143 results about "Image rotation" patented technology

A method for decoding an image includes receiving information on image partitioning for the image included in a bitstream; obtaining a plurality of image partitions included in the image based on the information on image partitioning; and decoding the plurality of image partitions. Each image partition comprises a plurality of default encoding parts, each default encoding part is selected from a plurality of candidate default encoding parts, each candidate default encoding part is composed of one or more encoding sub-units. The bitstream includes information on rotation of the image.

The invention discloses a vehicle gauge chip integrated packaging manufacturing process based on an etching technology. The vehicle gauge chip integrated packaging manufacturing process comprises the following steps: step 1, designing and preparing a substrate; step 2, preprocessing the chip; step 3, chip mounting and interconnection are carried out; step 4, plastic packaging and molding; 5, heat dissipation and reinforcement are carried out; step 6, image rotation correction and defect detection; step 7, performing post-working procedures and testing; high-density interconnection such as TSV (through silicon vias), RDL (redistribution layer) and micro bumps is realized by using an etching technology, the interconnection density is remarkably improved, the RDL wiring layer manufactured by using the etching technology has the line width less than 2 microns, high-density interconnection on the surface of a chip is realized, the integration level is further improved, the micro-channel is manufactured in the plastic package layer by using the etching technology, and the packaging efficiency is improved. And the copper radiating fins are embedded, so that the radiating efficiency is effectively improved, the problem of local heat accumulation caused by integration of the power chip and the logic chip is solved, a gap between the chip and the radiating cover plate is filled with a thermal interface material with high thermal conductivity, and the thermal resistance is reduced.

The invention relates to an image rotation angle recognition and correction method and system, and relates to the technical field of image processing, and the method comprises the steps: collecting an original image, marking and rotating a corresponding image rotation angle, obtaining a corrected image, and constructing a training set; selecting a corresponding neural network according to the model architecture, and constructing an initial regression model in combination with a preset improved loss function; rotating the corrected image according to a random angle, constructing a training sample, and iteratively training the initial regression model in combination with the training set to obtain a rotation angle recognition model; recognizing and judging the original image through the rotation angle recognition model, and outputting an image rotation angle; rotating the original image according to the image rotation angle to obtain a corrected image; the recognition of any rotation angle of the image is realized by improving the L1 loss function, and the robustness and accuracy of the rotation angle recognition model are improved.

The invention provides an anchor-frame-free remote sensing image rotating target detection method based on angle attention, and belongs to the technical field of computer vision and remote sensing image processing. The method comprises the following steps: inputting a remote sensing image and extracting multi-scale features by using a backbone network; an optimal angle feature is extracted and selected through an angle feature extraction module and a feature pyramid network; a detection head is added to multi-scale output of the feature pyramid network for classification, frame regression and angle parameter regression; a frequency domain channel attention mechanism is introduced into an angle regression branch through an angle prediction head module in a detection head, so that the angle of a rotating target is predicted more accurately; and carrying out joint training by utilizing ProbIoU loss and distributed focus loss to realize end-to-end high-precision rotating target detection. According to the method, relatively high detection precision and reasoning speed are realized on a plurality of public remote sensing image data sets, and the method has efficient performance.

An embodiment provides an electronic device comprising: a projector; a light guide device that guides light emitted from the projector; and an image rotation element provided between the projector and the light guide device, in which the image rotation element rotates light emitted from the projector about a longitudinal direction of the image rotation element. An embodiment provides a projection apparatus including: a light source unit disposed in a housing; a light modulator disposed at one surface of the housing; a lens unit disposed at a rear end of the light source unit; a reflection unit disposed between the light source unit and the light modulator; a projection lens unit; and a prism provided between the light modulator and the projection lens unit, in which the light source unit, the lens unit, and the reflection unit are sequentially provided and arranged to be inclined with respect to the light modulator or a top surface of the housing.

