147 results about "Sketch map" patented technology

The invention discloses a perceptual loss and deep convolutional neural network-based sketch simplifying method. The method comprises the steps of: 1) obtaining data: obtain the data by using a sketchmap and corresponding label data; 2) processing data: transforming the sketch map and corresponding line sketch map of an image data set into a format required for training a deep convolutional neural network through preprocessing; 3) constructing a model: constructing a deep convolutional neural network suitable for sketch simplification according to a training objective; 4) defining a loss function; 5) training the model: calculating the loss value of the network according to the loss function, calculating the gradient of each network layer parameter through back propagation and updating parameters of each layer of the network through a random gradient descent method; and 6) verifying the model: verifying the trained model by using a verification data set and testing the generalizationperformance. The method disclosed by the invention has the advantages that sketches with rougher miscellaneous lines and unclear main structure lines can be processed by a sketch simplifying network;and the network has strong robustness to the influence of illumination.

The invention discloses a geological survey method of a tunnel face based on a digital image. The geological survey method comprises the following steps of: A, confirming a target object and marking the tunnel face; B, photographing to the tunnel face by using a high-resolution digital camera to shoot a complete tunnel face region; C, calculating the proportion conversion coefficient of a real object and an image according to the distance of two marking points on the tunnel face and the pixel value of two points in the image; D, performing grey level transformation, color equalization, brightness/contrast adjustment, hue/saturation control and image filtering to logarithms; E, calculating all pixels in the digital image; F, calculating extended length of structural surface, total extension of unit area crack, linear fit of structural surface, viewing dip angle of structural surface and space structural surface parameter of structural surface by the binary image of the structural surface to obtain geological survey achievements of the tunnel face. The geologic sketch map of the tunnel face is obtained by the method, so that the geological sketch speed and the geological survey precision and accuracy are improved.

The invention provides a 3D printing preparation method of a polymeric gradient material. The method comprises the following steps: (1) preparing inorganic components, polymers and other raw materials; (2) establishing a three-dimensional computer sketch map of the polymeric gradient material, performing slicing processing, and establishing a scanning path; (3) calculating the component ratios at different positions; (4) setting the corresponding printing parameters; and (5) starting printing equipment for performing 3D printing, simultaneously injecting the inorganic particles and polymers into the printing head, heating, mixing, and printing and forming according to the set track and speed. The method disclosed by the invention has the advantages that the raw material component ratios at different positions are controlled by the computer, so that gradient distribution of the components is really realized. The 3D printing technology is a material increase manufacturing method, and compared with the traditional process, the method disclosed by the invention has the advantage that the raw materials are saved.

A project management system is disclosed and comprises a project progress database, a project progress sketch map, a project progress monitoring system, a project standard template picture which is generated via a project construction design drawing, a project implementation site imaging system, an implementation site processing system and a remote monitoring terminal. Via the project management system, a project implementation site can be monitored remotely; monitoring operation can be conducted in a simple, convenient and direct manner; manpower resources can be saved. Feedback on construction conditions of an implementation site can be given timely due to real time monitoring operation, convenience of determining operation can be realized, and high result accuracy can be achieved. Project progress can be automatically adjusted and monitored according to change of implementation conditions via the project management system, and project completion time can be effectively monitored.

The invention discloses a polarimetric SAR (synthetic aperture radar) terrain classification method based on semantic information and polarimetric decomposition. The polarimetric SAR terrain classification method includes performing mean shift on a span image, extracting a sketch map of the span image, extracting a line segment gathering region in the sketch map by an region extracting technology based on the semantic information, merging span image mean shift over-segmentation regions based on the line segment gathering region as well as by the aid of a critical region majority vote merging strategy and based on a polarimetric feature merging strategy, acquiring an image segmentation result, and fusing the image segmentation result based on the semantic information and an H/alpha-Wishart classification result based on an MRF (markov random field) to acquire a final classification result. The semantic information, an image processing technique and a polarimetric scattering property are combined by the polarimetric SAR terrain classification method, the problem of bad region consistency of existing classification techniques based on the polarimetric decomposition to classification results with gathering-featured surface features (like forest, building groups and the like) is mainly solved, and region consistency and boundary retainability of the classification results with the gathering-featured surface features are improved.

