30 results about "Colored graph" patented technology

The invention belongs to a metal model manufacturing technology, and in particular relates to a method for manufacturing a colored metal model by a three-dimensional (3d) digital model. The method comprises the following steps of: arranging the model, unfolding the model, composing graphs, manufacturing etching graphs, etching a metal sheet material, manufacturing an etching sheet surface coating and a colored printing graph, and printing and assembling etching sheets to form the metal model. A die is not required to be opened, and the 3D digital model can be composed to form a planar graph for processing a metal thin plate so as to produce various types of metal models. A planar metal plate can be directly printed and colored, various high-precision colored patterns can be printed or stamped on the metal plate, and the metal model is bright in color and vivid in pattern. The method is suitable for small-batch production and can meet humanized requirements of different objects.

The invention discloses a class picture conversion method in power load prediction. A new method for converting time series data into data of a class color picture structure is provided. The converteddata is inputted into an improved convolutional neural network using a special-shaped convolution kernel according to a certain logic, and the capacity of extracting local and overall implicit feature rules in a time * feature matrix is enhanced. In an actual power short-term load prediction experiment, the training time is shortened, and the prediction precision is improved in comparison with amainstream method.

The invention discloses a layered template matching method based on a multi-dimensional pyramid. The method comprises the steps of clustering template data rendered offline according to viewpoint parameters during rendering, a pyramid structure under multiple dimensions is established, efficiency optimization in the matching process is achieved, and the method comprises the following steps of 1, in the offline generation process, obtaining a color map and a depth map; 2, constructing a multi-dimensional template pyramid; 3, in the online matching process, obtaining an input feature map; 4, obtaining a high-level matching result; 5, obtaining an approximate interval where the object is located, and taking the approximate interval as an ROI on the two-dimensional image; 6, performing a matching test on the template corresponding to the low-level pyramid; and 7, carrying out random sampling consistent detection on the matched attitude to obtain a final detection and attitude estimation result of the object. According to the method, the CAD model is used, the method is suitable for industrial application, quick query can be realized, and the balance between the matching speed and the matching precision is ensured.

The invention relates to a semantic SLAM robustness improvement method based on instance segmentation, and the method comprises the steps: firstly carrying out the instance segmentation of a key framethrough an instance segmentation network, and building prior semantic information; calculating a feature point optical flow field to further distinguish the object, identifying a real moving object in the scene, and removing the feature points belonging to the dynamic object; and finally, performing semantic association, and establishing a semantic map without dynamic object interference. Compared with the prior art, the semantic map is established by adopting a method of combining deep learning and optical flow, and the depth map is added on the basis of the color map, so that the system isendowed with the capability of establishing the dense three-dimensional point cloud semantic map. In addition, a Mask-RCNN framework is adopted for real-time semantic segmentation, and object dynamicinformation can be calculated through mutual combination of dynamic feature points estimated by optical flow information and pixel-level semantic information. According to the method, deep learning and optical flow are mutually combined, so that the robustness of the whole system is remarkably improved, and the method can be applied to real-time semantic map construction in a dynamic scene.

The invention provides a method for generating a three-dimensional grid model by using multiple color pictures, which is used for processing a small number of color pictures of different visual angles and corresponding camera parameters based on deformation inference so as to generate a corresponding three-dimensional grid model, and is characterized by comprising the following steps: S1, preprocessing a pre-acquired three-dimensional model data set to obtain a training sample; s2, generating an initial grid template; s3, constructing an image feature extraction network for extracting geometric features and semantic features of the two-dimensional image; s4, constructing a graph convolutional neural network; s5, constructing a loss function; s6, training a three-dimensional model generation model composed of the image feature extraction network and the image convolutional neural network based on the loss function; and S7, inputting the plurality of color pictures and the corresponding camera parameters into the three-dimensional model to generate a model so as to generate the three-dimensional grid model.

The invention provides a color object rapid three-dimensional measurement method based on a color response model. The color object rapid three-dimensional measurement method based on the color response model comprises the following steps of S1, encoding a color pattern, and calibrating a crosstalk matrix, S2, projecting a color coded graph, and acquiring deformed stripes by a CCD, S3, descrambling the acquired image, extracting an RGB channel, calculating the surface reflectance of the object, and correcting phase shift stripes, S4, obtaining phase information by using three-step phase shift, and S5, unwrapping the wrapped phase by using the level information. The color object rapid three-dimensional measurement method based on the color response model provided by the invention has the advantages that the interference of the object surface texture information on the coding pattern can be well weakened, the measurement precision is improved, and a better measurement effect is obtained.