In a portable information terminal to which an image projection function can be added, there is provided a technique for improving visibility of a projected image. The portable information terminal has a terminal body and an attachment attached to the terminal body and communicatively connected to the terminal body. The attachment has a projector for projecting an image based on image information obtained from the terminal body, and is configured such that a projection opening direction of the projector is variable with respect to the terminal body. The terminal body or the attachment identifies a state of the portable information terminal including a position of the attachment with respect to the terminal body. The terminal body or the attachment performs a projection control of the image based on the identified state. The projection control includes, for example, an image rotation control, an image shape control, or a projection angle control.

A spin-estimation system may include an image-capturing sensor positioned and configured to capture images of an object within a field of view of the image-capturing sensor. The spin-estimation system may be configured to perform one or more operations to analyze spin properties of the object. The operations may include setting an image capture framerate that corresponds to a minimum spin motion of the object, printing an orientation marker on an outer surface of the object, and capturing, by the image-capturing sensor at the set image capture framerate, images of the object after starting motion of the object. The operations may include isolating the object in each image to generate isolated object images. The operations may include generating an object marker segmentation map based on the isolated object images. A spin rate and a spin axis may be estimated based on the object marker segmentation map using deep learning approaches.

The invention relates to the field of infrared thermal imaging, in particular to a miniaturized medium-wave infrared curved surface bionic compound eye optical system which comprises a curved surface multi-aperture array, an optical relay image rotation unit and an infrared detection unit. The curved surface multi-aperture array is of a compound eye structure formed by a plurality of sub-eyes with the same parameters. And the number of the lenses contained in each sub-eye is not more than 3. The optical relay image rotation unit converts a focal curved surface image of the curved surface multi-aperture array into a focal plane image, and the focal plane image is collected to the infrared detection unit. The number of lenses contained in the optical relay image rotation unit is not more than 6, and the optical relay image rotation unit has an object curved surface with the same curvature radius as the curved surface multi-aperture array and is matched with a diaphragm of the infrared detection unit. The system has the advantages of multiple compound eye apertures, light weight, miniaturization, large relative aperture, large target surface size, large field of view, high spatial resolution, cold diaphragm matching and the like, can perform operations such as quick capture, positioning and distance measurement on a moving target, and is particularly suitable for application scenes such as high-altitude over-the-ground wide-area detection, quick moving target positioning and tracking and the like.

The invention belongs to the field of natural resource monitoring, and discloses a spherical camera image set parameter resolving method, a medium and electronic equipment, and the method comprises the steps: collecting images, and extracting feature points in each image; selecting two images with the maximum overlapping area as an initial pair for processing; gradually introducing new images, and calculating a rotation parameter of each newly introduced image relative to the processed image by using a matching relationship of homonymy points; after the image rotation parameters are solved, selecting a plurality of control points, performing global constraint on the whole image set by using the control points, and performing corresponding matching on known positions of the control points and feature points in the image; and calculating final positions and angle parameters of all the images through a least square method. According to the invention, the field workload of the control point can be obviously reduced, the image positioning precision of the spherical camera is improved, and the spherical camera can be compatible with mainstream spherical camera equipment in the market.

The application relates to the technical field of SAR data processing and target detection, and discloses a SAR image speckle perception rotating target detection method based on physical priori, which comprises the following steps: firstly, collecting multi-source SAR image data and carrying out pretreatment, and preliminarily separating targets and background noise through low-rank sparse decomposition; combining the equivalent noise layer and contrast of the SAR image and other physical priori, constructing a conditional diffusion model, modeling the rotating frame prediction as a conditional diffusion denoising process, and guiding the step-by-step optimization of the frame parameters by using the physical priori; introducing a Gauss-Wasserstein distance as a regression loss, combining a shape consistency constraint, and effectively relieving the angle period discontinuity and deformation problem; finally, further optimizing the detection result through a geometric consistency constraint, and outputting a high-precision rotating target detection image after post-processing enhancement. The method significantly improves the precision, robustness and physical consistency of the SAR image rotating target detection in a complex scene.

