37results about How to "Rich image" patented technology

The invention discloses a deep learning and MRI (Magnetic Resonance Imaging) image-based ventricular function index prediction method and belongs to the field of medical image processing. A conventional ventricular index prediction method mainly depends on a process of artificially segmenting a ventricular muscle part of each phase and then performing measurement and calculation based on the artificial segmentation. The process needs to consume a large amount of manpower and time and has a serious artificial difference. The deep learning and MRI image-based ventricular function index prediction method is implemented by the following steps of clinically obtaining a heart MRI image; artificially drawing a ventricular outline and calculating related ventricular function indexes; preprocessing the heart MRI image; performing characteristic representation on heart MRI data by adopting a deep learning method; training a ventricular function index prediction model by adopting a machine learning method; and performing ventricular function index prediction on the heart MRI image obtained in the first step by adopting the trained model. According to the method, the ventricular function indexes can be quickly, accurately and automatically predicted to assist in diagnosis of clinical heart diseases.

The invention relates to a dynamic dragging and constructing module system with a 3D virtual element for an on-line exhibition, which comprises a network server, a client side computer and an Internet, wherein the network server is provided with a fire wall and is respectively connected with an organizer client side, an exhibiter client side and an audience client side by the Internet, and a virtual system of the on-line exhibition is constructed. The dynamic dragging and constructing module system is characterized in that the 3D virtual element which can be dynamically dragged and constructedfor the exhibition is stored in the network server and is a set of 3D virtual models which are manufactured based on the WEB3D of an X3D standard, the models are decomposed into different X3D subprogram bodies by aiming at the type of a virtual object to form a module icon library, icons are utilized to drag the 3D models so as to quickly construct a tableaux, and an individuation 3D virtual space is automatically generated. The invention solves the problems of singleness, independency, lower efficiency and the like of the prior method for constructing the on-line exhibition, reproduces the advantages of fullness, visualization and spatial impression of an entity exhibition vividly and forms man-machine interaction, man-network interaction and virtual-real interaction effectively.

The invention discloses a multiwindow tracking image stabilization method. The multiwindow tracking image stabilization method comprises the steps that a large-visual-field image of a target scene is acquired; multiple tracking windows used in the image stabilization tracking process are acquired according to the image characteristics of tracking windows; secondary imaging is conducted on the target scene, and a window image of each tracking window is acquired and stored and used as a reference image; the window image of each tracking window is updated and acquired, cross-correlation function calculation is conducted on a current window image and the corresponding reference image, weighting coefficients of each window image are acquired according to the change of related peak values of each window image, offset of each window image is weighted, and then overall offset of a focal plane image of an imaging optical system is acquired; by the utilization of the statistical characteristics of the offset of each window image, effective focal plane image offset is converted into rotating position control amount of a swing mirror in the imaging optical system and then converted into analog quantity, and according to the analog quantity, a swing mirror driver is controlled to drive the swing mirror in the imaging optical system to generate offset until the imaging optical system finishes high-resolution image imaging.

The present invention discloses a method and a device for management of household appliances which can solve the problems that a process for management of household appliances is tedious and is not easy to operate. A determines and displays a corresponding room identifier according to a login account at present, the terminal determines and displays a device identifier corresponding to the room identifier selected by the user according to the binding relation of the room identifiers and the device identifiers, at the same time, the terminal can display the three-dimensional device display information corresponding to the determined device identifier, and finally, the terminal sends control information to a device corresponding to the device identifier selected by the user according to a control command triggered by the user. The method can display the three-dimensional device display information corresponding to the determined device identifier to more vividly and concretely display thedevice information so as to bring new experience for users, omit the user operation for finding devices required being controlled from the whole family identifiers and more conveniently manage and control devices by the users.

The invention discloses a polarized image defogging method for obtaining transmissivity based on differential polarization, and belongs to the technical field of image defogging processing. The method comprises the following steps of acquiring two orthogonal polarization images of the same scene by using a polarization imaging method, establishing an atmospheric scattering model based on the foggy image physical model, and respectively calculating bright and dark channel images of the two sub-images in combination with bright and dark channel priori to obtain a transmissivity image t1 of the dark channel and an atmosphere light value at infinity, according to a difference image defogging method, approximately replacing the optical thickness by fog concentration to acquire a difference transmittance graph t2; and combining the dark channel transmissivity image with the differential transmissivity image to obtain a total transmissivity image, carrying out threshold processing on the total transmissivity image, and then carrying out refinement processing by using guide filtering; and substituting the obtained transmissivity and the atmospheric light into an atmospheric scattering model to obtain a fogless image. Image quality is obviously improved, more object reflected light information is obtained, and the definition of the restored image is improved.

