CN112669426B - Three-dimensional geographic information model rendering method and system based on generation countermeasure network - Google Patents
Three-dimensional geographic information model rendering method and system based on generation countermeasure network Download PDFInfo
- Publication number
- CN112669426B CN112669426B CN202011559363.0A CN202011559363A CN112669426B CN 112669426 B CN112669426 B CN 112669426B CN 202011559363 A CN202011559363 A CN 202011559363A CN 112669426 B CN112669426 B CN 112669426B
- Authority
- CN
- China
- Prior art keywords
- dimensional
- ground object
- rendering
- geographic information
- target area
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000009877 rendering Methods 0.000 title claims abstract description 95
- 238000000034 method Methods 0.000 title claims abstract description 41
- 238000012549 training Methods 0.000 claims abstract description 25
- 238000004891 communication Methods 0.000 claims description 9
- 230000008485 antagonism Effects 0.000 claims description 2
- 238000000605 extraction Methods 0.000 abstract description 3
- 238000013473 artificial intelligence Methods 0.000 abstract description 2
- 239000011159 matrix material Substances 0.000 description 6
- 230000000007 visual effect Effects 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
- 230000001502 supplementing effect Effects 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
Landscapes
- Processing Or Creating Images (AREA)
Abstract
The invention discloses a three-dimensional geographic information model rendering method, a system, equipment and a storage medium based on a generated countermeasure network, wherein the method comprises the following steps: the method comprises a geographic information three-dimensional modeling step, a three-dimensional model generation training step, an all-terrain element matching step, a three-dimensional geographic information collaborative rendering step and a three-dimensional geographic information integral rendering step. According to the invention, the generation of the geographic information model is realized through an artificial intelligence method and a generation countermeasure network frame, the matching of all-terrain elements and the extraction of similar-terrain main body parts are realized through a sliding matching mode, and the non-repeated rendering of the similar three-dimensional geographic information model main body parts is realized through a pre-rendering mode in a cache in advance, so that the rendering speed of a three-dimensional object is improved.
Description
Technical Field
The invention belongs to the technical field of three-dimensional visualization of geographic information, and particularly relates to a three-dimensional geographic information model rapid rendering method, system, equipment and storage medium based on a generated countermeasure network.
Background
With research and development of remote sensing technology and sensor technology, traditional two-dimensional maps have been gradually replaced by three-dimensional maps with larger information and better visual effect.
At present, the two-dimensional map is a vector map or an image map, can realize efficient rendering on the basis of the existing hardware performance and optimization algorithm, and meets the requirements of real-time map display and use.
For the three-dimensional map, the data volume is larger, and the traditional rendering method needs to render all points pixel by pixel, so that huge index and rendering time are required to be consumed, real-time interactivity is poor, and extra waiting time is required to complete three-dimensional rendering of geographic information of big data in mobile products with limited computing capacity, such as smart phones.
Disclosure of Invention
In view of the above, the invention provides a three-dimensional geographic information model rapid rendering method, a system, a device and a storage medium based on generation of an antagonism network, which are used for solving the problem of low rendering speed of the existing three-dimensional map.
By means of pre-matching the three-dimensional ground objects of the area to be displayed and pre-rendering in a cache, the three-dimensional map rendering speed is improved.
In a first aspect of the present invention, a three-dimensional geographic information model rendering method based on generation of an countermeasure network is disclosed, the method comprising:
a step of three-dimensional modeling of geographic information, in which typical ground object features in the three-dimensional geographic information are classified, three-dimensional ground object objects in a selected area are extracted and stored separately according to the categories; scaling each type of three-dimensional ground object to a cube space with standard size;
generating a training step of the three-dimensional model, namely respectively training a CGAN model through each type of scaled three-dimensional ground object; generating a new three-dimensional ground object with the same characteristics as the ground object of the corresponding category through the trained CGAN model;
a full-feature element matching step, namely sliding and matching the cube spaces corresponding to the new three-dimensional feature objects in the original three-dimensional map to be rendered one by one, extracting an overlapping area successfully matched as a main body part of the similar feature, and caching the overlapping area;
a three-dimensional geographic information collaborative rendering step, namely acquiring a target area selected by a user, extracting a main body part of the target area, and indexing and prerendering the main body part of the target area in a cache;
and a step of integrally rendering the three-dimensional geographic information, wherein pre-rendering features in the cache are selectively loaded according to the target area and the zoom level selected by the user.