The application discloses a kind of coronal mass ejection detection methods based on low rank restoration model, comprising the following steps: step 1, to the LASCO C2 image downloaded from SOHO homepage is carried out image preprocessing, to coronal image is denoised, screening processing;Step 2, the LASCO C2 image obtained after step 1 is carried out image gray, image rotation and flip operation;Step 3, the coronal image after processing in step 2 is converted to polar coordinates and is represented;Step 4, the moving target in coronal image is extracted using the method based on low rank restoration model;Step 5: fine-tuning low rank restoration model parameters, coronal mass ejection is distinguished from coronal stream, using image three frame difference is carried out moving target extraction, finally obtains coronal mass ejection.The method detection speed is fast, and precision is greatly improved.

A visual inspection system includes a processor, a rotation stage, an adaptor, a light source, a telecentric lens and an image sensor. The adaptor is connected to the rotation stage and located under the rotation stage for positioning the lyophilized bead. The light source is located at a lateral side of the lyophilized bead for emitting lights to illuminate the lyophilized bead. The telecentric lens is located at the lateral side of the lyophilized bead for imaging the lyophilized bead. The image sensor is located at a lateral side of the telecentric lens opposite to the lyophilized bead for capturing images of the lyophilized bead. The rotation stage drives the lyophilized bead to rotate 360 degrees as performing an inspection of the lyophilized bead, and the image sensor continuously captures images of the lyophilized bead at different angles, thereby obtaining a three-dimensional image of the lyophilized bead for the inspection.

The invention discloses an image rotation processing chip, a method thereof, equipment and a medium. The image rotation processing chip comprises a memory unit; the decoder compresses the initial image through a built-in compression module and writes an obtained original compressed code stream into the memory unit; the code stream rotation unit reads the original compressed code stream from the memory unit, performs image rotation processing on the original compressed code stream, and writes the obtained rotation compressed code stream into the memory unit; and the display unit reads and decompresses the rotary compressed code stream from the memory unit through a built-in decompression module. The code stream rotation unit can read the original compressed code stream from the memory unit and perform image rotation processing on the original compressed code stream to obtain the rotation compressed code stream, so that the code stream rotation unit in the image rotation processing chip does not need to be internally provided with a compression module and a decompression module, and the cost of the image rotation processing chip can be reduced.

The invention provides a remote sensing image rotating target detection method and system based on a joint attention mechanism, and the method comprises the steps: carrying out the feature extraction through employing a convolutional neural network based on obtained remote sensing image data, and obtaining a multi-scale feature map; performing semantic space coupling processing on the multi-scale feature map through a joint attention mechanism to obtain a target orientation feature descriptor, and performing feature refinement on the multi-scale feature map to obtain a refined feature map; a rotation candidate frame is generated according to the refined feature map, adaptive sample screening and rotation alignment processing are performed on the rotation candidate frame, an isovariant geometric convolution detection head is utilized to predict features after rotation alignment, and category information of the target and rotation bounding box parameters are output; according to the method, by combining an attention mechanism, isovariant geometric convolution and adaptive sample screening, the direction and geometric features of the target in the remote sensing image can be captured, the real-time reasoning efficiency is considered, and reliable technical support is provided for key tasks such as maritime affair supervision.

The invention is suitable for the technical field of image recognition, and provides a method and device for recognizing unsealing of a chip, and the method comprises the steps: collecting a current real-time image, rotating the current real-time image to a preset direction, and obtaining a current chip image; obtaining a region of interest in the standard template image; wherein the region of interest comprises a bare chip region, a wire bonding region and / or a pin region; extracting a to-be-identified region corresponding to the region of interest in the current chip image; and controlling a chip unsealing process according to the region similarity between the region of interest and the region to be identified corresponding to the region of interest. According to the scheme, high efficiency, accuracy and automation in the chip unsealing process are achieved, and the unsealing efficiency and the recognition accuracy are remarkably improved.

