CN109934911B - OpenGL-based three-dimensional modeling method for high-precision oblique photography of mobile terminal - Google Patents
OpenGL-based three-dimensional modeling method for high-precision oblique photography of mobile terminal Download PDFInfo
- Publication number
- CN109934911B CN109934911B CN201910196627.1A CN201910196627A CN109934911B CN 109934911 B CN109934911 B CN 109934911B CN 201910196627 A CN201910196627 A CN 201910196627A CN 109934911 B CN109934911 B CN 109934911B
- Authority
- CN
- China
- Prior art keywords
- model
- oblique photography
- image
- lod
- precision
- 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
Images
Abstract
The invention discloses a mobile terminal high-precision oblique photography three-dimensional modeling method based on OpenGL, belonging to the field of surveying and mapping engineering and computer combination; the method comprises the steps of firstly, rapidly obtaining image data with spatial position information in a region to be detected by using an unmanned aerial vehicle oblique photography technology, then, geometrically correcting an image, carrying out batch processing generation on oblique photography models on the corrected image data, then, carrying out LOD classification on the generated model data, then, carrying out data compression on the generated model data, merging adjacent nodes, and finally, dynamically loading a model at a mobile terminal by using an OpenGL-based technology; the invention can realize the large-batch loading of high-precision oblique photography models at the mobile terminal by combining the oblique photography technology with the mobile technology, can realize the loading and browsing of the oblique photography measurement three-dimensional digital city at the mobile terminal, and has certain social benefit and economic benefit.
Description
The technical field is as follows:
the invention relates to the technical field of surveying and mapping and digital earth model rendering, in particular to a mobile terminal high-precision oblique photography three-dimensional modeling method based on OpenGL.
Background art:
the Java programming language has a complete ecological chain, has the characteristics of safety, high efficiency and cross-platform operation, and is widely used in the field of mobile application. By utilizing the advantages of Java in the field of mobile application, the OpenGL-bound SFML library packaged by Java3d used in eclipse becomes a hot spot for constructing three-dimensional digital city models, digital earth and digital city developers. With the development of oblique photogrammetry technology, digital cities now begin to use such a realistic, highly automated model close to real ground features on a large scale. Compared with the traditional digital campus and digital city manually modeled, the digital campus and city built based on the oblique photography model can save a large amount of labor cost, the whole oblique photography process has no manual intervention, the effect is more real, the precision can reach the mapping level, the data presentation mode is full-factor presentation, and the method has the characteristics of high efficiency and short period. The oblique photography model is applied to the mobile field, can solve the problems in the industrial field range, such as city planning, urban side shape detection, major project engineering, emergency disaster relief, new rural and small town construction, smart cities and the like, and can bring important social and economic benefits into play. However, since the amount of data of the generated model is large in oblique photography, the rendering task of the model is heavier. Therefore, at present, the high-precision oblique photography model cannot be loaded to the mobile device in a large scale. At present, the solution in the industry is to cut the model in the physical range, divide a whole model into a plurality of sub-models, and then load and render the cut models in groups, and the physical range cutting of the model not only destroys the integrity of original data, but also destroys the original model textures, forms numerous broken floaters, and also influences the precision of the model, thereby causing the inaccurate matching of the model and the corresponding satellite images. In addition, the original image is not corrected and noise filtered based on the offset of the central image point in the prior industry, and only geometric correction and combined adjustment are carried out in batch generation model software, which is not enough to achieve the surveying and mapping precision. Therefore, how to improve the rendering/loading strategy/method of the large-batch high-precision oblique photography model with vivid texture, so that the loading of the large-scale high-precision oblique photography model by the mobile equipment can be met, and the requirement of browsing the oblique photography model in the mobile field is met, which becomes a difficult problem to be solved urgently.
The invention content is as follows:
the invention aims to overcome the defects of the prior art and provide a moving end high-precision oblique photography three-dimensional modeling method based on OpenGL; the problem that an existing high-precision oblique photography model cannot be loaded to a mobile device in a large scale is mainly solved.