Preferably, in the step of three-dimensional modeling of geographic information, the scaling each type of three-dimensional ground object to a cube space with a standard size specifically includes:
for each three-dimensional ground object, extracting the maximum numerical value in the length, width and height coordinates of the three-dimensional ground object;
and scaling the three-dimensional ground object to be reduced into a standard cube space by taking the maximum numerical value as a reference.
Preferably, the all-terrain element matching step specifically includes:
dividing an original three-dimensional map to be rendered into three-dimensional grid blocks, and numbering each three-dimensional grid block through a number;
sliding the cube spaces corresponding to the new three-dimensional ground object one by one in the original three-dimensional map to be rendered to perform matching degree comparison;
for the successfully matched area, the cube space corresponding to the new three-dimensional object is subjected to multi-angle rotation and multi-scale scaling, and is matched with the area in the original three-dimensional map again, and the largest overlapping area is obtained through multi-round comparison;
and taking the overlapped area as a main body part of the similar object, caching the main body part, and recording the corresponding three-dimensional grid block number of the main body part in the original three-dimensional map.
Preferably, the step of sliding the cube space corresponding to the new three-dimensional object one by one in the original three-dimensional map to be rendered to perform matching degree is specifically:
and each sliding is performed in a pixel-by-pixel sliding or self-defining sliding step length mode, after each sliding is stopped, SSIM is adopted to evaluate the matching degree of the new three-dimensional object and the region in the original three-dimensional map, and if the matching degree is greater than a preset threshold value, the matching is successful.
Preferably, the step of collaborative rendering of the three-dimensional geographic information specifically includes:
acquiring a target area selected by a user, and extracting a small-scale three-dimensional ground object main body part of the target area based on a large-scale three-dimensional ground object of the target area;
indexing and prerendering a main body part of the three-dimensional ground object with a small scale in a cache;
preferably, the step of integrally rendering the three-dimensional geographic information specifically includes:
determining a corresponding three-dimensional grid block number according to a target area and a scaling level selected by a user, and searching whether corresponding prerendered data exists in a cache according to the three-dimensional grid block number;
and if the pre-rendering data exists, loading the corresponding pre-rendering data to the corresponding position of the target area.
Preferably, the step of integrally rendering the three-dimensional geographic information further includes:
and judging whether each pixel in the cube space of the target area is empty, and if so, directly rendering the corresponding pixel point in the original three-dimensional map to the position.
In a second aspect of the present invention, a three-dimensional geographic information model rendering system based on generating an countermeasure network is disclosed, the system comprising:
the geographic information three-dimensional modeling module is used for classifying typical ground object features in the three-dimensional geographic information, extracting three-dimensional ground object objects in a selected area and independently storing the three-dimensional ground object objects according to the types; scaling each type of three-dimensional ground object to a cube space with standard size;
the three-dimensional model generation training module is used for respectively training the CGAN model through each type of scaled three-dimensional ground object; generating a new three-dimensional ground object with the same characteristics as the ground object of the corresponding category through the trained CGAN model;
the all-feature element matching module is used for sliding and matching the cube spaces corresponding to the new three-dimensional feature objects in the original three-dimensional map to be rendered one by one, and taking the overlapping area successfully matched as the main body part of the similar feature and caching the overlapping area;
the three-dimensional geographic information collaborative rendering module is used for acquiring a target area selected by a user, extracting a main body part of the target area, and indexing and prerendering the main body part of the target area in a cache;
and the three-dimensional geographic information integral rendering module is used for selectively loading the prerendered ground objects in the cache according to the target area and the scaling level selected by the user.
In a third aspect of the present invention, an electronic device is disclosed, comprising: at least one processor, at least one memory, a communication interface, and a bus;
the processor, the memory and the communication interface complete communication with each other through the bus;
the memory stores program instructions executable by the processor, which are called by the processor to implement the method according to the first aspect of the invention.
In a fourth aspect of the present invention, a computer readable storage medium is disclosed, wherein the computer readable storage medium stores computer instructions that cause the computer to implement the method according to the first aspect of the present invention.
