136 results about "Image transformation" patented technology

Disclosed are systems and methods including software processes for developing test cases for testing robustness of AI-based image-recognition models-under-test (MUTs) with respect to types of image variation transformations. The system may generate various types of robustness metrics for the MUT and output user-readable reports about the MUT's performance. The system trains machine-learning architectures to generate test cases including augmented images according to the types of image transformations, applies the IR MUTs, and then evaluates the image feature vector embeddings and predicted classification produced by the IR MUTs to determine the accuracy of the MUT with respect to each type of transformation.

A method for extracting at least one image feature from at least one image generated by at least one real camera, present in a motor vehicle, for monitoring the environment of the motor vehicle. The method includes, in a first step a), the image generated by the camera is transformed into a transformed image by means of a virtual camera which is assigned to the real camera and the optical axis of which extends along a horizontal direction with respect to the motor vehicle, according to which, in a second step b), at least one image feature is extracted from the transformed image.

The invention provides a continuously constrained and difficulty guided passive domain adaptive three-dimensional medical image segmentation method. The method comprises the following steps: in a source domain pre-training stage, carrying out full-supervised training on a segmentation model by utilizing source domain annotation data; in the pseudo source domain image generation stage, a thought of combining coarse generation and fine generation is adopted, style migration is performed by using a frozen source domain pre-training segmentation model and target domain unlabeled data in coarse generation, and a target domain image is converted into a pseudo source domain image with a source domain style; and in the fine generation step, Fourier transform is utilized to remove artifacts and noise in the coarsely generated image. In the target domain adaptation stage, a pre-training segmentation model, a pseudo source domain image and a target domain image are utilized, and continuity constraint between slices and a difficult sample mining mechanism are fused to carry out an adaptation process from a source domain to a target domain. According to the method, under the condition that source domain data does not need to be accessed, the spatial context constraint and the difficult sample mining mechanism of the three-dimensional medical image are effectively fused.

Systems, methods, and computer-readable media are provided for using generative AI enriched with metadata about historical document characteristics to transform documents of various formats, including images, to the fields and values they represent. A prompt template may be selected in association with a type of document. The prompt template indicates field definition(s) of field(s) to be detected in the document and location(s) in which the field(s) have been detected in prior documents. A large language model is prompted with a prompt generated using the prompt template to generate a result that assigns value(s) to the field(s). Output from the language model is used for identifying the field to value mapping for the document, such that data detected from the document may be stored in appropriate database structures of a database. Metadata stored in association with the prompt template is updated based on location(s) in the document in which the field(s) were detected, and the value(s) of the field(s) are stored in a database. Outbound documents may be similarly translated to detect values of corresponding fields requested by third parties, even if those values are not stored in the database. In this scenario, values for fields may be detected in outbound documents using the prompt templates enriched with metadata as processed by the large language model before such information is prepared to be sent to a third party.

Fisheye images that include objects at first, second, and third angles into rectilinear images are transformed with a first image transformation and the rectilinear images are transformed into bird's eye view images with a second image transformation. The bird's eye view images can be transformed into multiple images that include objects at multiple angles intermediate between the first, second, and third angles to generate a training dataset that includes ground truth regarding the multiple angles with a third image transformation. A machine learning model can be trained with the training dataset. A machine such as a vehicle can be operated with output from the machine learning model.

83 20 25 28 74 03 30 D ec 2 02 5 A B S T R A C T 2 0 2 5 2 8 7 4 0 3 3 0 D e c 2 0 2 5 8 3 IL LU M IN AT IO N PA TT ER N (S ) 11 0 A SE CO ND T YP E OF IM AG ES (E .G ., FL UO RE SC EN CE IM AG ES ) 12 4 A FI RS T TY PE O F IM AG ES (E .G ., BR IG HT -F IE LD IM AG ES 12 2 DI SC RI MI NA TO R 10 4 GE NE RA TO R 10 2 Tr ain in g Da ta 12 0 GA N Mo de l 1 00 FI G. 1 1 / 26 Training Data 120 GAN Model 100 ILLUMINATION A FIRST TYPE OF IMAGES (E.G., PATTERN(S) BRIGHT-FIELD IMAGES 110 122 GENERATOR 102 A SECOND TYPE OF IMAGES (E.G., DISCRIMINATOR FLUORESCENCE IMAGES) 104 124 FIG. 1 20 25 28 74 03 30 D ec 2 02 5 2 0 2 5 2 8 7 4 0 3 3 0 D e c 2 0 2 5 T r a i n i n g D a t a 1 2 0 ILLUMINATION A FIRST TYPE OF IMAGES (E.G., P A T T E R N ( S ) BRIGHT-FIELD IMAGES A SECOND TYPE OF IMAGES (E.G., FLUORESCENCE IMAGES)

