211 results about "Color map" patented technology

Embodiments of the present disclosure provide a three-dimensional scene reconstruction method and apparatus, a device, a medium, and a program product. The three-dimensional scene reconstruction method comprises: acquiring a scene image collected for a three-dimensional scene; on the basis of the scene image, determining sparse point cloud data and camera parameter information corresponding to the scene image; performing model initialization on the basis of the sparse point cloud data to obtain a current three-dimensional data model; performing rendering on the basis of the camera parameter information and the current three-dimensional data model to obtain a current color map, a current depth map and a current normal map from the perspective of the scene image, and determining a pseudo normal map from the perspective of the scene image on the basis of the current depth map; and on the basis of the current color map, the current normal map, the pseudo normal map, and an actual color map, training the current three-dimensional data model to obtain a trained target three-dimensional data model. By means of the technical solution provided by the embodiments of the present disclosure, higher-quality automatic reconstruction of three-dimensional scenes can be achieved, reducing reconstruction costs, and improving the level of detail and rendering quality of scene models.

The invention relates to the field of image processing, in particular to a low-illumination image enhancement method and system based on an HVI color space. Carrying out HVI color space conversion on the original RGB image to obtain an intensity graph and a horizontal and vertical color graph; performing feature extraction on the intensity image and the horizontal and vertical color image to obtain an intensity image feature and a horizontal and vertical color image feature with the same dimension; inputting the intensity graph feature and the horizontal and vertical color graph feature into a pre-constructed double-branch U-Net network to obtain an HV branch feature graph, and cascading the I branch feature graph and the HV branch feature graph to obtain an enhanced HVI feature graph; and carrying out inverse transformation processing on the enhanced HVI feature map to obtain an enhanced image. According to the dual-branch U-Net network provided by the invention, through parallel processing of the I branch and the HV branch, decoupling enhancement of brightness and color features is realized, scene information can be supplemented at each branch, and the visual effect of an enhanced image is further improved.

The invention relates to a light effect animation data generation method and apparatus, a device and a medium. The method comprises the steps of obtaining a target image for constructing light effect animation data; according to the color values of the pixels in the target image, constructing a palette containing a fixed number of reference colors, and determining attribution probability distribution corresponding to each reference color mapped to the palette by each pixel; corresponding to the parity of each pixel in the plane space position of the target image, selecting a reference color according to the attribution probability distribution so as to update the color value of the pixel; and packaging the target image with the updated color value into the light effect animation data. According to the method, the light effect animation data is generated by utilizing the attribution probability and odd-even spatial modulation, and the smooth pixelated light effect can be realized when the atmosphere lamp is played.

The invention relates to a depth map super-resolution method based on double-edge consistency constraint, belongs to the field of image quality enhancement, and solves the problem that the generalization ability is reduced due to the fact that the double-edge consistency constraint is not considered and feature fusion of a color map and a depth map is insufficient in an existing depth map super-resolution method. The method comprises the following steps: firstly, converting a depth map super-resolution problem into a three-task optimization model of double-edge consistency constraint, and secondly, according to an alternating direction multiplier method theory, converting the model into an iterative optimization process of color map edge updating, depth map edge updating, depth map updating and augmented Lagrange multiplier updating. Converting the optimization process into an expansion network, embedding a learnable module into an expansion model of an original pure expression to realize high-throughput transmission so as to enhance the expression capability, and further providing a multi-cross attention compensation fusion module; the module realizes approximation of a near-end operator by aggregating a color map edge, a depth map edge, cross-stage features and auxiliary information.

The invention relates to the technical field of laser measuring instruments, and provides a plug-in type handheld laser scanner and method based on light spot fine positioning, and the method is characterized in that a light spot scanning plug-in is accessed through a standardized interface of an equipment main body, and the light spot scanning plug-in supports a transverse arrangement mode and a longitudinal arrangement mode; acquiring centimeter-level coarse positioning information by using the equipment main body; according to the arrangement mode of the light spot scanning plug-in, controlling the light spot scanning plug-in to carry out light spot scanning according to a corresponding scanning strategy; in the light spot scanning process, depth data and a color image of the target object are synchronously collected; performing fusion processing on the centimeter-level coarse positioning information, the submillimeter-level fine measurement data, the depth data and the color image, and constructing point cloud data of the target object; and the detail precision and the texture effect of the point cloud data are optimized based on the depth data and the color image, so that the problem of target loss caused by contradiction between efficiency and precision of a traditional laser measuring instrument can be solved.