The present disclosure provides a continuous zoom stereoscopic microscope with an adjustable stereoscopic angle, consisting of a microscope stand, a first eyepiece module, a second eyepiece module, a first objective module, a second objective module, a first Risley prism, a second Risley prism, a first image-rotating prism, a second image-rotating prism, a drive module, a control module and an illumination module. The drive module provides preset drive values for individual liquid lenses in the liquid lens sets according to different magnifications to change the focal lengths of the liquid lenses, thereby changing the effective focal lengths of the objective module and that of the eyepiece module, and finally achieving continuous and fast zooming of the stereoscopic microscope to be adapted to different working scenarios. The control module controls the relative angle between the two wedge-angle prisms in the Risley prism, achieving the continuous adjusting of the stereoscopic angle of the stereoscopic microscope, and then acquiring stereo images with different stereo senses.

The application discloses a visual rich-text page restoration method based on an attention network, which comprises the following steps: inputting a document image into a trained document image rotation angle classification network to obtain the angle at which the document image needs to be rotated and correct the document image, and then using a projection correction algorithm to perform secondary correction on the document image; performing element detection on the document image through a page analysis network and performing corresponding processing to obtain element content corresponding to each element; using a plurality of deep learning algorithm network models connected in series to restore the document image into an editable document; and adding a contrast denoising training strategy in the page analysis network training to improve the accuracy of page analysis, because the plurality of network models are connected in series, according to the rule of the weakest link, the effect of page restoration will be affected by the page analysis model, therefore, the addition of the contrast denoising strategy can improve the accuracy of the network, promote the convergence of the network and shorten the training time.

The application discloses a remote sensing image super-resolution reconstruction method based on weight sharing and multi-scale multi-expert decision-making, which comprises image rotation, super-resolution expert network, super-resolution result of the rotated super-resolution expert network, average of four super-resolution expert network super-resolution results, feature vector splicing and fusion expert network; the image rotation comprises the following steps: after inputting an original low-resolution remote sensing image, aligning and rotating 0 degrees, 90 degrees, 180 degrees and 270 degrees, as shown in the following formula: I0=R O t0(I), I 90 =R O t 90 (I), I 180 =R O t 180 (I), I 270 =R O t 270 (I), wherein I is the original low-resolution remote sensing image. The super-resolution reconstruction system can solve the problems of the limitation of single expert decision-making and the imbalance of model parameter quantity and effect, so that the super-resolution reconstruction mechanism of multi-expert decision-making and the multi-scale weight sharing convolution module are introduced in the super-resolution reconstruction process, and more abundant detailed texture information can be obtained by using the mechanism of multi-expert decision-making.

The invention discloses an image data rotation optimization method and device, and the method is applied to a data rotation optimization network, and the method comprises the steps: obtaining packaged image data through an application layer, processing the image data into initial color space data containing brightness data and chrominance data, converting the initial color space data into a pointer type through a bridging layer; the code layer performs block matrix transposition operation on the brightness data and performs block matrix transposition and chrominance calibration processing on the chrominance data at the same time, and rotated brightness data and rotated chrominance data are obtained respectively; and combining the rotated brightness data and the rotated chrominance data into processed color space data in a code layer, returning the processed color space data to an application layer through a bridging layer, and outputting rotated image data. According to the invention, the overall processing duration of image rotation is reduced by adopting the block matrix while the matching between the bright and dark contours of the image and the color is ensured.

The invention discloses an auxiliary Alzheimer's disease assessment method and system based on a clock drawing test, and relates to the technical field of Alzheimer's disease auxiliary assessment. Firstly, a unified spatial reference direction is determined according to a subject drawing posture, rotation correction is performed on a test image, and the influence of posture difference on a spatial analysis result is reduced; and on the basis, performing partition processing on a clock face region, and constructing a lateral aggregation index through a digital quantity relationship in each partition, so as to quantitatively represent the abnormal degree of digital spatial distribution. Meanwhile, drawing duration is introduced into an evaluation process, and an evaluation threshold value is dynamically adjusted by comparing the drawing duration with a reference duration range, so that a judgment standard can adapt to drawing behavior characteristics of different subjects. According to the method, drawing result characteristics and drawing process information are taken into consideration, the stability and consistency of evaluation results are improved, and the objectivity and reliability of a clock drawing test in assisting Alzheimer's disease evaluation are improved.