The invention relates to a method for rapidly and safely predicting a tunnel face of an underground cavity on the basis of an image. The method comprises the following steps of 1, shooting by an infrared camera to acquire a tunnel face image in a current mileage; 2, carrying out image preprocessing on the tunnel face image; 3, carrying out clustering analysis; 4, carrying out lithologic identification to obtain lithologic classified information; 5, carrying out edge detection on the image preprocessed in the step 2; 6, carrying out boundary extraction to obtain bedding and jointing information of the tunnel face; 7, drawing a geologic sketch map of the current tunnel face by combining the lithologic classified information and the bedding and jointing information, and obtaining structural surface information of the current tunnel face; 8, moving to a next mileage and repeatedly carrying out the steps 1 to 7; 9, according to the extracted structural surface information, carrying out three-dimensional geologic modeling; 10, according to a model formed by the three-dimensional geologic modeling, carrying out rock mass analysis, predicting stability of rock masses and finally obtaining a cavity safety predicting result. Compared with the prior art, the method has the advantages of low noise, visual modeling and the like.

The present invention relates to a closed loop control system of a press, which pertains to the technical field of a control system of the press. The present invention is mainly characterized in that a coder is connected and arranged at the relevant moving part by an industrial field bus, the industrial field bus is connected with DeviceNet host station, and the DeviceNet host station exchanges data with a PLC programmable logic controller; a touch screen and an office computer adopt a closed loop system which consists of Ethernet and the PLC programmable logic controller; the present operation of a lathe can be reflected in real time under the support of various software programs; by using the closed loop control system, the safety of the lathe can be greatly increased, an automatic production line interface is arranged, the real-time operation information of the lathe can be displayed on the touch screen in the form of character, flow chart, sketch map, and the like, for convenient understanding, the statistics of relevant parts can be carried out according to the design parameters of the lathe so as to remind users of regular maintenance. The present invention can obviously increase the work efficiency and facilitate the arrangement and statistics of the work for supervisors.

The invention provides a gold ion imprinting magnetic adsorbent as well as a preparation method and application thereof. The adsorbent is prepared from a magnetic Fe3O4 core, a SiO2 layer, an MPS coating layer and a PVIM modified IIP coating layer, wherein the outside of the magnetic Fe3O4 core is sequentially wrapped by the SiO2 layer, the MPS coating layer and the PVIM modified IIP coating layer, and the IIP coating layer is in bond connection with imidazolyl. The preparation method of the adsorbent comprises the following steps: utilizing a magnetic Fe3O4 nano particle as the core and preparing Fe3O4SiO2MPS with a core-shell structure by utilizing a silica gel coupling method and utilizing TEOS and MPS as coupling agents; triggering polymerization reaction by utilizing a surface imprinting technology and combining free radical, utilizing Au(III) as a formwork ion, VIM as a functional monomer, EGDMA as a cross-linking agent and AIBN as an initiator, and synthesizing PVIM functionalized PVIM-Au-IIP on the surface of the MNPs to obtain a magnetic material coated with the IIP coating layer; utilizing an HCl solution containing thiourea to clean the material and removing the formwork ion to obtain the gold ion imprinting magnetic adsorbent. The synthesis sketch map of the adsorbent is shown in the description. The method can be used for simply and efficiently preparing the adsorbent which has high selectivity, high adsorption capacity and quick adsorption kinetic performance to Au(III), and the adsorbent has a good application prospect in the field of Au(III) adsorption.

The invention discloses a multi-type combined-hole-group numerical control processing method. The method comprises the following steps of step1, establishing a hole group distribution database, a hole type database, a hole processing technology parameter database and a numerical control processing code database; step2, establishing a man-machine interaction interface, logging and calling data of the hole group distribution database and generating a simulation hole group distribution sketch map; step3, establishing the man-machine interaction interface, logging and calling data of the hole type database, the hole processing technology parameter database and the numerical control processing code database; step4, setting a main shaft moving path optimization principle, optimizing calculation and automatically generating a main shaft moving optimization path; and step5, according to the main shaft moving optimization path, automatically generating a hole-group numerical control processing code program. In the invention, generation of single-type hole group and multi-type combined-hole-group processing codes can be realized; a work program is simple; an automatic programming function is possessed; cost is low and the method is suitable for parts used for processing regular hole groups.

