36 results about "Ultra high resolution" patented technology

The present invention discloses a method for ultra-high-resolution image denoising based on deep learning, which belongs to the technical field of image processing. The method comprises the following steps: step 1, constructing an ultra-high-resolution image cutting and coding model; step 2, inputting a noisy image and a clean image into the cutting and coding model to obtain image blocks; step 3, normalizing the image blocks to different degrees, constructing an image block sequence, and obtaining training and test data; step 4, constructing an image denoising network model; step 5, inputting the image blocks into the image denoising network model to obtain clean image blocks; step 6, decoding the clean image blocks output by the network and splicing them into the original image to obtain a denoised ultra-high-resolution image. The method proposed in the present invention has achieved good results on ultra-high-resolution images, and the network model can be trained end-to-end, which is highly practical in actual ultra-high-resolution image processing scenarios.

The present invention relates to a μLED display design method based on nanowire. The method comprises: first, growing different nanowire materials capable of generating three primary colors: red, green and blue on a substrate; then, respectively dissolving the different nanowire materials in insulated photocuring adhesives and injecting the mixtures into different grids of a grid pool, and performing curing; and finally, arranging electrodes on upper and lower surfaces of the grid pool. The method provided by the present invention can simplify the structure and enhance the industrial efficiency of the μLED.

The application discloses a kind of multi-camera imaging system based on microsecond level dynamic synchronization control, comprising: multi-camera imaging array includes several orderly arranged imaging camera, for imaging;Synchronization control module includes synchronization controller, for each imaging camera provides independent microsecond level time delay configuration;Timing management module includes time-sharing scheduler, for the exposure time slot of each imaging camera;Data acquisition and temporary storage module includes acquisition circuit and memory, for image acquisition and cache;Processing module includes image registration and alignment module and image fusion and interpolation module, wherein, the image registration and alignment module is used for image cooperation and alignment, the image fusion and interpolation module is used for image sequence interpolation image fusion.This application has the characteristics of high time synchronization accuracy, multi-channel acquisition, multi-resolution interpolation fusion space-time reconstruction, with ultra-high resolution and million frame per second level frame rate image acquisition and timing restoration.

This invention proposes a spaceborne ultra-high-resolution SAR imaging method with real-time tropospheric correction, which meets the requirements of high timeliness and high precision SAR imaging processing without being limited by the imaging area. Specifically, a corrected radiometer capable of covering the SAR observation field of view is used to simultaneously measure atmospheric temperature and water vapor pressure; the measurement data and SAR echo data are simultaneously transmitted to the ground to obtain the wet tropospheric zenith delay; using the total tropospheric delay estimation model, the total tropospheric zenith delay is obtained, and the slant range delay caused by the troposphere is further obtained; the slant range delay is used as a correction term to correct the SAR slant range model; in the ultra-high-resolution SAR imaging processing on the ground, the imaging algorithm adopts the slant range corrected by the real-time tropospheric correction, and finally obtains the ultra-high-resolution SAR image.

The application belongs to the technical field of image processing, and provides an image super-high resolution reconstruction method based on an attention mechanism, which comprises the following steps: shallow feature extraction is performed on a low-resolution image to be processed through a single convolution layer of a convolution network to obtain a shallow feature map; deep extraction is performed on the shallow feature map through RIR to obtain a deep feature map rich in high-frequency details; a high-resolution feature map is obtained by performing an up-sampling operation on the deep feature map through sub-pixel convolution; and the high-resolution feature map is reconstructed by using a single convolution layer of the convolution network, and a final super-resolution image is output. The application can reliably and accurately restore high-frequency details, avoid blurring / artifacts / false details, and improve the practicability and credibility in key applications.

1. The name of the design product: laser-induced breakdown spectroscopy equipment. 2. The use of the design product: analyzing and identifying the elemental composition of the measured substance by the method of laser-induced atomic spectroscopy (LIBS). By exciting the sample plasma with high-energy laser and combining with ultra-high resolution spectrometer and automatic focusing function, fast and accurate material composition analysis is realized. It is widely used in the fields of geology, metallurgy, environmental monitoring, material science, etc. 3. The design points of the design product: in shape. 4. The picture or photo that best indicates the design points: perspective view.