This invention discloses an image angle calculation method and system based on multiple candidate matching points. The method includes: placing a wheel hub of the same model as the one to be tested on a conveyor belt, aligning the center line of one spoke of the wheel hub with the horizontal line; taking a picture of the wheel hub to obtain a standard wheel hub image; placing the wheel hub to be tested on the conveyor belt and taking a picture of the wheel hub to be tested to obtain an image of the wheel hub to be tested; extracting feature point sets M and N from the background-removed standard wheel hub image and the image of the wheel hub to be tested, respectively; and obtaining any feature point M from M based on M and N. i Candidate matching point set P i M i With P i The feature points in the data form candidate matching point pairs; the feature point M is calculated. i The angle difference of candidate matching point pairs is used; the angle differences of candidate matching point pairs of all feature points in the feature point set M are clustered, and the mean of all angle differences in the class containing the most angle differences is calculated as the rotation angle. This invention can obtain more correct matching point pairs and obtain the correct image rotation angle through clustering.

A heads-up display unit and a method of adjusting a projected image with the heads-up display unit comprising projecting, by a projector, an image comprising an image ray toward a combiner. The method further comprising splitting, by a splitting element, the image ray before the image ray hits the combiner. The method further comprising reflecting, by a freeform mirror, the image ray or the split image rays before the image hits the combiner. The method further comprising projecting the split and reflected image rays onto the combiner.

The application discloses a camera module (1) of a field-of-view change camera, which can acquire high-quality images without losing the number of pixels. The camera module (1) comprises a field-of-view change mechanism (10), a lens (40), an image sensor (30) and an image rotation correction mechanism (20). The image rotation correction mechanism (20) is arranged between the field-of-view change mechanism (10) and the lens (40), and the image rotation correction mechanism (20) corrects the rotation of an image incident on the image sensor (30).

The application relates to a target tracking method for a high-speed rotating platform, which comprises the following steps: acquiring an initial frame image for target tracking; acquiring a next frame image, performing image rotation according to a rotation angle existing in the initial frame image, and obtaining a rotated image eliminating the rotation angle; taking a target point coordinate of a previous frame image as a center, adopting a correlation filtering model to perform target detection in the rotated image, and obtaining a target point coordinate in the rotated image; converting the target point coordinate in the rotated image to a target point coordinate in a pre-rotated image, and taking the target point coordinate in the pre-rotated image as a center to obtain a search region; updating the correlation filtering model, and performing target tracking on a next frame image; and further performing target tracking on all subsequent frame images. The application realizes stable tracking in the high-speed rotating process of the image.

The utility model relates to a coal mine inspection robot which comprises a robot body. A camera is arranged on the side wall of the front end of the robot body, a rotating support is installed on the rear side of the camera, and an arc-shaped rack is fixedly arranged at the free end of the rotating support. A driving motor is installed in the robot body, a circular gear is fixedly arranged at the driving end of the driving motor, the circular gear is in engaged connection with the arc-shaped rack, and the rotating support is driven to drive the camera to rotate to collect images; a through hole is formed in the middle of the rotating support, a positioning column is fixedly arranged in the robot body, and the rotating support is rotationally connected with the positioning column through the through hole. A mechanical arm is fixedly arranged at the top of the robot body, and a push block is fixedly arranged at the free end of the mechanical arm; according to the coal mine inspection robot disclosed by the utility model, the camera is arranged on the side wall of the front end of the robot body, and the driving motor is arranged to drive the camera to rotate, so that the camera can carry out omnibearing image acquisition on the interior of a coal mine, and the coal mine inspection robot is safe and efficient.