Compared with the prior art, the invention has the following beneficial effects:
1) Based on typical feature classification of the ground objects, the method generates rich three-dimensional ground object models by generating the countermeasure network, generates more matching objects for subsequent ground object matching, expands the number of the three-dimensional ground objects, improves the matching precision, and further improves the rendering quality;
2) The method comprises the steps of sliding and matching a three-dimensional object with an original three-dimensional map to be rendered, extracting an overlapping area successfully matched as a main body part of a similar object, and caching the main body part in a memory; when a user browses a map, three-dimensional features of a region to be displayed are pre-matched, pre-rendered in a buffer, and rendering of most features in the three-dimensional map can be realized only by directly loading a corresponding main body part from the buffer, so that response time is improved, the situation that all regions need to be repeatedly rendered point by point when the map is browsed every time is avoided, and rendering speed is greatly improved.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic flow chart of a three-dimensional geographic information model rendering method based on a generated countermeasure network;
FIG. 2 is a schematic diagram of a three-dimensional geographic information model rendering system based on generating a countermeasure network according to the present invention.
Detailed Description
The following description of the embodiments of the present invention will clearly and fully describe the technical aspects of the embodiments of the present invention, and it is apparent that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present invention without making any inventive effort, are intended to fall within the scope of the present invention.
In the embodiment of the invention, the rendering equipment adopts a computer with a CPU i7, a memory 16G and a video memory 8G, an operating system adopts Ubuntu16.04, a machine learning platform adopts Tensorflow2.0, a CGAN is adopted for generating an countermeasure network frame, and a three-dimensional map in 2019 in certain city is adopted for data.
Referring to fig. 1, the present invention proposes a three-dimensional geographic information model rendering method based on generating an countermeasure network, the method comprising:
s1, three-dimensional modeling of geographic information, namely classifying typical feature features in three-dimensional geographic information, extracting three-dimensional feature objects in a selected area and independently storing the three-dimensional feature objects according to the classification; scaling each type of three-dimensional ground object to a cube space with standard size;
specifically, a third party GIS tool can be used for classifying typical features of the ground in the three-dimensional geographic information, such as rivers, trees, grasslands, buildings, pavement and the like;
in all three-dimensional maps, a small area is selected, three-dimensional ground object objects such as rivers, trees, grasslands, buildings, pavement and the like in the area are extracted, and each type of three-dimensional ground object is independently stored.
For each three-dimensional ground object, extracting the maximum numerical value in the length, width and height coordinates of the three-dimensional ground object;
and scaling the three-dimensional ground object to be reduced to a cube space with unit standard by taking the maximum numerical value as a reference, and completing the standardized processing of the three-dimensional ground object.
For example, a three-dimensional object is 100 units long, 200 units wide, and 300 units high, the map needs to be scaled to a standard cube space of units 100 x 100, the height unit needs to be reduced to 100, the length unit needs to be reduced to 100/3 in an equal proportion, and the width needs to be reduced to 200/3 in an equal proportion; reducing the three-dimensional ground object to a standard cube space according to the largest dimension (one of length, width and height); the unit of the three-dimensional ground object can be set by different measurement units such as meters, decimeters, centimeters and the like by a user according to a specific three-dimensional object range, a scale and the like.
S2, generating a training step of the three-dimensional model, and respectively training a CGAN model through each type of scaled three-dimensional ground object; generating a new three-dimensional ground object with the same characteristics as the ground object of the corresponding category through the trained CGAN model;
specifically, a Tensorflow machine learning platform can be built under a Ubuntu operating system, and CGAN is selected as a generating countermeasure network frame;
converting the standardized three-dimensional ground object into a 3-dimensional matrix for storage, wherein the 3-dimensional matrix respectively identifies the relative coordinates of length, width and height, and the matrix for storage is used for mathematical calculation in the subsequent training process;
respectively training each type of three-dimensional ground object through a CGAN framework to generate a model, wherein the input of the training model is set as random noise; after iterative training, a new three-dimensional ground object with the same characteristics as the ground object is generated by using the trained model.
S3, performing full feature element matching, namely sliding and matching the cube spaces corresponding to the new three-dimensional feature objects in the original three-dimensional map to be rendered one by one, extracting an overlapping area successfully matched as a main body part of the similar feature, and caching the overlapping area;
because the main body part occupies a larger storage space and consumes larger resources for rendering, the step realizes the searching and positioning of the same main body part by matching the main body part of the ground object element, and is specifically as follows:
dividing an original three-dimensional map to be rendered into three-dimensional grid blocks, and numbering each three-dimensional grid block through a number;
sliding the cube spaces corresponding to the new three-dimensional ground object one by one in the original three-dimensional map to be rendered to perform matching degree comparison; each sliding is performed in a pixel-by-pixel sliding or self-defining sliding step length mode, if the sliding step length is larger, the matching speed is improved, but the matching precision is reduced.