The invention relates to the technical field of image processing, and discloses an image optimization method and device, equipment and a medium, and the method comprises the steps: recognizing the degradation information of a target image, carrying out the image transformation processing of the target image according to the image degradation information and a preset image transformation process, and obtaining an optimized image, performing multi-type visual task detection on the optimized image, generating a quantized image quality index, judging whether the image reaches a quality standard by comparing a detection parameter with a preset threshold value, if not, calculating a reward signal of a reinforcement learning agent model according to a detection result, and updating an image transformation process by using the signal, so as to improve the quality of the image. And then returning to the optimization step to process the image again, if the current strategy reaches the standard, indicating that the current strategy is effective, directly processing the subsequent to-be-processed image by the agent by using the updated strategy, and finally outputting a high-quality target optimized image. According to the invention, the efficiency and precision of image processing are improved.

An information processing device includes an image transformation unit, a left-right difference estimation unit, and an image generation unit. The image transformation unit performs warping to move positions of a feature point of a right-eye image and a feature point of a left-eye image based on right-eye and left-eye viewpoint information. The left-right difference estimation unit estimates a portion where a difference exceeding an allowable level occurs, due to the warping, between the right-eye image and the left-eye image as an inconsistent portion. The image generation unit makes a sharpness of the inconsistent portion different between the right-eye image and the left-eye image.

The present application provides a kind of image rain and fog removal method, device, equipment and storage medium of railway scene, the method comprises: obtaining the target image of railway scene under rain and fog to be removed;Based on the target image of rain and fog to be removed, the preprocessed image of target image is generated;The preprocessed image is input into target Unet network, and the high-resolution image of target image is encoded and decoded by target Unet network Process, and the target image of rain and fog removal is obtained by pixel by pixel inflation filtering and fusion operation;Wherein, target Unet network is obtained by training initial Unet network through paired training sample;Paired training sample includes multiple rain and fog free sample images under railway scene and the superimposed rain and fog image corresponding to each rain and fog free sample image generated based on image transformation method.The present application realizes that a large number of paired training samples are generated based on image transformation method, improves model training effect, and then improves the rain and fog removal effect of model.

A system and method for detecting and diagnosing faults in a building automation system (100). Time-series data (202) are received from the building automation system (100). Label rationality (214, 216, 218, 220, 222) is determined for each set of time-series data and the corresponding label associated with that set, based on a tree-based classifier (224) and an image transformation classifier (226). The tree-based classifier (224) and the image transformation classifier (226) receive the same data input and operate independently of each other.

This invention discloses a data generation method for correcting rotational nonuniformity of endoscopic probes, relating to the field of medical image processing technology. The method includes: selecting an acquired image composed of multiple scan line data; performing a random image transformation on the acquired image to obtain a reference image; simulating diverse nonuniform rotational distortion patterns using a mathematical model to randomly synthesize an offset vector corresponding to the number of line data in the reference image; applying the offset vector to the reference image, obtaining corresponding line data based on the offset value at each position, and arranging the newly arranged line data to form a distorted image; and generating a distorted-reference image data pair with known labels from the reference image and the distorted image. This invention can meet the quantitative and diverse requirements for training data in nonuniform rotational distortion correction at extremely low cost, making the training set more complete and exhibiting good applicability and generalization for correcting endoscopic probe imaging of different modalities and systems.