The technical scheme of the invention is as follows: the OpenGL-based three-dimensional modeling method for high-precision oblique photography at a mobile terminal is characterized by comprising the following steps of:
a, correcting 60% overlap ratio image data of a region to be detected, which is acquired by an unmanned aerial vehicle carrying a five-lens camera, based on an image center image point according to formulas 1-8, and performing Kalman filtering algorithm noise filtering processing on the corrected image to obtain a basic image;
formula 1: a = mx · cos t
Formula 2: b = my (k. Cos t-sin t)
Formula 3: d = mx · sin t
Formula 4: e = -1. My. (k. Sin t + cos t)
Formula 5: translation in C = x direction
Formula 6: translation in F = y direction
Formula 7: x1 = Ax + By + C
Formula 8: y1 = Dx + Ey + F
The method comprises the following steps that A is a scale factor of X, B and D are rotation terms, C and F are translation terms, E is a negative value of a scale factor of y, mx is a scale change in the X direction, my is a scale change in the y direction, t depends on a yaw angle, a value measured anticlockwise by taking the X axis as a starting point, and as an airplane coordinate system and a mathematical plane coordinate system are defined differently, a coordinate inverse calculation is needed to solve a correct yaw angle for a two-dimensional plane, wherein t is the yaw angle subjected to coordinate inverse calculation; k is the shear factor along the x-axis = tan (u), u is the difference between the yaw angle and the y-axis, when the tilt angle is measured relative to the y-axis;
b, using image processing software MATLAB in combination with a Kalman filtering algorithm to filter the influence of noise on the image quality; the method for removing the noise factors by Kalman filtering comprises the following steps:
1) Firstly, determining a noise state equation and an observation equation;
2) Importing external orientation elements of the image data;
3) Inputting various parameters: latitude, longitude, process noise variance, observation noise variance, initial value of filter vector and observation state variance at any moment;
4) Calculating the coefficient of a prediction equation, and finally realizing filtering through a difference equation;
c, importing the basic image into PC (personal computer) end automatic modeling software Smart 3D, generating an ultrahigh-density point cloud based on the real image through further geometric correction and joint adjustment processing procedures of the corrected basic image, and generating a high-resolution real-scene three-dimensional model based on the texture of the real image;
d, after the image correction is finished, manufacturing a high-precision oblique photography model, grading the generated oblique photography model data according to LOD (level of distribution), wherein the LOD grading can be used for expressing the texture fineness of the generated oblique photography model, dividing the model into pyramid grades by the LOD grading, and each grade corresponds to the precision grade of the oblique photography model;
e, fragmenting the classified oblique photography three-dimensional models according to the classification level, wherein each fragmentation model corresponds to a root node; each fragmentation model is provided with a plurality of child nodes; dividing each fragment model into a plurality of oblique photography models with small block quantity, wherein each fragment model corresponds to a sub-node; storing the number of the fragments, the root nodes of the fragment model and the corresponding longitude and latitude coordinates, the sub nodes of the block model and the longitude and latitude coordinates which are well divided according to the LOD grade; loading and grouping the sliced model, wherein the grouping step comprises the following steps:
1) Firstly, determining a whole measuring region model, and dividing the model grade according to an LOD technology;
2) Determining the number of the model fragments under each grade, wherein the number of the fragments under each grade is determined according to the LOD grading;
3) Then determining the area range of the number of the fragmentation models respectively displayed under the grade, and determining the central node corresponding to each fragmentation model and the corresponding longitude and latitude;
4) Finally, sub-nodes and corresponding longitudes and latitudes of a plurality of module sub-models under the fragmentation model are determined;
f, performing data compression on the high-precision oblique photography model completing the model classification fragmentation, reducing the data volume of the model, understanding that the node information of the current oblique photography model is of a tree structure, the whole classification model has a central node, each corresponding fragmentation model has a root node, a plurality of block models under each fragmentation model have corresponding child nodes, and realizing the quick browsing and loading of the mobile terminal by modifying the central node of the whole model, combining adjacent nodes, thinning out nodes and texture compression, wherein the data compression method comprises the following steps:
1) Modifying the central point of the whole oblique photography model;
2) Merging adjacent nodes;
3) A rarefaction node;
4) Texture compression;
g, texture construction and rendering are carried out on the generated high-precision oblique photography model by using an OpenGL open graphic library interface, the original data format of oblique photography is converted, loading and browsing of the high-precision oblique photography model by a mobile terminal are realized, and the method for constructing and rendering the model comprises the following steps:
1) Constructing and rendering a tilted photography model which is classified based on LOD grade in an OpenGL graphic library through a program interface by using an application program programming interface provided by an open graphic library;
2) Performing data conversion on a large-batch high-precision oblique photography model built by rendering, converting an original oblique photography data format (OBJ, OSGB and DAE) into an SFML format bound by an OpenGL graphic library to be converted into a mobile equipment format, uploading the data built by rendering to a server or configuring the data to a project local, reading an index file through a mobile equipment application program, and completing loading and browsing of mobile equipment;
3) Then carrying exterior orientation elements of a camera on the unmanned aerial vehicle, wherein the exterior orientation elements comprise a central point of the whole model, the attitude angle of the aircraft corresponds to degrees, the attitude angle of the aircraft comprises a pitch angle, a course angle and a roll angle, and an oblique photography model of the model needing to be loaded and based on LOD grading is determined in the camera lens range;
4) When the oblique photography model of the mobile equipment end is subjected to moving zooming operation, whether the level of the current model is the maximum model level or not is judged, the model loading capacity exceeds 60%, then the level of the model to be loaded is calculated, if the moving end is subjected to the operation of enlarging the oblique photography model, a new level model is loaded, the model loaded before is dynamically deleted, and if the moving end is subjected to the operation of reducing the oblique photography model, the loaded oblique photography model is not deleted, and finally the dynamic loading of a large batch of high-precision oblique photography models is realized.