After each sliding stop, the matching degree of the new three-dimensional ground object and the region in the original three-dimensional map is evaluated by adopting the structural similarity, and if the matching degree is larger than a preset threshold value, the matching is successful. The size of the three-dimensional ground object after the standardization processing is possibly inconsistent with that of the ground object in the original three-dimensional map, so that pixel-by-pixel matching cannot be performed;
for the successfully matched area, the cube space corresponding to the new three-dimensional object is subjected to multi-angle rotation and multi-scale scaling, and is matched with the area in the original three-dimensional map again, and the largest overlapping area is obtained through multi-round comparison;
and taking the overlapped area as a main body part of the similar object, caching the main body part, and recording the corresponding three-dimensional grid block number of the main body part in the original three-dimensional map.
S4, a three-dimensional geographic information collaborative rendering step, namely acquiring a target area selected by a user, extracting a main body part of the target area, and indexing and prerendering the main body part of the target area in a cache;
when a user selects to browse the three-dimensional map, a large-scale area is often selected first, and then small-scale ground objects are obtained in an amplifying, small-range moving and rotating mode, so that the target ground objects are used; therefore, when a user selects a certain area, a target area required to be observed by the user can be roughly confirmed, and because the visual display of the three-dimensional ground object with a large scale is realized during the area selection, at the moment, the small scale and more accurate three-dimensional ground object main body part of the area can be extracted based on the large scale area of the target area, and the index and the pre-rendering of the three-dimensional ground object main body part with the small scale of the target area are realized in the cache; according to the method, the cooperative rendering of the three-dimensional ground object of the region to be observed is realized by rendering the ground object main body part of the pre-browsing region through the multithread background buffer, so that the subsequent ground object rendering speed outside other current visible ranges is increased.
S5, the whole rendering step of the three-dimensional geographic information selectively loads the pre-rendered feature in the cache according to the target area and the zoom level selected by the user, and judges whether to render the rest according to whether the pixel data of the corresponding position in each three-dimensional grid block of the target area is empty or not. The method specifically comprises the following steps:
determining a corresponding three-dimensional grid block number according to a target area and a scaling level selected by a user, and searching whether corresponding prerendered data exists in a cache according to the three-dimensional grid block number;
if the pre-rendering data exists, the corresponding pre-rendering data is directly loaded to the corresponding position of the target area for display.
Because the pre-rendering data in the buffer memory is only the main body part of the three-dimensional ground object, the rest part needs to be supplemented for rendering, in order to achieve the rendering speed improvement of the rest part, whether each pixel of the corresponding position in each three-dimensional grid block of the target area is empty or not can be judged, if not, the pixel is occupied by the pre-rendering main body part, and if so, the corresponding pixel point in the original three-dimensional map is directly rendered to the position, so that the whole rendering of the three-dimensional geographic information is completed. Because only whether each pixel point is empty is judged, the method has a faster calculation speed, and the rendering speed of the whole three-dimensional ground object can be effectively improved.
The invention realizes the generation of the geographic information model by an artificial intelligence method and the generation of an countermeasure network frame; by means of sliding matching; realizing the matching of all-terrain elements and the extraction of main parts of similar terrain; the non-repeated rendering of the similar three-dimensional geographic information model main body part is realized in a pre-rendering mode in the cache in advance, so that the rendering speed of the three-dimensional object can be greatly improved.
Referring to fig. 2, corresponding to the method embodiment, the invention further proposes a three-dimensional geographic information model rendering system based on generating an countermeasure network, the system comprising:
the geographic information three-dimensional modeling module 10 is used for classifying typical feature features in the three-dimensional geographic information, extracting three-dimensional feature objects in a selected area and independently storing the three-dimensional feature objects according to the types; scaling each type of three-dimensional ground object to a cube space with standard size;
specifically, the geographic information three-dimensional modeling module 10 implements conversion of geographic information of an area to be displayed into three-dimensional object format data suitable for use in the present invention. Firstly, classifying typical feature characteristics in three-dimensional geographic information, such as rivers, trees, grasslands, buildings, pavement and the like, by using a third-party GIS tool; then, each type of three-dimensional ground objects is independently stored in a folder, so that the preprocessing of samples required by the anti-network training is facilitated, and meanwhile, each type of three-dimensional ground objects is required to be scaled to a uniform size standard in an equal proportion, and the three-dimensional ground objects are uniformly scaled to a cubic size of 5 m multiplied by 5 m; finally, repeating the steps, and scaling each ground object in each type of ground object to a cube space with standard size in a standardized equal proportion.