The invention relates to the technical field of computers, and discloses a digital twinning-based power grid monitoring data visual interaction method and system, and the method comprises the steps: carrying out the real-time monitoring of the operation state of a target power grid, and collecting the operation data of the target power grid; mapping the operation data to a digital twin power grid model of the target power grid to carry out anomaly diagnosis, and obtaining a power grid diagnosis result of the target power grid; based on the power grid diagnosis result, through a fault tree model of a target power grid and a fault tree logic relationship, determining occurrence states of each intermediate event and each bottom event, and obtaining an abnormality diagnosis result; and displaying abnormal equipment and a fault area in a display interface of the digital twin power grid model with different colors, icons and flickering effects based on the abnormal diagnosis result, and viewing detailed information of the abnormal equipment and the fault area through click operation. According to the invention, the abnormal source can be rapidly and accurately positioned, and the power grid operation stability is improved.

A method, apparatus, non-transitory computer readable medium, and system for data generation include obtaining a text prompt describing an object within a scene and generating, using a text-to-mask generation model and based on the text prompt, a color map corresponding to the scene. The color map indicates a region corresponding to the object from the text prompt. An image segmentation mask is generated based on the color map. The image segmentation mask comprises a plurality of regions corresponding to a plurality of image elements in the scene including the region corresponding to the object from the text prompt.

The invention discloses a three-dimensional scene reconstruction method for obtaining real three-dimensional data based on simple shooting, and the method comprises the steps: 1, carrying out the surrounding video shooting of a target object through a camera of a cellphone, and obtaining the video data containing the target object; 2, decomposing the video data into a plurality of image sequences according to frames, and optimizing the image sequences to generate an original image sequence; 3, performing streaming segmentation on the original image set frame by frame to obtain a mask sequence of the target object, and obtaining a foreground image sequence according to the mask sequence; step 4, performing camera pose estimation on the original image set; step 5, inputting the foreground image sequence and corresponding camera internal and external parameters into a 2DGS model to generate an initial three-dimensional scene; and step 6, performing fine processing on the generated three-dimensional scene, generating a color map and a depth map in the 2D three-dimensional scene for mapping a voxel network, and extracting a contour surface on the voxel network to obtain a final three-dimensional scene.

The invention discloses a color threshold adjustment method and system based on light brightness adaptive change. The method comprises the following steps: acquiring an image of any color; carrying out color correction on the acquired image so that the acquired image is restored to the real color of the object under different illumination; the corrected image is converted into a corresponding HSV color space, V channel information in the HSV image is extracted, a corresponding brightness image is obtained, and a brightness value is averaged; dynamically adjusting the threshold range of the color according to the average brightness value and dynamically adjusting the upper and lower limit values of the HSV color space of the color; segmenting the HSV image to obtain a regional mask of any color; carrying out filtering processing on a color region mask image in the processed binarized image to obtain a filtered color mask; and converting the originally collected image into a grayscale image, and performing masking operation on the grayscale image by using the filtered color mask to obtain a grayscale image only containing a color region. The method can effectively meet the actual requirements of industrial production.

The invention provides a multi-color printing control method, device and system of a thermal transfer printer, which are applied to a system comprising an editing and typesetting terminal and a thermal transfer printing mechanism, and the thermal transfer printing mechanism is provided with at least two printing heads which are respectively provided with colored tapes with different colors. The method comprises the following steps: acquiring a design page containing images and texts with various colors at an editing and typesetting terminal; identifying the color of each currently available color tape of the thermal transfer printing mechanism to form an available color set, and establishing a corresponding relationship between each printing head and the color of the color tape of the printing head; analyzing the design page pixel by pixel, matching the color of each pixel with the available color set to determine the target available color of the pixel, and merging the pixels with the same or similar target available colors into a corresponding printing object; distributing the printing objects to printing heads corresponding to the target available colors of the printing objects, and generating printing data of the printing heads based on distribution; and the printing data are sent to a thermal transfer printing mechanism, so that a multi-color printing task is completed.