Furthermore, the step f is implemented by data compression and node combination tools which are three-dimensional processing software; and recording the texture compression of the data and the information of the merging node in a config.
Further, in the step d, an LOD-based classification strategy is formulated by the import information, and the following conditions are satisfied:
1) The number of the slice models under the corresponding LOD grade is not more than 15;
2) Each slice model is adjacent to form a complete oblique photography model;
3) The number of the fragment models corresponding to the models with different LOD grades is obviously different;
4) The number of model slices with the lowest LOD level is the least.
Compared with the prior art, the OpenGL-based three-dimensional modeling method for high-precision oblique photography of the mobile terminal has the following prominent substantive characteristics and remarkable progress:
1. because each image data of the region to be detected is provided with an external orientation element, based on the combination of the external orientation element and an affine transformation algorithm, the offset of a central image point relative to an actual central image point can be solved, so that the influence of factors such as airplane posture and terrain fluctuation is eliminated, the precision of the oblique photography model is improved, the precision of the oblique photography model is approximately three times as high as the resolution of an orthographic image of an engineering, and compared with the traditional method of directly generating the oblique photography model without image correction, the precision of the model is improved by 20-30%; the difference of the formula (1-8) algorithm lies in that the calculated offset (relative to the route track) of the image center image point is used for correcting the image center image point, the center image point of each image corresponds to the center node of each block model, the node records longitude and latitude and elevation information, and the model precision can be integrally improved by using the data corrected based on the image center image point to make an oblique photography model;
2. because the basic image is corrected by the central point and has noise influence, the noise can cause the image to have periodic stripes, spots and the like, the corrected image data is combined with the longitude and latitude of the central point and the elements in the outer orientation to remove the noise by using a Kalman filter method, compared with the processing method of the traditional method, the filtering processing method combined with the elements in the outer orientation is improved in that the basic image noise is restrained and filtered by combining the longitude and latitude, the attitude angle and the flying height of the central point of the basic image, and the attitude angle of the airplane comprises a pitch angle, a course angle and a roll angle; the traditional remote sensing image noise filtering adopts high-pass or low-pass filtering to realize data correction, but the high-pass or low-pass filtering is carried out on basic image data required by an oblique photography model to realize a noise filtering effect is not obvious, and factors such as the spatial position and the attitude of image shooting are fully considered by adopting an external orientation element and a Kalman filter, so that the image result of noise on the basic image can be reduced to the maximum extent; compared with a model without noise filtering, the model with noise filtered has better texture fitting effect;
3. compared with a traditional region cutting model to divide the model method, the cutting model not only damages the data integrity to cause the loss of original data, is easy to form suspended matters, and has low model precision; the method has the improvement that the oblique photography model is classified according to the LOD model level according to the characteristics of the original model data formats (OBJ, OSGB and DAE), and is classified, fragmented and blocked according to the corresponding node information, so that the complete original model data information is stored under the condition of not damaging the original data structure, the model precision can reach 0.3-0.5CM, and the precision requirement of the surveying and mapping industry is met.
Description of the drawings:
FIG. 1 is a schematic diagram of an image rectification method according to the present invention;
FIG. 2 is a flow chart of the model hierarchical compression method of the present invention;
FIG. 3 is a flow chart of an OpenGL-based dynamic loading oblique photography model in the present invention;
fig. 4 is an overall flow diagram of the present invention.
The specific implementation mode is as follows:
in order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to be limiting.