Disclosed herein are various example systems, methods, and compositions, for extremely high-resolution in situ spatial analysis of a sample, with particular applicability to biological materials, by strategically multiplexing a. plurality of different imaging techniques. For instance, certain inputs can be configured to elicit multiple different types of output signals from the sample simultaneously, each of which encodes a different attribute of the sample. These various outputs, once detected, decoded, and assembled, provide for a more- robust analysis than their individual techniques, i.e., if performed sequentially. As just one illustrative example, the infrared light involved in photothermal-infrared (PI) microscopy can also be used, to simultaneously trigger fluorescence-encoded infrared (FEIR) excitation within the sample. Pegging the excitation signals to specific biomarker targets can drastically enhance the combined 3D imagery' that results. Additional techniques may be integrated (i.e., higher-order multiplexing) to resolve the imagery even further.

A tileable, low-cost image sensor array suitable for creating very large images (e.g., >1 gigapixel). The optical system can also be used to combine low resolution displays to create very high-resolution images (e.g., >1 gigapixel). It includes at least one beam splitter, a pair of sensor arrays and a mirroring means (physical mirror or electronic / digital mirroring algorithm). With 10 megapixel image sensors, four arrays 6×6 image sensors would provide a gigapixel image. For 48 megapixel image sensors, four array of 3×3 image sensors would provide a gigapixel image. The design of the image sensor arrays, when factoring in mirroring of the image array sensors, allows for the images to be stitched together to form a composite image.

The invention generally relates to systems and methods for utilizing interferometric ultra-high resolution 3D imaging. In certain aspects, the invention provides a system for achieving molecular resolution through tissues, the system comprising: an interferometric apparatus; and a control unit comprising one or more processors configured to: receive interferometric tissue data from the interferometric apparatus, wherein the interferonic tissue data comprises one or more tissue-induced abberations; apply a 4Pi in-situ point spread function (PSF) retrieval algorithm to the interferonic tissue data, thereby enabling direct retrieval of an aberrated, time-dependent in-situ model of single-molecule emission pattern from the interferonic tissue data, wherein the in-situ model reflects distortions, blurring, and / or partial coherence encountered during detection; and minimize disparities between the interferonic tissue data and the in-situ model to thereby provide accurate molecular localizations in the presence of tissue-induced aberrations.

Single-mask X-ray imaging simultaneously captures attenuation, differential phase contrast (DPC), and dark-field images in a single exposure by capturing multiple contrast features like dark field and differential phase contrast using mask alignment with relatively low-resolution detectors and smaller magnification. New light transport models are used for various system configurations to effectively extract contrast features from various single mask X-ray configurations, thereby eliminating the need for highly coherent X-ray sources, ultra-high-resolution detectors, and intricate gratings, and as such simplifying the imaging process and significantly reducing costs and improving the prospects of clinical translation. Three variations of the single-mask setup are each optimized for different contrast types, offering flexibility and efficiency in a variety of applications.

The invention relates to the technical field of optics, in particular to a photon nanometer jet flow generator based on a concentric ring structure and a design method, and the photon nanometer jet flow generator comprises a medium substrate and a medium concentric ring structure arranged on the medium substrate; the medium concentric ring structure is formed by circular ring structures with different inner and outer radiuses distributed along a concentric circle. According to the PNJ generator based on the concentric ring structure, the concentric ring structure meeting the requirement can be designed as required for any optimization target. Moreover, the concentric ring structure provided by the invention is compatible with a micro-nano processing technology, is easy to process, has the advantage of on-chip integration, and shows huge application potential in the fields of ultrahigh-resolution optical imaging, nano particle control, ultra-sensitive biosensing, nanoscale photoetching, enhanced spectroscopy and the like.

This disclosure provides a method and related apparatus for determining the location of ultrashort baseline lightning strikes using a multi-antenna monostatic three-dimensional layout. The method utilizes a multi-antenna three-dimensional layout to construct one set of regular pentagonal interferometer arrays without angular error and multiple sets of equilateral triangular interferometer arrays with angular error by different combinations of antennas. The time difference between each VHF antenna of the same pulse signal is obtained by different antenna combinations. Two-dimensional observations without system error are performed using the regular pentagonal interferometer arrays. The two-dimensional observation results with system error obtained from the equilateral triangular arrays are corrected. Then, the two-dimensional information obtained from the regular pentagonal interferometer arrays is used as a reference, and the corrected two-dimensional information obtained from the multiple equilateral triangular interferometer arrays is intersected in three-dimensional space to obtain three-dimensional information. This disclosure can achieve ultra-high-resolution three-dimensional location determination for lightning discharges, and can achieve nanosecond time resolution and digital-meter order spatial resolution of the lightning discharge process.