The embodiment of the invention provides a training method and device of a three-dimensional point cloud fusion model, equipment and a medium, and belongs to the technical field of computer vision. The method comprises the following steps: rendering an initial three-dimensional point cloud from color attributes and geometric attributes respectively based on camera parameters and an inherent parameter group at any camera visual angle to obtain a self-supervised color image and a self-supervised normal image; performing rendering processing on the sample three-dimensional point cloud based on the material parameters and the normal parameters to obtain a predicted color image, and performing rendering processing on the sample three-dimensional point cloud based on the normal parameters to obtain a predicted normal image; obtaining a luminosity loss value based on the self-supervised color image and the predicted color image, and obtaining a geometric loss value based on the self-supervised normal image and the predicted normal image; and training the initial fusion model according to the luminosity loss value and the geometric loss value so as to adjust the material parameter and the normal parameter, thereby obtaining a target fusion model. The reality sense of fusion of the multiple three-dimensional point clouds can be improved.

The invention relates to the technical field of machine vision detection, and discloses a color printed matter quality detection method which comprises the following steps: acquiring original light field response data of the surface of a printed matter in a preset structure light field; carrying out decoupling calculation on the data to generate a basic color graph representing intrinsic colors, a glossiness graph representing process characteristics and a surface normal graph representing geometrical morphology; the method is characterized in that a texture area is identified based on a surface normal graph of a standard sample, and a dynamic reference model is established for the area in combination with a basic color graph so as to provide a self-adaptive color detection threshold value; and finally, detecting color, structure and process flaws in parallel by utilizing each physical characteristic spectrum and the dynamic reference model. According to the method, through decoupling and information guidance of physical dimensions, interference of physical textures on color detection can be fundamentally inhibited, the false alarm rate is remarkably reduced, meanwhile, multi-dimensional comprehensive evaluation of flaws is achieved, and the detection result is objective and reliable.

The invention relates to the field of computer vision and navigation, and provides an SLAM acceleration method based on three-dimensional Gaussian rendering, and the method comprises the steps: initializing a three-dimensional Gaussian point cloud and a camera pose parameter based on input data, executing the camera tracking of a target frame in the input data, optimized camera pose parameters are output based on a staged switching strategy and an iteration mechanism, and camera tracking utilizes a floating point format with reduced precision; and executing a hierarchical screening mechanism in the optimized camera pose parameter to determine a key frame from the color image and the depth image, and performing mapping through a renderer supporting multichannel data by using the key frame and the optimized camera pose parameter to update the three-dimensional Gaussian point cloud. According to the method, through staged switching and iteration, pose optimization can be rapidly converged, a precision reduction floating point format is used, the numerical processing burden is relieved, a layered screening mechanism can be used for reducing the calculation cost lower than that of frame-by-frame fine comparison, and then the problem that the efficiency of an SLAM process is low is solved.

The invention provides a special effect rendering method and apparatus, an electronic device and a storage medium. The method comprises the steps of obtaining a primary color map and a flow map; generating a first mask based on the primary color map and the flow map; generating a second mask based on the first mask; and respectively applying the first mask and the second mask to different special effect layers to generate a composite special effect rendering result. Through the method provided by the invention, the associated multi-layer mask is calculated and generated based on the single primary color chartlet and the flow chartlet, the chartlet sampling frequency is reduced, and the rendering performance is remarkably improved; and meanwhile, the second mask is derived from the first mask, so that natural visual relevance is kept between the two layers of special effects, and the overall aesthetic feeling of the special effects is enhanced.

In the present invention, a map is prepared such that, on each of the surfaces of a cube 112 including an object 110, a distribution including a surface color value is projected when the object 110 is viewed from viewpoints corresponding to the surfaces. The cube 112 is arranged in a space to be displayed. When a ray generated for an arbitrary display viewpoint reaches the inside of the cube, the arrival to the object is determined by using a depth map 116, and the pixel value is determined by sampling the color value corresponding to the arrival point by using a color map 114.

Systems and methods for visualizing time-varying two-dimensional (2D) color maps are disclosed. The 2D color map may include luminance values ​​corresponding to hues of varying intensities that change over time. In some 2D color maps, the hue used for the intensity scale may change over time, while the luminance of the pixels may remain constant. In some 2D color maps, both hue and luminance may change over time relative to intensity.

A plurality of training examples is accessed, each training example comprising an image of a scene and a pose of a viewpoint from which the image was captured. A neural radiance field is trained using the training examples. A plurality of generated images is computed, by, for each of a plurality of randomly selected viewpoints, generating a color image and a depth image of the scene from the neural radiance field. A neural network is trained using the generated images.