Example 1, see fig. 1, 2, 3, 4; the image is corrected by the center image point, and the noise is removed by filtering, and the image correction method in the invention is shown in figure 1:
firstly, acquiring 60% overlapping rate image data of an area to be detected by using an unmanned aerial vehicle, carrying a five-lens camera by using the unmanned aerial vehicle, and performing image correction and image transformation on the acquired image data due to the influences of airplane postures, topographic relief and noise; each piece of acquired image data is provided with a central point coordinate, and the longitude and latitude of the unmanned aerial vehicle when the unmanned aerial vehicle shoots the area are recorded;
calculating the offset of the central point of the image by using the exterior orientation element of each image data and combining an affine transformation formula, and calculating the correct central point according to the offset of the central point; improving the precision of the oblique photography model, wherein the precision of the oblique photography model is approximately equal to three times of the resolution of the engineering orthographic image; the difference of the algorithm lies in that the calculated offset (relative to the route track) of the image center image point is used for correcting the image center image point, the center image point of each image corresponds to the center node of each block model, the node records longitude and latitude and elevation information, and the accuracy of the model can be integrally improved by using the data corrected based on the image center image point to make an oblique photography model;
the formula for calculating the offset of the center point of the image is as follows:
A = mx · cos t
B = my · (k · cos t - sin t)
D = mx · sin t
E = -1 · my · (k · sin t + cos t)
translation in C = x direction
Translation in F = y direction
x1 = Ax + By + C
y1 = Dx + Ey + F
Where A is the scale factor for X, B and D are the rotation terms, C and F are the translation terms, E is the negative of the scale factor for y, mx is the change in scale in the X direction, my is the change in scale in the y direction, t depends on the yaw angle, and is measured counterclockwise from the X axis as the starting point. Because the definition of a coordinate system of the body is different from that of a coordinate system of a mathematical plane, the correct yaw angle of the two-dimensional plane needs to be solved by coordinate back calculation, wherein t is the yaw angle after coordinate back calculation; k is the shear factor = tan (u) along the x-axis, and u is the difference between the yaw angle and the y-axis, when the tilt angle is measured relative to the y-axis; through the correction based on the image central point, the precision of the image data is obviously improved, and the manufacture of a high-precision oblique photography model has important influence on the surveying and mapping industry;
the image corrected by the central point still has noise influence, the noise can cause the image to have periodic stripes, spots and the like, the corrected image data is combined with the longitude and latitude and the exterior orientation element of the central point to remove the noise by using a Kalman filter method, and image processing software MATLAB is combined with a Kalman filter algorithm to filter the influence of the noise on the image quality, so that the interference of the noise on the image quality is filtered;
the Kalman filtering noise factor removing method comprises the following steps:
1) Firstly, determining a state equation and an observation equation;
2) Importing external orientation elements of the image data;
3) Inputting various parameters: latitude, longitude, process noise variance, observation noise variance, initial value of filter vector and observation state variance at any moment;
4) Calculating the coefficient of a prediction equation, and finally realizing filtering through a difference equation;
introducing high-precision images which are subjected to geometric correction and noise filtering into PC-side automatic modeling software in batches, carrying out further geometric correction, joint adjustment and other processing procedures by the automatic modeling software, completing aerial triangulation calculation, generating ultra-high density point cloud based on a real image, and generating a high-resolution live-action three-dimensional model based on real image texture;
the generated oblique photography model data is derived in a general format such as (OBJ, OSGB, DAE). The models are then graded. The LOD grading can be used for representing the texture fineness for generating the oblique photography model, the model is divided into pyramid grades by the LOD grading, and each grade corresponds to the precision grade of the oblique photography model;
the step of model hierarchical compression, the method for realizing the model hierarchical compression in the invention is shown in figure 2:
slicing the graded oblique photography three-dimensional models according to the grading, wherein each sliced model corresponds to a root node; each fragmentation model is provided with a plurality of child nodes; dividing each fragment model into a plurality of oblique photography models with smaller block quantity, wherein each fragment model corresponds to a sub-node; and storing the number of the fragments, the root nodes of the fragment model, the corresponding longitude and latitude coordinates, the sub nodes of the block model and the corresponding longitude and latitude coordinates which are well divided according to the LOD grade. Loading and grouping the sliced models, wherein the grouping step comprises the following steps:
firstly, determining a whole measuring region model, and dividing the model grade according to the LOD technical grade;
determining the number of the model fragments under each grade, wherein the number of the fragments under each grade is determined according to the LOD grade;