The three-dimensional model generation training module 20 is configured to train the CGAN model through each type of scaled three-dimensional ground object respectively; generating a new three-dimensional ground object with the same characteristics as the ground object of the corresponding category through the trained CGAN model;
specifically, the three-dimensional model generation training module 20 realizes generation training of the three-dimensional model of the same kind of ground objects, and generates more matching objects for subsequent ground object matching. Firstly, building a Tensorflow machine learning platform under a Ubuntu operating system, and simultaneously selecting CGAN as a generating countermeasure network frame; after the standardization treatment, converting the three-dimensional ground object with the side length of 5 meters into a 3-dimensional matrix, wherein the 3-dimensional matrix respectively marks the relative coordinates of length, width and height, and converting the three-dimensional ground object into matrix storage for mathematical calculation in the subsequent training process; further, training and generating a model for each type of three-dimensional ground object through the CGAN framework, wherein the input of the training model is set to be random noise; then, a trained model is used for generating a new three-dimensional ground object with the same characteristics as the ground object; finally, repeating the process, training a generation model for each type of three-dimensional ground object, and newly generating the type of ground object by using the generation model so as to expand the number of the three-dimensional ground objects.
The all-feature element matching module 30 is configured to slidingly match the cube spaces corresponding to the new three-dimensional feature objects in the original three-dimensional map to be rendered one by one, and take the overlapping area successfully matched as a main body part of the similar feature and cache the overlapping area;
specifically, the all-terrain element matching module 30 performs pre-rendering on the positioned terrain main body portion, so as to improve the three-dimensional rendering response time. Firstly, dividing all map areas with rendering into three-dimensional grid blocks, and numbering each grid block through numbers; then, sliding and comparing the three-dimensional object cube spaces with standard sizes in all the original three-dimensional maps to be rendered one by one, wherein each sliding can be performed in a pixel-by-pixel mode, or the sliding distance can be set in a self-defined step length mode, and the embodiment adopts the sliding distance of 10 unit pixels each time; after each sliding stop, comparing the matching degree of the three-dimensional ground object cube and the original three-dimensional map by adopting SSIM, and after the matching is successful, extracting the repeated area of the successfully matched area in the ground object cube and the original map, and obtaining the largest repeated area through multi-round comparison by rotating the ground object cube at multiple angles and scaling at multiple scales; finally, the body part is cached, and the three-dimensional grid number of the body part in the original three-dimensional map is recorded.
The three-dimensional geographic information collaborative rendering module 40 is configured to obtain a target area selected by a user, extract a main body portion of a three-dimensional ground object of the target area, and index and prerendering the main body portion of the target area in a cache;
specifically, the geographic information collaborative rendering module 40 implements the multi-thread background cache rendering of the ground object main body part of the pre-browsing area in the process of rendering the three-dimensional map in the visual range, so as to accelerate the subsequent ground object rendering speed outside the current visual range. Firstly, because the visual display of the three-dimensional ground object with a large scale is realized during the area selection, the pre-rendering of the ground object with a small scale and more accuracy can be started simultaneously; then, through the target observation area (usually around the center point of the display screen) selected by the user, the extraction of the main body part of the three-dimensional ground object with the small scale in the area is realized by the method described in the previous step; furthermore, the indexing and prerendering of the ground objects with the small scale of the observation area are realized in the cache; finally, the collaborative rendering of the three-dimensional ground object to be observed in the buffer memory is realized in the process of observing the current three-dimensional map by the user.
The three-dimensional geographic information overall rendering module 50 is configured to selectively load the pre-rendered feature in the cache according to the target area and the zoom level selected by the user, and determine whether to render the rest according to whether each pixel in each three-dimensional grid block of the target area is empty.