The embodiment of the invention provides a reconnaissance method, a reconnaissance device and network optimization equipment. The reconnaissance method comprises the following steps of: acquiring longitude and latitude parameters of an indoor space in which the equipment is by utilizing a global positioning system (GPS) and assisted global positioning system (A-GPS) module; acquiring directionality parameters of the indoor space by utilizing a compass module; acquiring a spatial distance parameter of the indoor space and a position parameter of the equipment in the indoor space by utilizing a laser ranging module; and marking a geographic sketch map of the equipment in the indoor space on a display module by utilizing the longitude and latitude parameters, directionality parameters and spatial distance parameter of the indoor space and the position parameter of the equipment in the indoor space. The embodiment of the invention can improve the reconnaissance efficiency and save the labor cost.

The invention provides a sketch map extraction method for a polarization SAR image. The method comprises the steps that the polarization SAR image to be classified is input, and refined Lee filtering is carried out on the SAR image; a weighted edge filter is constructed to detect the constant false alarm edge of the image processed in the step 1 to acquire a polarization energy map; a weighted gradient filter is constructed to detect the gradient edge of the image processed in the step 1 to acquire a gradient energy map; wavelet transformation is carried out on the polarization energy map and the gradient energy map; different policies are constructed to fuse high frequency and low frequency sub-bands acquired through wavelet transformation; the polarization energy map and the gradient energy map are fused, and non-maximal value suppression is carried out on the fused maps to acquire an edge map; sketch operator tracing is carried out on the edge map to acquire sketch lines; and a statistical hypothesis test method is constructed to select the sketch lines and acquire a polarization sketch map. Polygon and noise-induced pseudo-boundary phenomena are suppressed, and the structure information of the polarization SAR image is accurately characterized.

The invention discloses a ray completion region graph and characteristic learning-based SAR (synthetic aperture radar) image segmentation method. The problem that a large homogenous region such as a farmland cannot be accurately segmented by a conventional segmentation method is mainly solved. An implementation process comprises the following steps: 1, obtaining a sketch map of an input SAR image by using an initial sketch model; 2, completing sketch lines by virtue of a ray clustering method, and acquiring a region graph; 3, segmenting the SAR image into a concentrated region, a homogenous region and a structural region of a pixel space by utilizing the region graph; 4, performing characteristic learning and clustering on the concentrated region and the homogenous region by utilizing a bag-of-words model respectively, performing watershed segmentation and sketch line-guided superpixel combination on the structural region, and combining combined superpixels to the homogenous region to finally obtain an SAR image segmentation result by utilizing gray characteristics. Compared with the prior art, the method has the advantages that the segmentation result is high in region consistency, and boundaries and line targets are positioned more accurately.

The invention discloses a remote sensing image unsupervised target detection method based on prior information and aircraft shape, which mainly solves the problems of blind positioning and slow positioning speed of optical remote sensing image target detection in large and complex scenes at present. The invention adopts an unsupervised method for aircraft positioning and accurately positioning. The realization steps are as follows: 1. establishing the initial sketch model for the original drawing and extracting the initial sketch; 2. according to the position relation of the line segments in the sketch map, finding out the sketch line segments which may constitute the wing of the aircraft; 3. from the result of the previous step, the sketch line segment that can compose the symmetrical wing being found; 4. using the information of the sketch map and the line and surface of the original map, the region around the sketch line segment of the symmetrical wing being grown to obtain the surface of the aircraft; 5. the outer rectangular frame being obtained according to the surface area of the aircraft, and the outer contour of the aircraft being obtained at the same time, so that the precise positioning can be realized. Fast positioning of airplane target in optical remote sensing image is realized, which can be used for airplane target recognition and image interpretation in opticalremote sensing image.

The invention discloses a sketch map and priori constraint based high-resolution SAR image road network detection method, which mainly solves the problem that the road detection accuracy of the priorart is low. The scheme includes: 1) using a sketch model of an SAR image to obtain a sketch map of surface feature sparse representation; 2) extracting suspected road sketch line segments from sketchline segments representing multiple semantics according to a prior constraint and statistical information of a road; 3) calculating the spatial extension degree and the grayscale distribution connection degree of other sketch lines in a neighborhood region for the current suspected road sketch line segments, and combining the two features to further extract sketch line segments representing the road; and 4) performing heuristic connection on the sketch line segments and performing superpixel segmentation around the connected sketch line, and determining a final road region according to the segmentation result. The invention improves the accuracy of road detection, reduces a large number of false alarms, and can be used for high-resolution SAR image road network detection of complex large scenes.