then determining the area range of the number of the fragmentation models respectively displayed under the grade, and determining the central node corresponding to each fragmentation model and the corresponding longitude and latitude;
finally, sub-nodes and corresponding longitudes and latitudes of a plurality of module sub-models under the fragmentation model are determined;
the information formulation is based on LOD grading strategy, and the main following principles are as follows: 1) The number of the slice models under the corresponding LOD levels is not more than 15; 2) Each slice model is adjacent to form a complete oblique photography model; 3) The number of the fragment models corresponding to the models with different LOD grades is obviously different; 4) The number of the model fragments with the lowest LOD grade division is least;
storing the information after the model is classified and fragmented into an xml file, wherein the file stores information of each node of the inclined model, and the file is an index file and is used for acquiring the absolute position of each model;
the directory structure of the XML file is:
{FileName:/Tile_+000_+000.osgb:centerX,centerY,centerZ:Radius},
the relative path of the model file relative to the central node is only loaded with one osgb file of the master control under the same directory;
wherein the Tile folder stores oblique photography data, X, Y, Z correspond to node longitude and latitude and height information,
radius corresponds to the range distance, and each level of information is stored in an XML index file by the information;
importing three-dimensional processing software into an xml file which completes the hierarchical fragmentation of the model, performing data compression, and reducing the data volume of the model, wherein the node information of the current oblique photography model is understood to be a tree structure, the whole hierarchical model has a central node, each corresponding fragmentation model has a root node, a plurality of block models under each fragmentation model have corresponding child nodes, and the node information can be divided into a quadtree, an octree and any according to the specific project requirements for data compression by modifying the central node of the whole model, combining adjacent nodes, thinning out nodes and texture compression. The method for the hierarchical compression comprises the following steps:
1) Modifying the central point of the whole oblique photography model;
2) Merging adjacent nodes;
3) A rarefaction node;
4) Texture compression;
after data compression is completed, the number of folders of the model is combined and thinned, a configuration file config.scp is generated, data compression under different LOD levels is completed respectively according to the configuration file config.scp, the LOD levels divide the model into pyramid levels, and each level corresponds to the precision level of the oblique photography model;
the dynamic loading step of the model, the dynamic loading method of the batch high-precision oblique photography model based on OpenGL in the invention is shown in figure 3:
5) The method for constructing and rendering the high-precision oblique photography model by using the OpenGL open graphics library in a large batch mode, converting the original data format of oblique photography, loading and browsing the high-precision oblique photography model by a mobile terminal, and constructing and rendering the model comprises the following steps:
5.1 Applying Java programming language, establishing and rendering a model by using an eclipse development platform and combining an application programming interface provided by an OpenGL open graphics library;
5.2 Carrying out data conversion on the large-batch high-precision oblique photography model built by rendering, and converting an original oblique photography data format (OBJ, OSGB and DAE) into a format built by binding SFML rendering by an OpenGL graphic library. Uploading the rendered and constructed data to a server or configuring the data to a project local, and completing reading of an index file through a mobile equipment application program to complete loading and browsing of mobile equipment;
5.3 Then carrying outer orientation elements of the camera on the unmanned aerial vehicle, including the central point of the whole model and the corresponding degrees of the attitude angle of the airplane; the attitude angle of the airplane comprises a pitch angle, a course angle and a roll angle, and then a tilted photography model which is determined to be loaded in the range of a camera lens and is based on LOD grading is calculated;
6) When the oblique photography model on the mobile device is subjected to the moving zooming operation, whether the level of the current model is the maximum model level or not is judged, the model loading amount exceeds 60%, then the level of the model to be loaded is calculated, if the moving end is subjected to the operation of enlarging the oblique photography model, a new level model is loaded, the model loaded before is dynamically deleted, and if the moving end is subjected to the operation of reducing the oblique photography model, the loaded oblique photography model is not deleted. And finally, dynamically loading a large batch of high-precision oblique photography models.
The method accurately calculates the offset of the central image point by combining the exterior orientation element of the airplane with the affine transformation projection algorithm, realizes geometric correction on the image, and filters out the influence of the strip and the spot on the image caused by noise by using a Kalman filter; and finally, displaying a large quantity of high-precision oblique photography three-dimensional models in a high-precision physical model form in the surveying and mapping engineering field through a hierarchical dynamic loading strategy, and quickly browsing in an APP form in the mobile field.
It will be understood that modifications and variations may be resorted to as will be apparent to those skilled in the art, and that all such modifications and variations are intended to fall within the scope of the appended claims.