The application provides a large-scale satellite image automatic registration and enhancement method based on adaptive multi-source fusion, comprising: performing orthorectification and multi-resolution unified projection on a multi-source original data set obtained to obtain a coarse registration image set; calculating a local tensor structure of the coarse registration image set, performing repeated texture recognition, generating a final structure representation and a structure reliability; performing regional registration to obtain a full-image dense deformation field, then performing image transformation to be aligned and registration uncertainty estimation to obtain a high-precision registration image and structure registration uncertainty estimation; combining a pseudo-change probability of the high-precision registration image and the structure registration uncertainty estimation to obtain a change reliability weight; under the constraint of the change reliability weight, performing consistency enhancement on the high-precision registration image to output a quantitative and reliable enhanced image. The application realizes high-precision, low-false alarm and quantifiable and reliable automatic registration and enhancement of multi-source large-scale satellite images.

The application discloses a temperature measurement error attribution analysis method and device for an infrared thermal imager. The method comprises the following steps: based on a preset temperature value set of a blackbody radiation source, temperature measurement of the blackbody radiation source is performed by using the infrared thermal imager to obtain a measurement temperature set; image transformation processing is performed on the measurement temperature set to obtain a measurement temperature image set; temperature measurement error attribution analysis processing is performed on the measurement temperature image set to obtain temperature measurement error region information of the infrared thermal imager.

Described are systems and methods for training a machine-learning model to generate image of biological samples, and systems and methods for generating enhanced images of biological samples. The method for training a machine-learning model to generate images of biological samples may include obtaining a plurality of training images comprising a training image of a first type, and a training image of a second type. The method may also include generating, based on the training image of the first type, a plurality of wavelet coefficients using the machine-learning model; generating, based on the plurality of wavelet coefficients, a synthetic image of the second type; comparing the synthetic image of the second type with the training image of the second type; and updating the machine-learning model based on the comparison.

This invention belongs to the field of image processing technology, specifically relating to an image stitching method for a low-altitude unmanned aerial vehicle (UAV) monitoring network. The method includes: acquiring a reference image and an image to be stitched; extracting matching point pairs between the two images; determining the static confidence level of the matching point pairs based on their disparity changes; extracting the static matching point pairs and performing multiple linear transformations to obtain multiple static projection matrices; selecting the optimal static projection matrix for image transformation; and extracting dynamic targets within the overlapping area. When generating the stitched image, the gray values ​​of the regions identified as dynamic targets in the reference image are replaced with the gray values ​​of the same regions in the image to be stitched, thus completing the image stitching. This invention, through a dynamic-static separation strategy, effectively avoids the contamination of geometric transformation calculations by dynamic targets, significantly improving stitching accuracy and visual quality.

This invention relates to the field of image recognition, specifically to an automated tunnel construction monitoring system based on image data. The system includes an edge computing node that acquires the original tunnel image and constructs an atmospheric scattering physical model using measured data from an optical dust concentration sensor. The edge computing node transforms the original image to the frequency domain, extracts low-frequency brightness features and high-frequency texture features using discrete cosine transform, and adaptively compensates the high-frequency texture features with weights based on the transmittance matrix calculated by the physical model to suppress diffuse light interference. The compensated features are then input into a lightweight convolutional neural network via inverse transform. This scheme introduces dust physical parameters as prior constraints into the frequency domain feature weight calculation, reducing the probability of image edge artifacts and color distortion under conditions of localized strong light and high dust, ensuring the physical authenticity of target edge features in the reconstructed image, and improving the accuracy of feature extraction under harsh conditions.

The invention provides an image registration method and electronic equipment, which can be applied to the fields of remote sensing, computer vision, digital image processing, artificial intelligence and embedding. The method comprises the following steps: according to respective first positions of a plurality of first modal image blocks of a first modal image and respective second positions of a plurality of second modal image blocks of a second modal image, obtaining a plurality of first modal image blocks of the first modal image; determining matched candidate first modal image blocks and candidate second modal image blocks from the plurality of first modal image blocks and the plurality of second modal image blocks to obtain a plurality of candidate matching pairs; processing the plurality of candidate matching pairs by using a target student model to obtain a plurality of first modal feature vectors and a plurality of second modal feature vectors; and obtaining an image registration result indicating image transformation information between the first modal image and the second modal image according to the plurality of first modal feature vectors and the plurality of second modal feature vectors.