The invention discloses a method for calculating working performance of static-pressure sliding seats on the basis of a fluid-solid coupling effect. The method comprises the following steps of: (1) analyzing stress condition of a static-pressure sliding seat according to an actual working condition, and establishing a sliding seat stress analysis sketch map; (2) simplifying the sliding seat to a thin plate model and an oil pad model according to a thin plate deformation theory and a basic structure size of the static-pressure sliding seat; (3) calculating pressure distribution and oil film thickness change of the static-pressure sliding seat oil pad model, and establishing an equation; (4) calculating deformation condition of the static-pressure sliding seat thin plate model under the action of external force, and obtaining a deflection function of the static-pressure sliding seat thin plate model; (5) establishing a sliding seat-oil pad fluid-solid coupling model through a mutual coupling relationship between sliding seat deformation and the oil film thickness change, and establishing a dynamic balance equation of the sliding seat-oil pad fluid-solid coupling model; and (6) dynamically analyzing the working performance of the static-pressure sliding seat under the actual working condition by utilizing matlab software so as to finally obtain fluid-solid coupling effect-based working performance valuation of the static-pressure sliding seat under the actual working condition.

The invention discloses a drawing method and device for sketching. The method comprises the following steps of acquiring a selected image, and carrying out gray processing on the selected image to obtain a processed image; according to the gray value of pixel points in the processed image, establishing the gray distribution curve of all pixel points in the processed image; dividing all pixel points of the processed image into several groups by taking the minimum points in the graph as segmentation points; normalizing the gray value of each pixel point located in the same group to obtain the average gray value of each group; according to the position of pixel points in each group in the processed image, determining the corresponding drawing line sub-paths of each group, and determining thedrawing times of each drawing line sub-paths according to the average gray value of each group; according to the drawing times of each drawing line sub-path and each drawing line sub-path, calculatingthe total drawing line path, and drawing the sketch map of the selected image according to the total drawing line path. By implementing the embodiment of the invention, the sketch map of the selectedpicture can be automatically drawn according to the selected picture, and the sketch map of the selected picture can be obtained.

The invention discloses a computer virtual three-dimensional scenario library-based synthetic shooting system, which comprises shooting equipment, a three-dimensional scenario library, a shot information acquisition and real-time synchronization module, a shooting assisting module, an information integration module, a scenario rendering module and an image synthesis module, wherein the shooting equipment can project an effect sketch map of a virtual three-dimensional scenario onto a backdrop, and does not produce any projection on the backdrop at the moment of shooting; the information integration module integrates the position and direction information of the shooting equipment and information shot by the shooting equipment to obtain virtual spatial position information; the scenario rendering module renders a virtual scenario picture which is completely matched with a real shooting parameter and a real shooting environment; and the image synthesis module synthesizes a virtual scenario image and an original image to generate a synthetic image. The synthetic shooting system is low in cost and free from the influence of natural conditions, has authentic effects, and can be widely applied to the photographic industry, and a sufficient selective space and a sufficient reaction space can be provided for staff in the photographic industry.

The invention discloses a large field of view image sharpening device and method based on a focal plane Hartmann wavefront sensor. Spatial information is recorded by a light spot array image detectedby a CCD detector in the focal plane Hartmann wavefront sensor, and a microlens array records phase information which are incident light waves of different fields of view or angles. Based on a speciallight field structure, wavefront information of multiple fields of view can be measured at one time so as to achieve an effect of a large field of view. A traditional wavefront detector is generallysmall field of view or zero field of view, and only an intermediate field of view can be detected. After original image information is acquired by an image acquisition device, the device and the method of the invention can be used to restore an entire blurred image area, but only a part of the image can be restored based on the traditional wavefront detector. In the invention, a wavefront error large field of view measurement problem is solved, and an image restoration range is greatly increased and recovery precision is enhanced when the wavefront detector has a certain detection caliber. A,full recovery is achieved; b, a traditional wavefront sensor can only restore areas in a circle; and c, multi-field-of-view simultaneous measurement of the focal plane Hartmann wavefront sensor is combined into a large field of view sketch map.