Claims (3)
1. The OpenGL-based three-dimensional modeling method for high-precision oblique photography of the mobile terminal is characterized by comprising the following steps of:
a. correcting 60% overlap ratio image data of a region to be detected, which is acquired by carrying a five-lens camera on an unmanned aerial vehicle, based on an image center image point according to formulas 1-8, and performing Kalman filtering algorithm noise filtering processing on the corrected image to obtain a basic image;
formula 1: a = mx · cos t
Formula 2: b = my (k. Cos t-sin t)
Formula 3: d = mx · sin t
Formula 4: e = -1. My. (k. Sin t + cos t)
Formula 5: translation in C = x direction
Formula 6: translation in F = y direction
Formula 7: x1 = Ax + By + C
Formula 8: y1 = Dx + Ey + F
The method comprises the following steps that A is a scale factor of X, B and D are rotation terms, C and F are translation terms, E is a negative value of a scale factor of y, mx is a scale change in the X direction, my is a scale change in the y direction, t depends on a yaw angle, a value measured anticlockwise by taking the X axis as a starting point, and as an airplane coordinate system and a mathematical plane coordinate system are defined differently, a coordinate inverse calculation is needed to solve a correct yaw angle for a two-dimensional plane, wherein t is the yaw angle subjected to coordinate inverse calculation; k is the shear factor along the x-axis = tan (u), u is the angle of inclination of the yaw to the y-axis, when the angle of inclination is measured relative to the y-axis;
b. filtering the influence of noise on image quality by using image processing software MATLAB in combination with a Kalman filtering algorithm; the method for removing the noise factors by Kalman filtering comprises the following steps:
1) Firstly, determining a noise state equation and an observation equation;
2) Importing external orientation elements of the image data;
3) Inputting various parameters: latitude, longitude, process noise variance, observation noise variance, initial value of filter vector and observation state variance at any moment;
4) Calculating the coefficient of a prediction equation, and finally realizing filtering through a difference equation;
c. importing the basic image into PC-side automatic modeling software Smart 3D, further performing geometric correction and joint adjustment processing on the corrected basic image to generate an ultrahigh-density point cloud based on the real image, and generating a high-resolution real-scene three-dimensional model based on the texture of the real image;
d. the method comprises the steps of manufacturing a high-precision oblique photography model after image correction is finished, grading generated oblique photography model data according to LOD (level of distribution), wherein the LOD grading can be used for expressing texture fineness of the generated oblique photography model, dividing the model into pyramid grades by the LOD grading, and each grade corresponds to the precision grade of the oblique photography model;
e. slicing the graded oblique photography three-dimensional model according to the division level, wherein each slice model corresponds to a root node; each fragmentation model is provided with a plurality of child nodes; dividing each fragment model into a plurality of oblique photography models with smaller block quantity, wherein each fragment model corresponds to a sub-node; storing the number of the fragments, the root nodes of the fragment model and the corresponding longitude and latitude coordinates, the sub nodes of the block model and the longitude and latitude coordinates which are well divided according to the LOD grade; loading and grouping the sliced models, wherein the grouping step comprises the following steps:
1) Firstly, determining a whole measuring region model, and dividing the model grade according to an LOD technology;
2) Determining the number of the model fragments under each grade, wherein the number of the fragments under each grade is determined according to the LOD grade;
3) Then determining the area range of the number of the fragmentation models respectively displayed under the grade, and determining the central node corresponding to each fragmentation model and the corresponding longitude and latitude;
4) Finally, sub-nodes and corresponding longitudes and latitudes of a plurality of module sub-models under the fragmentation model are determined;
f. the method comprises the following steps of carrying out data compression on a high-precision oblique photography model for completing model grading fragmentation, reducing the data volume of the model, understanding that the node information of the current oblique photography model is of a tree structure, the whole grading model has a central node, each corresponding fragmentation model has a root node, a plurality of block models under each fragmentation model have corresponding child nodes, and realizing quick browsing and loading of a mobile terminal by modifying the central node of the whole model, combining adjacent nodes, thinning out nodes and texture compression, wherein the data compression method comprises the following steps:
1) Modifying the central point of the whole oblique photography model;
2) Merging adjacent nodes;
3) A rarefaction node;
4) Texture compression;
g. the method comprises the following steps of using an OpenGL open graphics library interface to construct and render the texture of the generated high-precision oblique photography model, converting the original data format of oblique photography, realizing loading and browsing of the high-precision oblique photography model by a mobile terminal, and constructing and rendering the model:
1) Constructing and rendering a tilted photography model which is classified based on LOD grade in an OpenGL graphic library through a program interface by using an application program programming interface provided by an open graphic library;
2) Performing data conversion on a large-batch high-precision oblique photography model built by rendering, converting the original OBJ, OSGB and DAE data formats of oblique photography into formats of mobile equipment by binding SFML (open graphics library) to an OpenGL (open graphics library), uploading the data built by rendering to a server or configuring the data to the local engineering, completing the reading of an index file through a mobile equipment application program, and completing the loading and browsing of the mobile equipment;
3) Then carrying exterior orientation elements of a camera on the unmanned aerial vehicle, including the central point of the whole model, and the corresponding degrees of attitude angles of the aircraft, wherein the attitude angles of the aircraft include a pitch angle, a course angle and a roll angle, and calculating an oblique photography model of the model to be loaded, which is determined in the lens range of the camera and is based on LOD grading;
4) When the oblique photography model of the mobile equipment end is subjected to moving zooming operation, whether the level of the current model is the maximum model level or not is judged, the model loading capacity exceeds 60%, then the level of the model to be loaded is calculated, if the moving end is subjected to the operation of enlarging the oblique photography model, a new level model is loaded, the model loaded before is dynamically deleted, and if the moving end is subjected to the operation of reducing the oblique photography model, the loaded oblique photography model is not deleted, and finally the dynamic loading of a large batch of high-precision oblique photography models is realized.