Specifically, the three-dimensional geographic information overall rendering module 50 selectively loads the pre-rendered ground objects in the cache according to the three-dimensional map area and the zoom level selected and observed by the user, so as to finally increase the rendering speed of the three-dimensional ground objects. Firstly, searching whether pre-rendering data exists in a cache according to a three-dimensional grid number corresponding to a region selected by a user and a zoom level, and if the pre-rendering data exists, automatically pasting the data at a corresponding position of a region to be displayed of a three-dimensional map; then, since the prerendered data in the buffer memory is only the main part of the three-dimensional ground object, the rest part needs to be complementarily rendered; by judging whether each pixel is empty in the cube space area; if the map is empty, the point of the corresponding pixel of the original map zone rendering part is directly rendered to the position, and thus the whole rendering process of the ground object main body part and the supplementing part is completed.
The invention also discloses an electronic device, comprising: at least one processor, at least one memory, a communication interface, and a bus; the processor, the memory and the communication interface complete communication with each other through the bus; the memory stores program instructions executable by the processor, and the processor invokes the program instructions to implement the steps of the three-dimensional geographic information model rendering method based on generating the countermeasure network, for example, the three-dimensional geographic information model rendering method includes a geographic information three-dimensional modeling step, a three-dimensional model generating training step, an all-terrain element matching step, a three-dimensional geographic information collaborative rendering step and a three-dimensional geographic information integral rendering step.
The invention also discloses a computer readable storage medium which stores computer instructions for causing a computer to realize the steps of the three-dimensional geographic information model rendering method based on the generation countermeasure network, such as a geographic information three-dimensional modeling step, a three-dimensional model generation training step, an all-terrain element matching step, a three-dimensional geographic information collaborative rendering step and a three-dimensional geographic information integral rendering step. The storage medium includes: a usb disk, a removable hard disk, a ROM, a RAM, a magnetic or optical disk, or other various media capable of storing program code.
Those of skill would further appreciate that the modules and algorithm steps described in connection with the embodiments disclosed herein may be embodied in electronic hardware, in computer software, or in a combination of both, and that the various example components and steps have been described generally in terms of functionality in the foregoing description to clearly illustrate this interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
In addition, it should be noted that the above-described system embodiments are merely illustrative, and do not limit the scope of the present invention, and in practical applications, one skilled in the art may select some or all modules according to actual needs to achieve the purpose of the embodiment, which is not limited herein. In addition, technical details not described in detail in this embodiment can be referred to other embodiments of the present invention, and are not described herein.
The foregoing description of the preferred embodiments of the invention is not intended to be limiting, but rather is intended to cover all modifications, equivalents, alternatives, and improvements that fall within the spirit and scope of the invention.
Claims (7)
1. A method for rendering a three-dimensional geographic information model based on generating an antagonism network, the method comprising:
a step of three-dimensional modeling of geographic information, in which typical ground object features in the three-dimensional geographic information are classified, three-dimensional ground object objects in a selected area are extracted and stored separately according to the categories; scaling each type of three-dimensional ground object to a cube space with standard size;
generating a training step of the three-dimensional model, namely respectively training a CGAN model through each type of scaled three-dimensional ground object; generating a new three-dimensional ground object with the same characteristics as the ground object of the corresponding category through the trained CGAN model;
a full feature element matching step, namely sliding and matching the cube spaces corresponding to the new three-dimensional feature objects in the original three-dimensional map to be rendered one by one, extracting the largest overlapping area successfully matched as a main body part of the similar feature, and caching the main body part; the all-terrain element matching step specifically comprises the following steps:
dividing an original three-dimensional map to be rendered into three-dimensional grid blocks, and numbering each three-dimensional grid block through a number;
sliding the cube spaces corresponding to the new three-dimensional ground object one by one in the original three-dimensional map to be rendered to perform matching degree comparison;
for the successfully matched area, the cube space corresponding to the new three-dimensional object is subjected to multi-angle rotation and multi-scale scaling, and is matched with the area in the original three-dimensional map again, and the largest overlapping area is obtained through multi-round comparison;
taking the overlapped area as a main body part of a similar object, caching the main body part, and recording the corresponding three-dimensional grid block number of the main body part in an original three-dimensional map;
a three-dimensional geographic information collaborative rendering step, namely acquiring a target area selected by a user, extracting a main body part of a three-dimensional ground object of the target area, and indexing and prerendering the main body part of the target area in a cache;
a step of integrally rendering three-dimensional geographic information, in which pre-rendering features in a cache are selectively loaded according to a target area and a zoom level selected by a user, and whether the rest part is rendered is judged according to whether pixel data of corresponding positions in three-dimensional grid blocks of the target area are empty or not; the step of integrally rendering the three-dimensional geographic information specifically comprises the following steps:
determining a corresponding three-dimensional grid block number according to a target area and a scaling level selected by a user, and searching whether corresponding prerendered data exists in a cache according to the three-dimensional grid block number;
if the pre-rendering data exist, loading the corresponding pre-rendering data to the corresponding position of the target area;
and judging whether to render the rest part according to whether the pixel data of the corresponding position in each three-dimensional grid block of the target area is empty, and if so, directly rendering the corresponding pixel point in the original three-dimensional map to the position.