This invention discloses an ultra-high resolution FMCW-based spaceborne sliding spotting mode SAR imaging method, which receives the reflected signals of sub-blocks in the region to be imaged; performs Fourier transform on the fast time portion of the reflected signal and then phase compensation on the slow time portion, converting the resulting signal into a virtual equivalent direct sampling matched filter echo signal; performs range pre-filtering; and performs frequency domain f i The domain flip translation is the translation amount under a unified reference slant range; fast time domain sign and scale replacement is performed; the range time domain is converted to the frequency domain and downsampled; the downsampled signal is imaged to obtain sub-block images; several sub-block images are fused to obtain the SAR image of the area to be imaged; by using a spaceborne ultra-high resolution SAR fast imaging method based on range-to-frequency domain downsampling, the computational load of block imaging processing can be greatly reduced.

The present invention relates to a microlens array sheet of ultra-high resolution using quantum dots and a method for manufacturing same, for implementing an ultra-high resolution RGB display, and a color-conversion microlens array sheet according to one embodiment of the present invention is formed by first creating a barrier structure to determine the diameter of the lenses and the distance between adjacent lenses, and then filling empty spaces between the barriers with a microlens-shaped structure containing quantum dots.

A method for fabricating a semiconductor device is provided. The method for fabricating the semiconductor device includes forming a semiconductor pattern including a first layer and a second layer on a substrate, forming a coating layer on a surface of the first layer, forming a dyeing substance in which one of an antibody or a protein is combined with a fluorophore, attaching the dyeing substance to a surface of the coating layer to form a dyeing layer, and photographing the fluorophore with an ultra-high resolution microscope to detect the semiconductor pattern.

The invention discloses a multi-image panoramic stitching method based on depth matching, a camera and a storage medium, and the method comprises the following steps: S1, image preprocessing and multi-scale zooming: firstly inputting a multi-image sequence, and after inputting the multi-image sequence, according to a self-defined working resolution, a seam resolution and a synthesis resolution, carrying out the multi-scale zooming of the multi-image sequence; the method comprises the following steps of: performing adaptive scaling on a plurality of input images to generate a multi-scale image set; s2, deep learning end-to-end matching: performing end-to-end feature matching on the multi-scale image set by using a pre-trained ONNX model to generate direct matching information between the image pairs; s3, performing global consistency screening; s4, estimating and optimizing camera parameters; and S5, performing exposure compensation and image fusion. The method has the advantages that the matching precision and the calculation efficiency are remarkably improved, the overall consistency of the splicing result is ensured, the calculation efficiency and the splicing quality are balanced, and the method is suitable for ultrahigh-resolution image sequences.

The invention relates to an end-side adaptive picture super-division processing method and system based on weak network perception, and belongs to the technical field of image processing and mobile computing, and the method comprises the steps: continuously monitoring a network state at a client, and setting a weak network mark; the method comprises the following steps of: integrating a lightweight picture super-division SDK (Software Development Kit), wherein a special super-division model which is obtained through a knowledge distillation and model compression technology and has a parameter quantity between 0.1 B and 0.5 B is arranged in the lightweight picture super-division SDK; according to the target resolution and the network state of the to-be-requested picture, adaptively selecting a picture acquisition strategy: directly downloading a low-resolution picture; downloading a low-resolution version of the medium-resolution picture in a weak network, and processing the low-resolution version into a high-resolution picture in real time by an end-side super-resolution SDK; and the ultra-high-resolution picture goes through a local ultra-resolution process by default. According to the method, the contradiction between high-precision picture display and network transmission delay is effectively solved, the user experience in a weak network environment is remarkably improved, and efficient balance between network bandwidth and terminal computing resources is realized.

This invention relates to a method for fabricating horizontal heterojunction photodetector matrices based on single-crystal nanoribbons, belonging to the field of nanomaterials and optoelectronic devices. A two-step chemical vapor deposition process is employed: first, a first type of single-crystal nanoribbon array is directionally grown on a single-crystal substrate with a stepped surface; then, using this as a template, a second material is epitaxially grown along the lateral edges of the nanoribbons, directly forming a large-area, oriented, continuous single-crystal horizontal heterojunction nanoribbon array. This avoids material damage at the heterojunction interface caused by traditional patterning processes, achieving high-quality, high-uniformity, and controllable interface fabrication. The nanoribbon array is then segmented into a high-density discrete PN junction matrix using micro / nano fabrication techniques, and electrodes are fabricated to construct the photodetector matrix. This achieves controllable growth of a high-quality single-crystal heterojunction array and efficient integration of the optoelectronic device matrix. The resulting device exhibits advantages such as high pixel density, no crosstalk between units, low dark current, and fast response, providing a feasible solution for ultra-high resolution optoelectronic integrated systems.