2. The OpenGL-based mobile terminal high-precision oblique photography three-dimensional modeling method according to claim 1, wherein the step f performs data compression and merges node tools into three-dimensional processing software; and recording the data texture compression and the information of the merging node in a config.
3. The OpenGL-based three-dimensional modeling method for high-precision oblique photography at a mobile terminal of a mobile terminal, as claimed in claim 1, wherein an LOD-based hierarchical strategy is formulated by the imported information in the step d, and the following conditions are satisfied:
1) The number of the slice models under the corresponding LOD grade is not more than 15;
2) Each slice model is adjacent to form a complete oblique photography model;
3) The number of the slice models corresponding to the models with different LOD levels is obviously different;
4) The number of model slices with the lowest LOD level division is the least.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910196627.1A CN109934911B (en) | 2019-03-15 | 2019-03-15 | OpenGL-based three-dimensional modeling method for high-precision oblique photography of mobile terminal |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910196627.1A CN109934911B (en) | 2019-03-15 | 2019-03-15 | OpenGL-based three-dimensional modeling method for high-precision oblique photography of mobile terminal |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109934911A CN109934911A (en) | 2019-06-25 |
| CN109934911B true CN109934911B (en) | 2022-12-13 |
Family
ID=66987188
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910196627.1A Active CN109934911B (en) | 2019-03-15 | 2019-03-15 | OpenGL-based three-dimensional modeling method for high-precision oblique photography of mobile terminal |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109934911B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110634184B (en) * | 2019-09-11 | 2023-01-17 | 西安恒歌数码科技有限责任公司 | Loading method of mass oblique photography data |
| CN110796152A (en) * | 2020-01-06 | 2020-02-14 | 杭州鲁尔物联科技有限公司 | Group building earthquake damage extraction method and system based on oblique photography |
| CN111325848B (en) * | 2020-03-02 | 2022-04-15 | 中国电建集团昆明勘测设计研究院有限公司 | Automatic batch deletion method for suspended matters of OSGB inclined model |
| CN111583404B (en) * | 2020-05-08 | 2021-08-31 | 广西壮族自治区自然资源遥感院 | Massive oblique photography three-dimensional model data scheduling method and system |
| CN112365598B (en) * | 2020-10-29 | 2022-09-20 | 深圳大学 | Method, device and terminal for converting oblique photography data into three-dimensional data |
| CN113868349B (en) * | 2021-08-31 | 2023-01-13 | 广东省测绘工程有限公司 | Rural house-ground integrated investigation and library building method based on hybrid network |
Family Cites Families (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002074323A (en) * | 2000-09-01 | 2002-03-15 | Kokusai Kogyo Co Ltd | Method and system for generating three-dimensional urban area space model |
| JPWO2008152740A1 (en) * | 2007-06-13 | 2010-08-26 | 株式会社情報科学テクノシステム | Digital aerial 3D measurement system |
| WO2012132237A1 (en) * | 2011-03-31 | 2012-10-04 | パナソニック株式会社 | Image drawing device for drawing stereoscopic image, image drawing method, and image drawing program |
| US9947128B2 (en) * | 2013-01-29 | 2018-04-17 | Andrew Robert Korb | Methods for improving accuracy, analyzing change detection, and performing data compression for multiple images |
| CN103617650A (en) * | 2013-11-29 | 2014-03-05 | 哈尔滨工程大学 | Displaying method for complex three-dimensional terrain |
| CN104766360B (en) * | 2015-03-09 | 2017-12-22 | 腾讯科技（深圳）有限公司 | The rendering intent and device of a kind of three-dimensional space data |
| CN106327573B (en) * | 2016-08-25 | 2019-03-12 | 成都慧途科技有限公司 | A kind of outdoor scene three-dimensional modeling method for urban architecture |