2. The method for rendering a three-dimensional geographic information model based on a generated countermeasure network according to claim 1, wherein in the step of three-dimensional modeling of geographic information, scaling each type of three-dimensional ground object equally into a cube space of a standard size specifically includes:
for each three-dimensional ground object, extracting the maximum numerical value in the length, width and height coordinates of the three-dimensional ground object;
and scaling the three-dimensional ground object to be reduced into a standard cube space by taking the maximum numerical value as a reference.
3. The method for rendering the three-dimensional geographic information model based on the generation countermeasure network according to claim 1, wherein the matching degree comparison of sliding the cube spaces corresponding to the new three-dimensional ground object one by one in the original three-dimensional map to be rendered is specifically as follows:
each sliding is performed in a pixel-by-pixel sliding or self-defining sliding step length mode, after each sliding is stopped, structural similarity is adopted to evaluate the matching degree of the new three-dimensional ground object and the corresponding area in the original three-dimensional map, and if the matching degree is larger than a preset threshold value, the matching is successful.
4. The three-dimensional geographic information model rendering method based on the generation countermeasure network according to claim 1, wherein the three-dimensional geographic information collaborative rendering step specifically comprises:
acquiring a target area selected by a user, and extracting a small-scale three-dimensional ground object main body part of the target area based on a large-scale three-dimensional ground object of the target area;
and indexing and prerendering the main body part of the three-dimensional ground object with the small scale in the target area are realized in the cache.
5. A three-dimensional geographic information model rendering system based on generating a countermeasure network using the method of any of claims 1 to 4, the system comprising:
the geographic information three-dimensional modeling module is used for classifying typical ground object features in the three-dimensional geographic information, extracting three-dimensional ground object objects in a selected area and independently storing the three-dimensional ground object objects according to the types; scaling each type of three-dimensional ground object to a cube space with standard size;
the three-dimensional model generation training module is used for respectively training the CGAN model through each type of scaled three-dimensional ground object; generating a new three-dimensional ground object with the same characteristics as the ground object of the corresponding category through the trained CGAN model;
the full-feature element matching module is used for sliding and matching the cube spaces corresponding to the new three-dimensional feature objects in the original three-dimensional map to be rendered one by one, and taking the largest overlapping area successfully matched as the main body part of the similar feature and caching the main body part;
the three-dimensional geographic information collaborative rendering module is used for acquiring a target area selected by a user, extracting a main body part of the target area, and indexing and prerendering the main body part of the target area in a cache;
and the three-dimensional geographic information integral rendering module is used for selectively loading the pre-rendered ground object in the cache according to the target area and the zoom level selected by the user and judging whether to render the rest part according to whether the pixel data of the corresponding position in each three-dimensional grid block of the target area is empty or not.
6. An electronic device, comprising: at least one processor, at least one memory, a communication interface, and a bus;
the processor, the memory and the communication interface complete communication with each other through the bus;
the memory stores program instructions executable by the processor, the processor invoking the program instructions to implement the method of any of claims 1-4.