The invention relates to the technical field of defect detection, and relates to a probability graph visualization method and device for defect detection. The problems that in the prior art, probability information of pixel point defect categories cannot be visually displayed, and background category related information is omitted are solved. The method comprises the following steps: inputting a to-be-detected target image into a semantic segmentation model to obtain a multi-channel probability graph corresponding to the to-be-detected target image; according to the multi-channel probability graph, obtaining the probability of the same pixel point in each category, and comparing to obtain the maximum probability of each pixel point; based on the category corresponding to each maximum probability, recording an index number of the category corresponding to each maximum probability; according to an index number corresponding to a preset category, judging a category corresponding to each maximum probability; according to the category, processing each maximum probability to obtain a single-channel probability graph containing the probability that each pixel point belongs to a certain category; and converting the single-channel probability graph into an RGB image to obtain a pseudo-color graph so as to realize visualization of the probability graph.

The invention provides a sketch coloring method and system based on a deep convolution generative adversarial network, and the method comprises: S100, extracting features of a sketch through a gray-scale map generator GGN, and carrying out deconvolution of the features of the sketch to generate a gray-scale map corresponding to the sketch; and S110, performing feature extraction on the grey-scalemap and the graffiti information of the sketch by the user by using a color map generator CGN, and performing deconvolution on the extracted grey-scale map and the graffiti information to generate a corresponding color map, wherein both the GGN and the CGN belong to a deep convolutional generative adversarial network. In the CGN network in the coloring stage, the advanced features of the sketch are extracted by skillfully utilizing the training model and are input into the CGN, so that the CGN network not only obtains the features of the grey-scale map, but also obtains the features of the sketch, and the coloring result is more accurate.

The invention discloses a depth map confidence estimation method based on a convolutional neural network, which is used for quality evaluation and post-processing operation of a depth map generated by a multi-view stereo matching algorithm, and comprises the following steps: calculating a truncated symbol distance function graph and a normal graph by using the depth map generated by the multi-view stereo matching algorithm; performing feature extraction on the truncated symbol distance function graph, the normal graph and the color graph by using a U-shaped network structure to obtain a feature graph; and using a convolutional long-short term memory structure, a prediction module, a refinement module and a multi-supervision method to predict the confidence of the depth map from the feature map and refine an estimation result. According to the method, quality evaluation can be carried out on the depth maps generated by various multi-view stereo matching algorithms, and the depth map confidence degree in multi-view stereo matching can be estimated robustly, so that evaluation of the multi-view stereo matching algorithms and post-processing of the depth maps are facilitated.

The invention discloses a test result display method and device based on a traffic network, equipment, and a computer readable storage medium. The method comprises the steps of performing functional unit splitting on to-be-tested systems according to service features; constructing an association relationship graph between the to-be-tested systems according to the functional unit splitting result;and coloring the association relationship graph according to the test results of the functional units of the to-be-tested systems and the calling relationship between the functional units to form a colored graph.

The invention provides an unmanned ship intelligent target detection method and system adopting a binocular camera. The method comprises the following steps: establishing a system directory; downloading or performing selfie on a network to establish a graph of the obstacle, processing the graph into a color picture, and placing the color picture in a directory; marking a picture by using software, and respectively placing the generated txt file and xml file in a system directory; randomly generating a training set, a test set and a verification set according to a proportion of 7: 2: 1, and placing the sets in a system directory; a voc.names file is newly built in a system directory, the content is the name of an obstacle, and one file is built in each row; a voc.data file is newly built in a system directory, and a parnet.cfg file is newly built in the system directory and used for defining a lightweight network structure. The target detection network realized by the lightweight network is relatively shallow, simple in structure and easy to realize; the target detection network realized by the lightweight network is symmetrical in structure, few in parameters and convenient to train and upgrade.

The invention discloses a graph-based generic genome data organization method, system and equipment and a computer readable storage medium. The method comprises the following steps: acquiring a group of generic genome sequence data; performing composition on the generic genome sequence data to obtain a generic genome coloring graph; marking and acquiring characteristics of an access state of a single node of the colored graph, and traversing the colored graph to obtain a cSupB data model after the colored graph is decomposed and data information of the cSupB data model; and determining an inclusion relation between the cSupB data models based on the data information of the cSupB data models, and constructing a cSupB structure tree model according to the inclusion relation. According to the method, the problems that the data organization mode is disordered and the readability, the effectiveness and the integrity of the sequence are poor when a large amount of genome data is targeted at present are solved.