Method of obtaining a color subpalette from a first moodboard containing one or more digital color images, said method comprising the steps of: (1) generating a first color palette with k1 representative colors from the first moodboard by a method that takes account of underrepresented colors, (2) providing a second color palette with k2 colors, (3) ranking the colors in the first color palette according to a scoring algorithm, which assigns a score to each of the colors in the first color palette based on its color difference with each of the colors in the second color palette, and (4) selecting a subset of ksub colors from the first color palette, which have the scores corresponding to the least color difference with the colors in the second color palette, said subset of colors being the color subpalette. The disclosure also provides a data processing device and a computer program product.

The invention relates to the technical field of synthetic aperture radar image processing, and discloses a PIX2PIX model-based SAR image grey-scale map coloring method and system, and the method comprises the steps: building a PIX2PIX model, carrying out the filtering and downsampling of the SAR image grey-scale map, and converting the SAR image grey-scale map into an SAR image color map through the PIX2PIX model; according to the RGB ratio of the color image of the SAR image and the gray value of the gray-scale image of the SAR image, constructing a linear unary equation about the red channel value of the color image of the SAR image; according to the red channel value of the SAR image color map and the RGB ratio of the SAR image color map, the blue channel value and the green channel value of the SAR image color map are calculated, and the technical problems that an existing SAR image gray-scale map coloring method is low in automation degree, poor in color authenticity, weak in channel coordination and insufficient in detail reservation are solved.

A memory structure stores a plurality of color maps, the entity, when decompressed by a printer controller, comprising a plurality of first level wrappers, each first level wrapper including at least one characteristic to identify a corresponding color map, and a recipe to define and locate packets for constructing the color map, the recipe having at least one locator for each second level wrapper; and a plurality of second level wrappers of different types, each second level wrapper wrapping a corresponding packet.

This invention discloses a self-supervised depth completion method, system, and terminal based on 3D feature extraction and fusion. The method includes: acquiring a discrete depth map and a color map; preprocessing the discrete depth map to obtain a target discrete depth map; downsampling the color map to obtain a target color feature map; performing a preset number of downsampling and feature extraction operations on the target discrete depth map and the color map to obtain a first color feature map and a first target discrete depth feature map; performing channel concatenation and upsampling operations to obtain a fused image feature map; inputting the fused image feature map and the target discrete depth map into a cross-attention feature fusion module to output the fused feature map; and performing channel concatenation and upsampling on the target color feature map and the fused feature map to obtain a completed depth map. This invention can obtain a completed depth map after information completion, thereby enabling accurate processing of subsequent computer vision tasks.

A method for driving a display device is performed by the display device. The method includes: displaying a first image including user convenience information through a display panel; generating sensing data corresponding to a fingerprint sensing area by using a photoelectric sensor; and determining whether a sensed fingerprint corresponding to the sensing data is a fake fingerprint by comparing expected sensing light illuminance information based on the first image with sensed light illuminance information of the sensing data. The first image includes a first color pattern in the fingerprint sensing area of the display panel. A fake-determination image pattern includes the first color pattern and a second color pattern different from the first color pattern.

The embodiment of the invention provides an image editing method and device, a computer readable medium and electronic equipment. The method comprises the steps that an original image, a single-color image, first description information of a first object in the original image and second description information of a second object in a target image are acquired; encoding the original image into a first feature, encoding the single-color image into a second feature, and constructing a text feature according to the first description information and the second description information; splicing the first feature and the second feature, and adding the splicing result and the noise graph to obtain a first hidden space feature; performing forward diffusion operation based on the first hidden space feature to obtain a second hidden space feature; performing denoising operation of multiple time steps on the second hidden space feature based on the text feature to obtain a third hidden space feature; and generating a target image matched with the second description information according to the third hidden space feature. According to the scheme provided by the embodiment of the invention, the image editing efficiency and the image editability can be greatly improved.

Provided is a compensation device for a display apparatus, including an image-acquiring assembly and a control assembly, wherein the image-acquiring assembly is configured to acquire a first image in a case that the display apparatus displays a picture of a first color; the control assembly is configured to determine a reference region in the first image based on a position of a region to be compensated in the first image in a first direction, wherein the reference region has a same position as the region to be compensated in the first direction, the reference region being greater than the region to be compensated; and the control assembly is configured to determine compensation data of the region to be compensated based on average brightness of the reference region and average brightness of the region to be compensated.