This application discloses a visual tracking method, apparatus, device, and medium for industrial robots, relating to the field of industrial robot technology. The method includes: acquiring a continuous image stream and global motion data of the industrial robot's end effector camera; performing optical flow tracing on the current frame and the previous frame to obtain feature point displacement vectors, and calculating mechanical vibration energy indices based on these vectors; adaptively adjusting a smoothing factor based on the vibration indices, filtering the global motion data to obtain a stable image transformation matrix, and performing a geometric transformation on the current frame image to generate a stable image; inputting the stable image into a Transformer architecture detection model, and obtaining the detection result through cross-attention calculation; extracting the target feature vector of the current frame based on the target bounding box, performing dual checks of occlusion gating and consistency gating, executing a corresponding memory update strategy, and outputting the target tracking result. This method can achieve accurate tracking.

Fisheye images containing objects at first, second, and third angles are transformed into rectilinear images using a first image transformation. These rectilinear images are then transformed into bird's-eye views using a second image transformation. The bird's-eye views can be further transformed into multiple images containing objects at various angles between the first, second, and third angles using a third image transformation. This generates a training dataset that provides ground truth regarding objects at multiple angles. A machine learning model can then be trained on this dataset. Finally, a machine, such as a vehicle, can be operated using output from this machine learning model.

This invention discloses an event camera calibration method based on wavelet transform, which can quickly extract feature points and improve calibration accuracy and speed. The target, displayed as a striped pattern on an LCD screen, serves as the two-dimensional calibration target, showing both horizontal and vertical stripe images. The screen flashes at a certain frequency to trigger events. During calibration, the event camera acquires and captures horizontal and vertical stripe event stream data. An appropriate time window is selected, and the event stream is accumulated into two-dimensional event frames, which are then converted into image information. Wavelet transform is then used to calculate the wrap-around phase map in both directions. The feature points are located at the 2π phase value between the horizontal and vertical wrap-around phases, thus obtaining the pixel coordinates of the feature points. This invention uses image transformation technology to extract the phase map, eliminating reliance on image intensity for feature point detection and achieving accurate feature point extraction.

The present application belongs to the technical field of image processing, and particularly relates to an airborne imaging non-uniform low-light scene image enhancement method. The method comprises the following steps: S1, processing a non-uniform low-light scene image to obtain H, S and V channel images; S2, performing dark area correction calculation on the V channel image to obtain a first image; S3, performing bright area suppression calculation on the V channel image to obtain a second image; S4, fusing the first and second images to obtain a fused image; S5, inputting a pixel distribution optimization image into a nonlinear image transformation tone adjustment model for processing to obtain a mid-term enhancement image; and S6, performing pixel value range stretching processing on the mid-term enhancement image, and combining the H and S channel images to obtain an optimized enhancement image. The present application can effectively improve the visibility of the dark area of the image and retain the detail feature information of the bright area while retaining the color richness and level of the image through the regional differentiation calculation and adaptive fusion mechanism.

This invention provides a method and system for automatic vertex picking of DAS-based microseismic events, relating to the field of microseismic location technology. The method includes acquiring multi-channel data corresponding to microseismic events and performing demodulation analysis to form multi-channel waveform data; performing correlation analysis on the waveform data from different channels to establish an automatic picking correlation model; performing image transformation based on the automatic picking correlation model to form corresponding grayscale images; and performing location discrimination analysis based on the grayscale images to determine the event vertex channel. This method improves the interpretation accuracy and reliability of geological analysis of seismic images by removing noise while preserving geological structural features such as faults and strata.