This application relates to the field of medical image processing technology, and discloses a fully automated lung nodule segmentation system and method based on ultra-high resolution target scanning CT. The system includes a data preprocessing module, a 3D target detection module, an adaptive resampling module, a 3D fine segmentation module, and a post-processing and feature extraction module. The detection module outputs a 3D bounding box to provide nodule in-plane size information for adaptive resampling; the resampling module maintains the Z-axis unchanged while adaptively downsampling the XY axes; the segmentation module adopts a dual-cascaded feature pyramid network architecture to achieve two-stage optimization of coarse and fine segmentation; the post-processing module generates a binarized segmentation mask and calculates quantitative features. This application effectively mitigates the impact of some volume effects on the segmentation accuracy of subsolid nodule boundaries.

The present invention proposes a method for acquiring a virtual rotation point in a sliding spotlight mode of a spaceborne SAR, which is suitable for acquiring a virtual rotation point in a sliding spotlight mode of a spaceborne SAR in ultra-high resolution conditions. The present invention segments the entire imaging time, acquires the closest point of the beam pointing vector for each time period, and then averages the positions of the closest points of the beam pointing vectors for all acquired time periods to obtain the position coordinates of the virtual rotation point. Compared to the prior art, the present invention fully considers the impact of changes in scanning rate and greatly reduces this impact, ensuring ultra-high resolution and wide-angle observation. It can meet the requirements of spaceborne SAR imaging with an ultra-high resolution better than 0.1m and a spaceborne SAR azimuth scanning angle of ±50°.

The invention discloses a billion-pixel-level high-resolution long-focus power monitoring lens and electronic equipment. The lens sequentially comprises a first lens, a second lens, a third lens, a fourth lens, a fifth lens, a sixth lens, a seventh lens, an eighth lens, a ninth lens, a tenth lens, an eleventh lens and a twelfth lens from the object side to the image side along the optical axis. The first lens has positive diopter, the object side surface is a convex surface, and the image side surface is a concave surface; the second lens has negative diopter, the object side surface is a convex surface, and the image side surface is a concave surface; the third lens has negative diopter, the object side surface is a convex surface, and the image side surface is a concave surface; the fourth lens has negative diopter, the object side surface is a concave surface, and the image side surface is a convex surface; the fifth lens has positive diopter, the object side surface is a plane, and the image side surface is a convex surface. Through the configuration of diopters and surface types of the lenses and the introduction of the gluing group, the invention aims to systematically solve the core problem of coexistence of ultrahigh resolution and harsh environmental adaptability.

This invention discloses a computational spectral measurement system based on spatial and angular coupling modulation. The system introduces the electromagnetic radiation to be measured through a point light source, which is then subjected to secondary modulation by a spectroscopic filtering unit and a phase interferometry unit with different positional combinations. The spectroscopic filtering unit restricts the wavelength passband in the spatial dimension, while the phase interferometry unit generates interference fringes that vary highly with wavelength by utilizing the coupling relationship between the incident angle and phase delay. A composite coded image containing the spatial wavelength characteristics and the angular phase characteristics is acquired by a detection unit. Then, by establishing a forward physical model and employing a computational reconstruction algorithm, ultra-high resolution spectral reconstruction is achieved across the entire wavelength range. This system effectively solves the problem of aliasing in high-order interference through dual modulation at the physical level, and has advantages such as simple structure, extremely high resolution, and strong band versatility.

A flat optical device includes a substrate, a first nanostructure layer on the substrate, and a second nanostructure layer on the first nanostructure layer. The first nanostructure layer includes a first array of meta-elements in a first array of segments and characterized by a same first height in a first direction. The second nanostructure layer includes a second array of meta-elements in a second array of segments and characterized by a same second height in the first direction. The second array of segments is shifted with respect to the first array of segments in one or more directions perpendicular to the first direction. In one example, the flat optice device includes an array of flat micro-lenses and is configured to collimate and / or deflect light beams emitted by an array of micro-light emitting diodes.