The invention relates to the technical field of 3D rendering, and discloses a rendering method based on video texture rendering, and the method comprises the steps: carrying out the rendering of a to-be-rendered object, and carrying out the time domain anti-aliasing processing, and obtaining a color image and a depth image; and performing video texture rendering on the to-be-rendered object based on the color image and the depth image to obtain a video texture template edge image, and performing pixel jitter elimination on pixels at the junction of the to-be-rendered object based on the video texture template edge image. According to the invention, the problem that the junction of the image shakes in the prior art is solved.

An image sensor includes a pixel array including a plurality of unit pixels configured to receive light of a color filter array having a first color pattern of size k×l arranged therein, where k and l are natural numbers of 3 or more, the image sensor configured to generate first image data including the first color pattern and an image signal processor (ISP) configured to generate second image data including a second color pattern of size p×q, by remosaicing the first image data, where p and q are natural numbers of 2 or more, where a size of the second color pattern is smaller than a size of the first color pattern, and the image sensor is configured to transmit the second image data an application processor that is external to the image sensor.

The application discloses a thunder and lightning comprehensive detection early warning system and a thunder and lightning detection method thereof, which comprises an atmospheric electric field detection module, a meteorological radar module and a system module. The atmospheric electric field detection module is used for detecting characteristic value A of a thundercloud in a specific area, which at least contains electric field data. The meteorological radar module is used for detecting characteristic value B of the thundercloud in the specific area, which at least contains geographic position data. The system module is used for connecting the atmospheric electric field detection module and the meteorological radar module, and combining the received characteristic value A with the characteristic value B to form thunder and lightning early warning information. The application combines the meteorological radar and the atmospheric electric field equipment, embeds the meteorological radar device into the atmospheric electric field instrument, detects echo data by the radar, converts the echo data into an echo intensity color map, and superimposes the color map on a map to form a thunder and lightning detection early warning radar system.

Methods, apparatus, systems, and articles of manufacture are disclosed to automatically process an image based on a detection of a face mask. An example article of manufacture include instructions that, when executed, cause programmable circuitry to at least: map a characteristic of an upper area of a face of a person in an image to a color plot including a face skin tone cone; map a characteristic of a lower area of the face to the color plot; and identify a presence or an absence of a face mask based on the respective positions of the characteristic of the upper area and the characteristic of the lower area relative to the skin tone cone.

The invention provides a hybrid gripper control method for grabbing a disordered scene and a hybrid gripper, relates to the technical field of robot grabbing, and aims to realize accurate robot grabbing in a dense disordered environment. The method comprises the following steps: acquiring a color image and a depth image collected for a current capture scene; performing target detection and segmentation according to the color image to obtain the position of a target object and a target object segmentation mask; according to the type of the target object and the position of the target object, a target grabbing mode is determined, and the target grabbing mode is one of clamping jaw grabbing, suction cup grabbing and electromagnet grabbing; generating a target grabbing posture corresponding to the target grabbing mode according to the target object segmentation mask and the depth image; and according to the target grabbing posture, generating a collision-free grabbing track, and according to the grabbing track, controlling a hybrid gripper to grab the target object according to the target grabbing mode.

The invention belongs to the technical field of burn wound analysis. According to the skin burn analysis method and system based on contrast agent blood perfusion, firstly, a fluorescence sequence after a contrast agent is injected is obtained, and a preprocessing sequence is generated through area cutting and down-sampling; a fluorescence intensity time curve of each pixel point is extracted, peak time is determined, and statistics is carried out to construct a peak time histogram; performing smooth denoising, extreme value detection and weak peak removal on the histogram, and adaptively selecting a threshold value for distinguishing normal and burnt tissues; generating an initial mask by using the threshold value, and performing connected domain analysis and morphological post-processing to obtain a target burn area mask; and finally, generating a pseudo-color image in combination with peak time. According to the method, the time dynamic characteristic of blood perfusion is fully utilized, the burn area can be accurately segmented and the microcirculation disturbance degree can be visually presented without depending on skin textures or a large amount of training data, and the objectivity and accuracy of burn depth evaluation are remarkably improved.