The application discloses a wafer overlay error measurement method based on sub-pixel edge positioning, and belongs to the technical field of semiconductor optical detection and machine vision. The application solves the problem that the traditional symmetric Gaussian fitting model is prone to fitting distortion when processing actual wafer mark edge gradient asymmetry and image noise interference, and the sub-pixel edge positioning precision is limited, and finally the overlay error measurement accuracy is poor. The application includes the following steps: preprocessing an image to obtain a binary image; extracting a contour, performing affine correction and distinguishing an inner frame and an outer frame; intercepting a left edge and a right edge region of interest; extracting a one-dimensional gradient signal along a column direction, intercepting a fitting interval to obtain a maximum gradient amplitude sequence; achieving sub-pixel edge positioning by using weighted asymmetric Gaussian fitting, and calculating an X-direction overlay error; rotating the image by 90 degrees, and repeating the above steps to obtain a Y-direction overlay error; and synthesizing two-dimensional overlay error output, thereby improving the precision and stability of overlay error measurement. The application is mainly applied in a semiconductor lithography process.

To provide a head-up display device capable of more accurately adjusting the inclination of a projection image, and an image rotation angle setting method.SOLUTION: The head-up display device displays a virtual image by projecting onto a windshield, display light corresponding to a display image D emitted from a display surface 11a for displaying the display image D. The head-up display device comprises a correction display processing part which displays the display image D on the display surface 11a in a state where the display image D is rotated and inclined by display image rotation angle θd with the center position of the display image D as rotation center Cp so as to adjust the inclination of the virtual image in a left-right direction when viewed from a viewer. The display image rotation angle θd is set to be an angle larger than virtual image rotation angle which is inclination angle of the virtual image in a left-right direction generated when the display image D is displayed without inclination on the display surface 11a. The display image D and the virtual image have correspondence in which the virtual image rotates by the same angle as the virtual image rotation angle, by rotating the display image D by the display image rotation angle θd.SELECTED DRAWING: Figure 2

This invention provides a three-dimensional registration method for neutron / X-ray heterogeneous source images of cylindrical multilayer structures. The method includes edge extraction operators and PCA principal component analysis for vertical adjustment of image pairs; an edge contour extraction module to determine the height mapping of heterogeneous source tomographic images; a circle extraction algorithm and calculation of the circle center positioning translation; an image super-resolution reconstruction algorithm to achieve image scaling; and image rotation matching based on feature tomography and phase cross-correlation methods. This method effectively corrects geometric and modal differences between heterogeneous source images, achieving accurate alignment of metallic and non-metallic regions. It has significant advantages in registration accuracy, edge preservation, and modal feature alignment. It exhibits high adaptability and stability for complex structures and multimodal characteristic characterization, and is suitable for three-dimensional registration of neutron / X-ray heterogeneous source images of cylindrical multilayer structures such as cylindrical lithium batteries.

The application relates to the field of image processing, in particular to an image correction method and device and a storage medium. The method comprises the following steps: acquiring a first image photographed by a camera in an optical image stabilization mode, acquiring first displacement data corresponding to the first image, the first displacement data indicating a lens displacement condition when the camera photographs the first image; converting the first displacement data into a first image rotation angle according to calibration information, the calibration information comprising a corresponding relationship between an image rotation angle of a calibration image photographed by the camera and calibration displacement data, the first image rotation angle indicating an image rotation condition of the first image; performing image correction on the first image according to the first image rotation angle, and outputting a second image. According to the application, the first displacement data is converted into the first image rotation angle according to the calibration information, and the image correction is performed based on the first image rotation angle, so that the image rotation error caused by the OIS device in the photographing process is eliminated or reduced.

Disclosed are a method and apparatus for learning equivariant and invariant representations for rotation of an image based on a graph convolutional network. The method of learning an equivariant and invariant representation for rotation of an image based on a graph convolutional network performed by a computer device includes learning an equivariant representation for rotation of an image by using a self-weighted nearest neighbors graph convolutional network (SWN-GCN); and finally obtaining the equivariant representation for the rotation of the image obtained from the self-weighted nearest neighbors graph convolutional network as an invariant representation of the rotation of the image by using permutation invariance of global average pooling (GAP).