The present application relates to a kind of based on heterodyne technique's hyperspectral detection method and the online hyperspectral detector using this method, belong to photoelectric detection technical field, solve the technical problem that the detection spectral range is limited.The method includes the following steps: after the optical collimation system collimation of the light signal to be measured, enter optical fiber and couple;Local light laser of the frequency band to be measured emits local light signal, and is divided into probe signal and reference signal by optical beam splitter.For the frequency band below wavelength 760nm, when the central wavelength of laser is not tunable, the present application utilizes the central frequency of band-pass filter to change, realizes the ultra-high resolution spectrum scanning of laser frequency band below 760nm.The present application changes from limited to possible for the heterodyne detection method of laser wavelength non-tunable range, and the implementation method is simple, the pressure of laser drive is changed into the implementation pressure of electronic filter, and digital control method is more flexible.

This invention discloses a method and system for enhancing ultra-high resolution low-light images based on block overlap inference, belonging to the field of image processing technology. The method segments the ultra-high resolution image into sub-blocks with overlapping regions, and uses a deep learning model based on the UNet architecture to independently enhance each sub-block. The model incorporates a Feature Linear Modulation (FiLM) module to adaptively modulate the underlying features using target brightness parameters. It calculates the ratio of the local average brightness of the sub-block to the global average brightness, and uses logarithmic scaling to determine the local target brightness, ensuring brightness continuity. The overlapping regions are progressively fused using a cosine window function to eliminate block artifacts. This invention solves the problems of memory limitations, block artifacts, and brightness discontinuities in low-light enhancement of ultra-high resolution images, achieving high-quality real-time enhancement, and is suitable for scenarios such as wide-area monitoring and astronomical observation.

The invention provides a renal tubule and interstitial lesion grading system, and relates to the technical field of renal tubule and interstitial lesion grading, and the system comprises a preprocessing module, a slice processing module, a lesion segmentation network module, a mask splicing module and a grading network module. The preprocessing module is used for segmenting renal cortex and renal medullary regions in the kidney pathological full-dose image and removing the renal medullary regions. The slice processing module is used for segmenting the preprocessed renal cortex area image into slice images with a first resolution. The lesion segmentation network module is used for outputting an IFTA lesion segmentation mask. And the mask splicing module is used for splicing the lesion segmentation masks into a mask image with a second resolution. And the grading network module is used for generating an IFTA lesion grading result. The system provided by the invention can realize synchronous breakthrough in calculation precision, calculation efficiency, clinical suitability and generalization ability, and can effectively solve the problems of processing bottleneck, lesion characteristic dispersity and manual evaluation subjectivity of ultrahigh-resolution WSI.

Apparatuses and methods for uplink (UL) beam tracking. A method performed by a user equipment includes transmitting a physical uplink shared channel (PUSCH). The PUSCH includes a first set of resource elements (REs) including a first demodulation reference signal (DMRS) and a first UL-shared channel (SCH), the first DMRS and UL-SCH based on a first spatial transmission filter, and a second set of REs including first N reference signals (RSs) based on second N spatial transmission filters, respectively, where N is larger than or equal to 1. The method further includes receiving a first channel including information based on measurement of the first DMRS and the first N RSs, determining, based on the first channel, a third spatial transmission filter for the first DMRS and the first UL-SCH, and transmitting the first set of REs based on the third spatial transmission filter.

The application discloses a kind of hardware computing system and method based on Canny algorithm's image upsampling, it is related to image processing field.When current 4 times or above times upsampling operation is carried out to image, hardware resource consumption is big, calculation is slow, cannot be simply realized by integrated circuit design.This application presents upsampling system and method, and the upper and lower boundary of image is generally interpolated, and the core interpolation calculation is used to the core part of image.Core interpolation algorithm repeatedly calls image core interpolation module, can be interpolated by row, system only needs to store a part of picture each time, can improve the improvement of image edge and detail in upsampling process with low hardware resource consumption as cost, improve operation speed by multi-path parallel computation, realize real-time processing to image, can be completed by integrated circuit from low resolution to super high resolution RGB image upsampling.

The application discloses a kind of based on pixel reorganization sampling block's super high resolution image denoising method.The specific steps are as follows: the pixel reorganization undersampling block of paired noisy and noiseless super high resolution image is carried out to generate data set;Depth learning network constituted by two denoising networks and a fusion network is constructed using image block and its pixel adjacent image block;Based on three loss functions respectively constraining two denoising networks and a fusion network, the whole network is trained jointly;Inference, the pixel reorganization undersampling block of noisy super high resolution image is carried out, image block and its pixel adjacent image block are sent into network denoising, pixel reorganization upsampling splicing, to obtain the full image denoising result of super high resolution image.This method makes full use of the local information and non-local information of super high resolution image, makes up the shortcoming that non-local information is difficult to use when block denoising, and traditional undersampling will destroy image details texture.