The invention provides a training method of an optical character recognition model, a recognition method of the optical character recognition model, equipment and a storage medium. The training method of the optical character recognition model comprises the following steps: synthesizing a document by using text content; rendering the element simulation printing display effect in the document into a canvas; performing image transformation of various simulation degeneracy on the canvas; if the image conversion is completed, performing post-processing on the canvas to obtain formatted image data; an optical character recognition model is trained according to the image data and the text content. According to the embodiment of the invention, the project is extensible and supports the generation of high-concurrency data, the distribution difference between the synthetic data and the real data can be remarkably reduced by rendering the simulation printing display effect and the simulation degeneration, and the recognition accuracy of the optical character recognition model in an actual scene is improved.

Methods, systems, and apparatus for receiving a request to perform a plurality of image transformation operations on an input image to generate a plurality of output image frames. An identifier is assigned to each respective image transformation operation and a corresponding output image frame. Image data for the input image is obtained from memory and stored in a local cache. An output image frame portion is generated based on a portion of image data for the input image stored in the local cache and an image transformation operation corresponding to the identifier of the respective output image frame. The output image frame portion is stored in memory within a separate container associated with each of the plurality of output image frames.

This invention provides an image processing device, an image processing method, and an image processing program that enable individuals with color vision deficiency to distinguish between different colors through simulation and clustering. [Solution] The image processing device is an image processing device that converts input image data input to an input device into output image data within a set color gamut on an output device, and includes: a simulation image generation unit that performs a dichromacy simulation on the input image data in the input device to generate a color vision simulation image; an input color information conversion unit that converts the simulation image generated by the simulation image generation unit into an XYZ color system; a confusion color information storage unit that stores confusion color information relating to confusion color lines that show the distribution of colors that are difficult for colorblind persons to distinguish; and a clustering processing unit that performs clustering to classify multiple similar colors and aggregate them into a representative number of colors.

The application provides an arrow picture data augmentation method, which comprises the following steps: obtaining a horizontal homography matrix and a field of view transformation matrix of a target camera based on intrinsic parameters, extrinsic parameters and preset inverse perspective transformation parameters of the target camera; obtaining a perspective transformation matrix and an inverse perspective transformation matrix based on the horizontal homography matrix and the field of view transformation matrix, and transforming a first front view into a bird's eye view according to the perspective transformation matrix; transforming the types of arrows in the bird's eye view according to the aspect ratio characteristics of multiple types of arrows; and transforming the bird's eye view after the type transformation into a second front view based on the inverse perspective transformation matrix. The application transforms a front view into a bird's eye view, then transforms multiple types of arrows in the picture in the bird's eye view, and then transforms the bird's eye view after the transformation into a front view, so that the transformation conforms to the front picture imaging rule when the front view is locally transformed, and the generalization ability of a deep learning model is improved through local image transformation.

The application relates to a motor bearing fault identification method, device, equipment, medium and product, wherein the method comprises the following steps: converting a current signal of a motor bearing to be identified into a two-dimensional Markov image by using a Markov transform field; processing the two-dimensional Markov image by using a student model to obtain an identification result of the motor bearing; the identification result is used to represent the probability that the motor bearing belongs to a preset state category; and the identification result is obtained by converting the two-dimensional Markov image into an initial capsule vector by the student model, and performing dynamic routing inference based on a preset sparse ratio and the activation intensity of the initial capsule vector. By using the above method, the accuracy, interpretability and robustness of the identification result of the running state of the motor bearing can be improved.

This invention discloses a plug-and-play model inversion attack method based on a generative model in a collaborative reasoning environment. This method uses a pre-trained StyleGAN2 as the target-independent image prior, leveraging its mapping and synthesis networks in the generator structure to map latent vectors into intermediate representations to generate images. During the attack optimization process, to address the gradient vanishing problem easily caused by traditional cross-entropy loss, a Poincaré loss function is introduced, utilizing the gradient preservation properties of non-Euclidean space to ensure the stability of the optimization process. Simultaneously, through standard image transformation and random image transformation techniques, the difference between the target data distribution and the image prior is effectively reduced, enhancing the robustness of the generated features. This method effectively overcomes limitations such as high computational resource consumption, insufficient flexibility, and sensitivity to changes in dataset distribution. In high-resolution scenarios such as face recognition, it demonstrates outstanding attack efficiency and generated image quality.