The invention relates to an image augmented reality method and device. The method comprises the steps: calling image collection equipment to scan an image; carrying out the augmented reality recognition of the image, and obtaining an augmented reality content identification of the image; obtaining the URL (uniform resource locator) of an augmented reality content corresponding to the augmented reality content identification according to the augmented reality content identification; transmitting an augmented reality content request according to the augmented reality content identification and the URL; receiving the augmented reality content returned according to the augmented reality content request; obtaining a framing image, carrying out the overlapping of the framing image and the augmented reality content, and obtaining an augmented reality image. According to the invention, the method and device enable the framing image and the augmented reality content to be overlapped, obtain theaugmented reality image, enable the image contents to be abundant, and displays abundant information.

The invention discloses an image data processing method and apparatus and a computer readable storage medium. The method comprises the steps of obtaining a first sample image and a second sample image; wherein the first sample image comprises a sample source object; wherein the second sample image comprises a sample virtual object; obtaining key point position information of the sample source object, and generating a to-be-discriminated image corresponding to the first sample image based on the initial image generation model, the first sample image and the key point position information of thesample source object; wherein the to-be-discriminated image comprises a to-be-discriminated virtual object corresponding to the sample source object; correcting network parameters of the initial image generation model based on the sample source object, the sample virtual object and the to-be-discriminated virtual object; and determining the initial image generation model containing the correctednetwork parameters as an image generation model. The image content of the to-be-discriminated image can be enriched, the matching between the to-be-discriminated image and the first sample image can be improved, and the display effect of the to-be-discriminated image can also be improved.

A modulator and an apparatus comprising the modulator comprising a Tunable Diffraction Grating (TDG) component is arranged with a plurality of individual modulator sections, wherein a prism with appropriate chromatic and polarization filters provides at least two independent overlaid two dimensional images providing stereographic displays for three-dimensional images.

The invention discloses an image-based air quality grade evaluation method, and relates to a machine learning related technology, in particular to an image processing technology. The objective of theinvention is to solve the problems of high air quality detection cost and incapability of covering each corner in a city in the prior art. The method comprises the following steps: firstly, performingHSI space conversion on an image; performing DCT transformation on the saturation channel S and the brightness channel I of the image, converting the saturation channel S and the brightness channel Iinto a frequency domain, finally, the edge pixel ratio, the high-brightness pixel ratio, the low-saturation pixel ratio and the low-frequency pixel ratio of the image are extracted to serve as features, the features are input into an SVM to be classified, and therefore an evaluation model of the air quality grade is constructed. The air quality grade evaluation model constructed by the method canaccurately estimate the air quality condition of an image acquisition place, air quality detection becomes low in cost and fine in granularity, and people can conveniently monitor the surrounding airpollution condition anytime and anywhere. And the method is more suitable for evaluating the image with moderate air humidity, rich image content and sky area proportion less than 1/2.

The invention relates to a method for generating a visual industrial configuration interactive graph, which is used for simulating a visual interactive industrial configuration graph method of actual equipment by selecting a virtual object component from a user-defined component library to bind with a database and a field data point, so that the industrial configuration graph can be quickly established. The process of the method mainly comprises: uploading a self-defined configuration graph source, generating a virtual model component, defining a node template, defining a pipeline template, binding a data source, giving an alarm by equipment, and checking a historical trend graph. The method is mainly characterized by comprising: constructing simplified operation through a self-defined template and a layout component, monitoring mouse operation of a user, and performing corresponding event processing at any time, so that the interactivity of a configuration graph is improved. The industrial configuration can be adjusted in a self-defined manner, the flexibility is high, the universality is high, the display is more visual, and the production equipment condition of an industrial process control system can be well simulated.

The invention discloses a combined type candy comprising a head part, an upper half part, a lower half part, and arms which are arranged at two sides of the upper half part. The head part, the upper half part, the arms which are arranged at two sides of the upper half part, the lower half part and connection pieces are made of candies; and each individual candy can be directly eaten and is varied in shape having the choices of a human shape, an animal shape or plant and similar shapes.

The invention discloses a combined firework capable of displaying a projected image. The combined firework comprises a smoke effect barrel combination, a firework effect barrel combination, an image projector and a smoke ejecting nozzle device. The smoke effect barrel combination is arranged at the front end of a combined firework body. The firework effect barrel combination is arranged several rows behind the smoke effect barrel combination. The image projector is arranged in the middle position in the rear part of the combined firework body and comprises a box body, a reflective mirror, a collecting mirror, a magic lamp picture, a bracket, a projecting lens, a projecting tube, a light emitter, a U-shaped slot, a smoke exhaust window and a picture inserting opening. The above parts are installed in the box body orderly. The collecting lens, the magic lamp picture and the projecting lens are arranged in the projecting tube. The smoke exhaust window and the picture inserting opening are further arranged on the top end surface of the box body. The U-shaped slot is arranged in the middle of the projecting tube. The light emitter is formed by a light emitting core and a core conductive lead. The image projector can be further assembled and matched with a sliding rod in a sliding manner. The combined firework has the advantages of simple structure and easy implementation and the like.