7. A computer readable storage medium storing computer instructions for causing a computer to implement the method of any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011559363.0A CN112669426B (en) | 2020-12-25 | 2020-12-25 | Three-dimensional geographic information model rendering method and system based on generation countermeasure network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011559363.0A CN112669426B (en) | 2020-12-25 | 2020-12-25 | Three-dimensional geographic information model rendering method and system based on generation countermeasure network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112669426A CN112669426A (en) | 2021-04-16 |
| CN112669426B true CN112669426B (en) | 2024-01-02 |
Family
ID=75408892
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011559363.0A Active CN112669426B (en) | 2020-12-25 | 2020-12-25 | Three-dimensional geographic information model rendering method and system based on generation countermeasure network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112669426B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115690284A (en) * | 2021-07-28 | 2023-02-03 | 华为云计算技术有限公司 | Rendering method, device and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20100136604A (en) * | 2009-06-19 | 2010-12-29 | (주)이지스 | Real-time visualization system of 3 dimension terrain image |
| KR101566167B1 (en) * | 2015-04-22 | 2015-11-05 | (주)이지스 | Method for Rendering of Object using Geomorphic data in 3D space information |
| CN112069285A (en) * | 2020-11-11 | 2020-12-11 | 湖北亿咖通科技有限公司 | Map generation method and device based on three-dimensional high-precision map slice and electronic equipment |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100520708B1 (en) * | 2003-10-20 | 2005-10-14 | 엘지전자 주식회사 | Method for displaying three dimensional map |
-
2020
- 2020-12-25 CN CN202011559363.0A patent/CN112669426B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20100136604A (en) * | 2009-06-19 | 2010-12-29 | (주)이지스 | Real-time visualization system of 3 dimension terrain image |
| KR101566167B1 (en) * | 2015-04-22 | 2015-11-05 | (주)이지스 | Method for Rendering of Object using Geomorphic data in 3D space information |
| CN112069285A (en) * | 2020-11-11 | 2020-12-11 | 湖北亿咖通科技有限公司 | Map generation method and device based on three-dimensional high-precision map slice and electronic equipment |
Non-Patent Citations (5)
| Title |
|---|
| 基于OpenGL ES的二三维地图可视化客户端设计与实现;王亚美;鲁田;;计算机应用与软件(09);全文 * |
| 基于SOA网格管理模型的空间信息共享平台设计;林秀玉;廖磊;;现代测绘(03);全文 * |
| 大规模三维地形高效可视化方法研究;刘振东;《中国优秀硕士学位论文全文数据库信息科技辑》;全文 * |
| 手机三维地图显示引擎的设计与实现;顾茂强;;电子技术与软件工程(07);全文 * |
| 网络环境下基于缓存的大规模地形渲染;曹力等;《第五届全国几何设计与计算学术会议》;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112669426A (en) | 2021-04-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111860233B (en) | SAR image complex building extraction method and system based on attention network selection | |
| CN117015806A (en) | High resolution neural rendering | |
| CN108224895B (en) | Article information input method and device based on deep learning, refrigerator and medium | |
| US20240193923A1 (en) | Method of training target object detection model, method of detecting target object, electronic device and storage medium | |
| CN105894551B (en) | Image drawing method and device | |
| CN109325996B (en) | Method and device for generating information | |
| CN104299241A (en) | Remote sensing image significance target detection method and system based on Hadoop | |
| CN116740668B (en) | Three-dimensional object detection method, three-dimensional object detection device, computer equipment and storage medium | |
| CN115457492A (en) | Target detection method and device, computer equipment and storage medium | |
| CN114387289B (en) | Semantic segmentation method and device for three-dimensional point cloud of power transmission and distribution overhead line | |
| CN115272887A (en) | Coastal zone garbage identification method, device and equipment based on unmanned aerial vehicle detection | |
| CN112883900A (en) | Method and device for bare-ground inversion of visible images of remote sensing images | |
| CN112669426B (en) | Three-dimensional geographic information model rendering method and system based on generation countermeasure network | |
| CN107038199B (en) | Drawing method and device | |
| CN114240949A (en) | Cervical cell segmentation network training method, cervical cell segmentation method and cervical cell segmentation device | |
| CN106445445B (en) | Vector data processing method and device | |
| CN116596935B (en) | Deformation detection method, deformation detection device, computer equipment and computer readable storage medium | |
| CN116704324A (en) | Target detection method, system, equipment and storage medium based on underwater image | |
| CN115953330A (en) | Texture optimization method, device, equipment and storage medium for virtual scene image | |
| CN116310308A (en) | Image segmentation method, device, computer equipment and storage medium | |
| KR101768913B1 (en) | Method of partitioning a data including geographical information, apparatus performing the same and storage medium storing a program performing the same | |
| CN114238541A (en) | Sensitive target information acquisition method and device and computer equipment | |
| CN116612474B (en) | Object detection method, device, computer equipment and computer readable storage medium | |
| CN118212430A (en) | Image feature extraction method, device, computer equipment and storage medium | |
| CN116612287B (en) | Image recognition method, device, computer equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |