinical adaptability;Using space-channel-sequence triple attention module, adaptive processing of indefinite number of full images is realized, and information utilization and classification accuracy are greatly improved;Using decision module, reasoning speed and logical rigor are improved;Standardized diagnosis unifies classification staging standard, and reduces subjective differences of different doctors.

This invention discloses a multi-image encryption method based on 3D face keys and multiplexed digital holography. The encryption steps include: using 3D face high-order data acquisition technology, chaotic face structured light phase mask generation technology, grating modulation technology, multiplexed digital holographic coding technology, and watermark embedding technology based on discrete wavelet domain singular value decomposition, multiple plaintext information images can be encrypted into an amplitude-type holographic ciphertext and embedded into a host image; the decryption steps include: the user must first calculate the face similarity using a 3D face multilayer perceptron neural network. If the similarity exceeds a preset threshold, authentication is successful. Then, the final decryption result is generated using watermark image extraction technology based on discrete wavelet transform, image decryption technology based on the Cramer-Kroni relation, and spectral filtering technology; otherwise, authentication fails. This method has advantages such as a large key space, high sensitivity, strong robustness of the 3D face key, and low decryption crosstalk.

This invention relates to an electrical drawing recognition method and system. The recognition method includes: acquiring multiple images corresponding to the electrical drawing and generating a dataset directory structure that conforms to the training requirements of the YOLO model; loading the YOLO model and training the classifier model and recognizer model for each component, and saving the training results; initializing the classifier model, recognizer model, and OCR engine; loading the trained classifier model and performing YOLO object detection on the input image, outputting the component type, bounding box coordinates, and confidence score; loading the trained recognizer model and inputting the corresponding component region image to detect key regions within the component; the OCR engine extracting textual information from the key regions within the component image and the rule engine integrating the fields corresponding to the key regions within the component to generate structured data after recognition; and outputting the recognition results for all components. This invention can automatically and accurately recognize component information in electrical drawings.

Example methods and systems for image reconstruction using multiple image reconstruction chains for radiation therapy are described. In one example, a computer system may obtain projection image data 110 associated with a target structure within a patient. The computer system may generate, using a first image reconstruction chain 121, 122, 122N, first volume image data 131, 132, 132N based on the projection image data. The computer system may generate and display, on a display device, a first user interface (UI) 141 view for a user to interact with the first volume image data. The computer system may also generate, using a second image reconstruction chain, second volume image data based on at least one of the following: the projection image data and the first volume image data. The computer system may generate and display, on the display device, a second UI 142 view for the user to interact with the second volume image data.

The embodiment of the application discloses a vehicle re-identification model construction method based on multi-image learning, comprising: acquiring a vehicle image set, each vehicle image comprising a single vehicle and being labeled with a vehicle category; according to the space-time information of each vehicle image, combining multiple vehicle images taken at the same space-time position into an image package, and taking the vehicle category with the most labels in the image package as the labeled category of the image package; and training a vehicle re-identification model based on deep learning by using each image package. The embodiment improves the accuracy of vehicle re-identification.

We provide the system. [Solution] A means of inputting multiple image data, Means for recognizing objects and backgrounds based on the aforementioned image data and input related information, A means for generating narrative data based on the aforementioned recognition results, A means for automatically generating animation data based on the aforementioned story data, Means for encoding and outputting the aforementioned animation data, A system that includes this.

The application relates to the fields of information security and image processing, and discloses a multi-image parallel encryption processing method based on an improved chaotic mapping, which comprises the following steps: separating and stacking each color channel of multiple images to be encrypted to construct a three-dimensional pixel matrix; extracting a matrix hash digest to be converted into a sub-key; inputting the sub-key into an improved chaotic mapping model of a historical state variable to iteratively generate a scrambling and diffusion sequence; performing sorting scrambling on the three-dimensional matrix in multiple orthogonal dimension sections by using the sequence; mapping the sequence into a shift step length to perform a cyclic shift and a bitwise XOR linkage diffusion on the pixels; and finally, outputting ciphertext images by segmentation. The application breaks the physical isolation of a single image to realize spatial depth mixing by constructing a three-dimensional matrix and performing cross-dimension scrambling; the chaotic dynamics degradation under limited precision is overcome by means of a historical state delay; and the efficiency of multi-image encryption and the system attack resistance are improved by using a shift and XOR linkage to realize bit-level depth confusion.

Example methods and systems for image reconstruction using multiple image reconstruction chains for radiation therapy are described. In one example, a computer system may obtain projection image data associated with a target structure within a patient. The computer system may generate, using a first image reconstruction chain, first volume image data based on the projection image data. The computer system may generate and display, on a display device, a first user interface (UI) view for a user to interact with the first volume image data. The computer system may also generate, using a second image reconstruction chain, second volume image data based on at least one of the following: the projection image data and the first volume image data. The computer system may generate and display, on the display device, a second UI view for the user to interact with the second volume image data.

Approaches to improving frame rate and visual smoothness in real-time video streams through multi-frame interpolation are revealed. A neural classification network analyzes two consecutive images and outputs confidence scores indicating the reliability of motion data for each pixel. These scores determine whether the motion of a pixel is accurately described by the motion vectors or should be treated as static. The classification results are reused to generate intermediate frames by warping the original images based on their motion properties. Blended weights are calculated by combining the confidence scores of the warped motion vector with static values, and a second neural network refines the alignment and blending of the candidate images.This second network predicts intermediate streams and generates new blend weights that are used to warp and blend the candidate images, ultimately producing a final interpolated image that improves the visual smoothness and consistency in the video stream.

This invention relates to the field of artificial intelligence image processing technology, specifically to a blurred image diffusion restoration method incorporating background-guided constraints. The method includes integrating background-guided constraints into the restoration loss and combining it with a pre-trained diffusion generation module to restore blurred images. To improve the accuracy and realism of the restoration, the similarity between the generated image and the background of the noisy image is used to guide the restoration process, forming background-guided constraints. This addresses the problem of unrealistic restored images caused by overly smooth backgrounds. To preserve more details and contour information in the image, the pre-trained diffusion generation module is used again to generate more image details based on the restored image. Finally, a blurred image restoration network incorporating background-guided constraints is proposed, which can effectively alleviate image degradation caused by noise, shooting conditions, and other factors, improving image quality. Image restoration can make the main subject of the image clearer.

The invention relates to a method of processing an image representative of a scene facing a vehicle. A light intensity map is generated (501) from a generation model configured to receive as input an image representative of a scene facing the vehicle. The determined light intensity map is transmitted (502) to the light module, for projection (503) of a pixelated light beam according to the light pattern corresponding to the generated light intensity map. A first image representative of the scene facing the vehicle is obtained (503), following the projection of the at least one pixelated light beam according to the light pattern. The method includes determining (505; 506) at least a first image processing result by applying a first image processing function to the first image and a second image processing result by applying a second image processing function to the first image.

Images of an object may be captured via a camera at a mobile computing device at different viewpoints. The images may be used to identify components of the object and to identify damage estimates estimating damage to some or all of the components. Capture coverage levels corresponding with the components may be determined, and then recording guidance may be provided for capturing additional images to increase the capture coverage levels.

1. The name of the design product: upload file graphical user interface for electronic device. 2. The use of the design product: for an electronic device. 3. The design points of the design product: in the graphical user interface of the electronic device. 4. The picture or photo that best indicates the design points: interface change state figure 3. 5. The use of the graphical user interface: the graphical user interface is an upload file graphical user interface, mainly showing users how to upload UFF files. 6. The human-computer interaction mode of the graphical user interface: after clicking the "login to continue" button in the main view, interface change state figure 1 is displayed, after inputting the account and password and clicking the login button in interface change state figure 1, interface change state figure 2 is displayed, after clicking the "upload more images" button in interface change state figure 2, interface change state figure 3 is displayed, after clicking the "upload UFF file" button in interface change state figure 3, interface change state figure 4 is displayed, after clicking the "UFF" type button in interface change state figure 4, interface change state figure 5 is displayed, after clicking the "select UFF file" button in interface change state figure 5, interface change state figure 6 is displayed. The cross in each interface represents text and / or numbers, and the gray part in each interface is the deleted content screen.

A processing method for a plurality of image sources, for a display device includes inputting a plurality of image sources; detecting the plurality of image sources according to metadata of the a plurality of image sources to determine electro-optical transfer function (EOTF) corresponding to the a plurality of image sources; and dynamically applying different EOTFs on corresponding image sources.

Provided in the present application are a multi-image processing unit core parallel image rendering method, a processing unit and a device. The method comprises: acquiring all API calls of multiple image frames to be rendered, and respectively writing all the API calls of said image frames into different target API call buffers among preset M×N API call buffers, N being a positive integer greater than 1, M being a positive integer greater than 1, each M API call buffers corresponding to one image processing unit core, and at least two of the target API call buffers corresponding to different target image processing unit cores among N image processing unit cores; parsing all API calls in each target API call buffer to obtain all rendering tasks corresponding to each of said image frames; and sending all rendering tasks corresponding to said image frames to target image processing unit cores, rendering tasks corresponding to each target image processing unit core being: all rendering tasks obtained by parsing all API calls in a target API call buffer corresponding to the target image processing unit core.

The application discloses a kind of based on FPGA's pulse neural network handwritten numeral identification system, belong to integrated circuit field.It includes: CPU / GPU end software training module, parameter fixed-point quantization and export module, test input generation module, FPGA end model prediction module and simulation verification module.From snnTorch training to FPGA inference end-to-end code closed loop, make training parameter and RTL inference structure consistent;Through.coe / .mem double format export, make parameter initialization and simulation verification repeatable, can regress;Through the unified scheduling of TOP internal state machine "loading-replay-interval-complete" rhythm and interlayer parallel conversion, make two full connection and LIF, ROM IP core and LIF update strictly aligned, facilitate timing verification and parameter tuning;Through the generation of test input generation script with label and multi-image simulation test platform, realize stable regression test before board and fault location.

This application discloses a training-free multi-image fusion method, device, and medium based on a diffusion model, relating to the field of computer vision technology. The method includes: acquiring multiple source images to be fused; simultaneously inputting the multiple source images into a diffusion model; performing a preset diffusion step on each source image to obtain a noisy image of the same size as the source image; and simultaneously performing a denoising process on the noisy image corresponding to each diffusion step until the noisy image corresponding to the 0th diffusion step is determined as the fused image of the multiple source images. This application aims to solve a series of problems in existing technologies when performing multi-image fusion, such as poor fusion effect, long processing time, and low efficiency, achieving fast and efficient multi-image fusion.

Further refinement is needed for extension applications that extend the functionality of standard drivers to provide better performance. [Solution] An application program that supports a standard driver provided by an operating system provider, characterized in that it causes the standard driver of an information processing device to perform a modification process to change the page order of multiple image data of multiple pages generated by an image editing application program to a page order for reverse printing, which prints the images in the reverse order of the original page order; and, after the standard driver has finished the modification process, it causes the operating system of the information processing device to instruct the standard driver to generate print data based on the image data.

A system for the forensic origination of digital image and video data, comprising: • an enrollment unit for capturing a multi-image set consisting of at least three reference images per physical camera, • a fingerprint extraction unit for generating a camera-specific PRNU reference fingerprint from the multi-image set, • a storage unit with an enrollment data record containing at least one enrollment identifier, the PRNU reference fingerprint, and a capture source, • an analysis unit for generating a query PRNU fingerprint from an image or video medium to be examined, • a comparison module for determining a correlation-based similarity value between the query PRNU fingerprint and the PRNU reference fingerprint, • and a classification module, characterized in that the classification module, using a first threshold value T_min, generates an unassigned output if the highest determined similarity value is below T_min.An ambiguous output is generated if the largest determined similarity value is at least T_min and at least two enrollment records have similarity values ​​that lie within a specified tolerance interval around the maximum value; and an associated output is generated if the largest determined similarity value is at least T_min and the ambiguity case does not exist.

The application discloses a two-stage zoom camera multi-image splicing method and system based on SuperGlue, and the method comprises the following steps: step one, collecting a large field angle image and completing splicing to obtain a template base map; step two, upsampling and blocking the template base map to generate template base map blocks; step three, calculating the position of a shooting point, automatically collecting a clear image, and making all the collected clear images cover a specified area; step four, selecting a clear image and a template base map block, performing feature matching, calculating a homography matrix, and transforming the clear image; step five, calculating the optical flow between the clear image after transformation and the template base map covering the area; step six, performing pixel-by-pixel registration on the clear image after transformation according to the optical flow, so that the clear image is aligned with the template base map; and step seven, transforming all the clear images to the coordinate system of the template base map, and fusing the images. The image splicing method has good splicing efficiency and splicing accuracy.

The article to which the design pertains in this design registration application (hereinafter referred to as "the Application") is an image generation GUI using AI technology. The image diagram shows the state in which five images with a 1:1 aspect ratio have been generated by inputting prompts. The user can generate further images, as shown in the changed image diagram, by clicking the "edit" button located below the generated images and entering additional prompts. The changed image diagram shows the state in which five images have been generated based on the first image generated and the additional prompts. In each generation phase, the number of images generated, the aspect ratio, and other conditions of the generated images will vary depending on the input prompts, so the aspect ratio is not limited to 1:1 and the number of generated images is not limited to five.