The invention discloses a digital twinning method based on a digital-analog separation data circulation mode, in engineering project management application based on digital twinning, information of various projects is closely associated with a digital twinning model through uniform codes, a user obtains required information through human-computer interaction with a three-dimensional model, and the information is stored in the digital twinning model. And the three-dimensional model is combined and visually displayed to a user. The three-dimensional model is connected with the structure tree through codes in a hanging mode, attributes are not attached to the three-dimensional model any more, and digital-analog separation is truly achieved. According to the method, plug-ins are designed for the model, a WebGL digital twin graphic engine is used for rendering the three-dimensional model, the three-dimensional model is displayed in a picture stream mode in a virtual desktop mode, and lightweight conversion of digital twin model source files is achieved.

The invention discloses a role generation system for a game, and the system comprises a scanning module, a bone generation module, a skin generation module, a filling and forming module, a clothes making module, a wearing module, and a human-machine interaction module. In the invention, based on the simulation object, a game role model is generated so that a game user could generate the game role model using himself or herself as the blueprint or simulating the idols according to its own inclination. The clothes making model enables the game user to have a wider possibility to select clothes, maximizing the aesthetics of the user to the fullest and making the personalities in the game richer and more colorful. In the invention, the wearing module can load the adjusted target clothes model to the role model after the target clothes model is adjusted according to the role model, therefore, benefiting to increase the matching degree of the clothes and the role and improving the role image.

The invention discloses a binocular vision positioning method based on a semantic target, and relates to the binocular vision positioning method based on the semantic target. The invention aims to solve the problems that the existing indoor positioning methods need to know the scene layout before positioning, the layout cost is high, the positioning accuracy is low, and indoor positioning cannot be carried out in an unknown environment. The method comprises the steps of 1, acquiring a left image and a right image; 2, identifying semantic targets contained in the current left and right images and angular point coordinates corresponding to each semantic target; 3, selecting, by a user, a common semantic target contained in the left image and the right image in the semantic targets, and building a three-dimensional coordinate system of the target based on the angular point coordinates corresponding to the semantic target; 4, determining corresponding pixel coordinates of the angular point of the selected semantic target in the left image and the right image; 5, calibrating a binocular camera; and 6, solving the position coordinates and the steering angle of the current user relative to the target to realize the positioning of the user. The invention belongs to the field of image processing.

The embodiment of the invention provides an image processing method and device, a terminal and a storage medium. The image processing method can comprise the steps: carrying out the merging of a pixelat a target position in a first image layer and a pixel at the target position in a second image layer, and obtaining a first merging result; obtaining an Alpha value of the target position in the Alpha layer; performing merging processing on the Alpha value and the pixel of the target position in the third image layer to obtain a second merging result; and generating a target image according tothe first merging result and the second merging result. By adopting the image processing method and device, the image content and hierarchy are richer and are not single any more.

The invention discloses a virtual character image processing method and device, equipment and a storage medium, and belongs to the technical field of human-computer interaction. The method comprises the following steps: displaying a virtual character image and a color adjusting control of a first independent dyeing area on a user interface, wherein the first independent dyeing area is one of at least one independent dyeing area on an appearance part of the virtual character image, and the color adjusting control is used for adjusting color parameter values on the first independent dyeing area; in response to an adjustment operation on the color adjusting control, displaying the virtual character image after color adjustment; and in response to an application operation on the color-adjusted virtual character image, applying the color-adjusted virtual character image when the virtual task is executed. According to the method, independent dyeing areas are divided for appearance components of a virtual character image, then color parameter values on the independent dyeing areas are self-defined, and a unique appearance is customized for a three-dimensional virtual character.

The invention relates to the technical field of utilization, research and development of renewable resources, in particular to a renewable silica gel heat transfer label and a manufacturing method thereof. The renewable silica gel heat transfer label comprises a base layer, a transition layer, a first bonding layer, a raw silica gel layer, a second bonding layer, a renewable silica gel layer and asilica gel protective layer, wherein the base layer, the transition layer, the first bonding layer, the raw silica gel layer, the second bonding layer, the renewable silica gel layer and the silica gel protective layer are sequentially arranged in a laminated mode from bottom to top. According to the invention, the renewable silica gel heat transfer label is made of a renewable silica gel material; the utilization rate of renewable silica gel is remarkably increased, and the content of renewable silica gel powder in the renewable silica gel heat transfer label reaches 20% or above, which meets the global recovery standard (GRS) certification; and in the aspect of the use method of the renewable silica gel powder, the renewable silica gel powder does not need to be made into a gel solution, is directly attached to the heat transfer label and is wrapped by the silica gel protective layer, so the renewable silica gel powder is protected from falling off.