| CN106373199B (en) * | 2016-08-31 | 2019-05-14 | 中测新图(北京)遥感技术有限责任公司 | A kind of oblique photograph building model rapid extracting method |
| CN107590854A (en) * | 2017-09-15 | 2018-01-16 | 福建四创软件有限公司 | Reservoir region three-dimensional live methods of exhibiting based on WEBGIS |
| CN107862744B (en) * | 2017-09-28 | 2021-05-18 | 深圳万图科技有限公司 | Three-dimensional modeling method for aerial image and related product |
| CN108090956B (en) * | 2017-11-10 | 2021-12-21 | 中国科学院空间应用工程与技术中心 | Geometric correction method and system |
| CN108171787A (en) * | 2017-12-18 | 2018-06-15 | 桂林电子科技大学 | A kind of three-dimensional rebuilding method based on the detection of ORB features |
| CN108665536B (en) * | 2018-05-14 | 2021-07-09 | 广州市城市规划勘测设计研究院 | Three-dimensional and live-action data visualization method and device and computer readable storage medium |
| CN109186551B (en) * | 2018-08-08 | 2021-02-12 | 广州市城市规划勘测设计研究院 | Method and device for extracting characteristic points of oblique photogrammetry building and storage medium |
| CN109238239B (en) * | 2018-09-12 | 2021-04-06 | 成都坤舆空间科技有限公司 | Digital measurement three-dimensional modeling method based on aerial photography |
-
2019
- 2019-03-15 CN CN201910196627.1A patent/CN109934911B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN109934911A (en) | 2019-06-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109934911B (en) | OpenGL-based three-dimensional modeling method for high-precision oblique photography of mobile terminal | |
| CN102506824B (en) | Method for generating digital orthophoto map (DOM) by urban low altitude unmanned aerial vehicle | |
| Lo Brutto et al. | UAV platforms for cultural heritage survey: first results | |
| CN113192193B (en) | High-voltage transmission line corridor three-dimensional reconstruction method based on Cesium three-dimensional earth frame | |
| CN109242862B (en) | Real-time digital surface model generation method | |
| CN113607135B (en) | Unmanned aerial vehicle inclination photogrammetry method for road and bridge construction field | |
| CN105427377A (en) | Unmanned plane based geological disaster data collection processing method and device | |
| JP2003288613A (en) | System, method, and computer program for generating 3-d model | |
| CN104809759A (en) | Large-area unstructured three-dimensional scene modeling method based on small unmanned helicopter | |
| US11087532B2 (en) | Ortho-image mosaic production system | |
| KR101405891B1 (en) | Reality display system of air inteligence and method thereof | |
| CN115393240A (en) | GIS and BIM fusion method and device, electronic equipment and readable storage medium | |
| CN111667569A (en) | Three-dimensional real-scene earthwork visual accurate measuring and calculating method based on Rhino and Grasshopper | |
| KR100732915B1 (en) | Method for three-dimensional determining of basic design road route using digital photommetry and satellite image | |
| CN114140593B (en) | Digital earth and panorama fusion display method and device | |
| KR20060127519A (en) | Spatial information structure method based image and system thereof | |
| CN115797256A (en) | Unmanned aerial vehicle-based tunnel rock mass structural plane information processing method and device | |
| KR102587445B1 (en) | 3d mapping method with time series information using drone | |
| Pacina et al. | Pre‐dam valley reconstruction based on archival spatial data sources: Methods, accuracy, and 3D printing possibilities | |
| WO2022104251A1 (en) | Image analysis for aerial images | |
| Crues et al. | Digital Lunar Exploration Sites (DLES) | |
| JP7122901B2 (en) | breakline extraction program, breakline extraction system | |
| CN114119892B (en) | Three-dimensional digital road network construction method based on BIM and GIS technologies | |
| KR102538157B1 (en) | Method for producing 3 dimension reality model using unmanned aerial vehicle | |
| Picu | Updating geospatial data by creating a high resolution digital surface model |
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 |