CN104616345A - Octree forest compression based three-dimensional voxel access method - Google Patents
Octree forest compression based three-dimensional voxel access method Download PDFInfo
- Publication number
- CN104616345A CN104616345A CN201410765462.2A CN201410765462A CN104616345A CN 104616345 A CN104616345 A CN 104616345A CN 201410765462 A CN201410765462 A CN 201410765462A CN 104616345 A CN104616345 A CN 104616345A
- Authority
- CN
- China
- Prior art keywords
- voxel
- block
- octree
- data
- dimensional
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
- G06T17/005—Tree description, e.g. octree, quadtree
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Computer Graphics (AREA)
- Geometry (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Generation (AREA)
- Processing Or Creating Images (AREA)
Abstract
The invention discloses an octree forest compression based three-dimensional voxel access method. The method comprises the steps of (1) uniformly dividing a voxel space according to the thickness, and creating an octree in each area; (2) performing dynamic construction and growth update for an octree forest according to each frame depth map and camera parameters; (3) quickly searching leaf data by the layered search method according to the constructed forest structure. According to the method, the octree forest sturctuer is utilized to effectively compress the consumption of a storage resource of space voxel data, and meanwhile, the data searching speed and the surface reconstruction precision can be ensured.
Description
Technical field
The invention belongs to three-dimensional reconstruction field, be specifically related to a kind of three-dimensional voxel access method based on the compression of Octree forest.
Background technology
Three-dimensional reconstruction is that one utilizes two-dimensional projection to recover the computer technology of object dimensional information (shape etc.).Traditional three-dimensional surface rebuilding utilizes video camera to take object in different visual angles, the angle point in abstract image, edge, texture, lines, and the essential characteristics such as border recover object general profile according to visual angle relation.In recent years along with the development of depth measuring devices, people start by laser measuring apparatus, and ToF (the time m-light) equipment such as depth camera and structured light depth camera carries out more accurately denser resurfacing.Along with the develop rapidly of computer software and hardware, drafting real-time that is extensive, high-precision three-dimensional scene improves constantly, and the difficulty of three-dimensional reconstruction significantly reduces, and precision is higher, and speed is faster.Three-dimensional reconstruction has a wide range of applications in fields such as reverse-engineering, video display amusement, industrial design and historical relic's protections.In satellite exploration landforms field, utilize the great amount of images information taken by satellite, three-dimensional reconstruction is carried out to geology surface, thus obtain looks information more, and then related science research is offered help; In medical domain, by the surface model of three-dimensional reconstruction human internal organs, doctor can be helped more effectively to carry out illness analysis and diagnosis; In historical relic's protection field, three-dimensional reconstruction is carried out to the historical relic of laying special stress on protecting, preserve the information data of historical relic in time, make heritage information have diversity and continuity.
In the end of the year 2011, Microsoft Research will propose with the depth camera of consumer level (Kinect) be collecting device real-time dense resurfacing and tracing algorithm (KinectFusion).The core of KinectFusion algorithm blocks distance function value (TSDF), first it define an individuality in actual physics space, then this individuality is divided into many voxels of uniform size, wherein each voxel stores a TSDF value to represent that this voxel is to distance surperficial recently, namely the voxel that distance solid object surface is nearer, its TSDF value is less.
Comparatively large and the impartial division methods adopted makes the voxel data in KinectFusion algorithm need to take larger storage space due to number of voxel, which limit its application under different platform.According to analysis, only have the number of regions of close body surface to be effective in TSDF body, and the ratio shared by surface in physical space is very little, therefore TSDF volume data has a large amount of redundant informations, compression method can be utilized to compress it, reduce taking of storage resources.
Summary of the invention
For the above-mentioned technical matters existing for prior art, the invention provides a kind of three-dimensional voxel access method based on the compression of Octree forest, can under the prerequisite ensureing former algorithm reconstruction accuracy, the effective occupancy reducing storage resources, realizes the efficient access of TSDF data.
Based on a three-dimensional voxel access method for Octree forest compression, as follows:
Storage is operated: first, rough even partition is carried out to three-dimensional space, obtain the block of multiple identical square; Then, in described block, build Octree, namely utilize depth camera back projection dynamic tree correspondence carries out block segmentation from root node; Finally, TSDF data are stored in the voxel corresponding to Octree bottom leaf node;
For read operation: first, according to the coordinate waiting to fetch data in three-dimensional space, determine the block belonging to it; Then, adopt the searching method of depth-first to search for this block Octree, to search the bottom leaf node matched with coordinate to be fetched data in tree; Finally, from voxel corresponding to the leaf node searched, corresponding TSDF data are read.
The concrete grammar of described structure Octree is: using corresponding for the block root node as Octree, if by judging to divide block, then block is evenly divided into the voxel of eight identical squares, eight child nodes under eight corresponding root nodes of voxel difference; In like manner, if by judging to divide voxel, then further voxel is evenly divided into eight little voxels, eight little voxels, eight child nodes respectively under corresponding large voxel node; When the Octree number of plies reaches the default maximum number of plies, even if voxel meets the requirement of Further Division, also it is not split again.
Judge that the conditioning process whether block or voxel need to carry out dividing is as follows:
First, by coordinate transform, the central point C of block or voxel is mapped in the projection plane of depth camera, obtains corresponding depth image vegetarian refreshments U;
Then, depth image vegetarian refreshments U is returned in three-dimensional space according to its depth value back projection, obtain the subpoint P of central point C in space on body surface;
Finally, need be divided block or voxel by following judging whether according to the three-dimensional coordinate of central point C and subpoint P:
If block or voxel do not exist with body surface in three-dimensional space intersection, then meet following relation and then block or voxel divided:
||p-c||≤T
If block or voxel exist with body surface in three-dimensional space intersection, then meet following relation and then block or voxel divided:
Wherein: p and c is respectively subpoint P and the coordinate vector of central point C in three-dimensional space, and l is the length of side of block or voxel, T and ε is the intercept parameter of setting.
If block does not make Further Division, then the Octree of its correspondence is empty tree, does not have any TSDF data in this block.
The expression formula of described coordinate vector p is as follows:
p＝R·π(K,u,Z(u))+t
Wherein: u is the coordinate vector of depth image vegetarian refreshments U in three-dimensional space, K is the internal reference matrix of depth camera, Z (u) is the depth value of depth image vegetarian refreshments U, R and t is respectively rotation matrix and the translation vector of depth camera external parameter; π (K, u, Z (u)) is the coordinate vector be converted to according to its depth value Z (u) by depth image vegetarian refreshments U under depth camera coordinate system.
The expression formula of described coordinate vector u is as follows:
u＝KR
-1(c-t)
Determine that the method waiting to fetch data affiliated block is: the position vector r being calculated affiliated block of waiting to fetch data by following formula, can be found by the arrangement sequence number in position vector r and wait the particular location of affiliated block in three-dimensional space that fetch data;
Wherein: x, y, z corresponds to the coordinate figure waiting to fetch data on three-dimensional space X-axis, Y-axis, Z axis, L is the length of side of block,
for downward bracket function, Num
x, Num
y, Num
zcorrespond to and wait the arrangement sequence number of affiliated block in three-dimensional space X-direction, Y direction, Z-direction of fetching data.
When block Octree is searched for, first judge that whether this tree is empty tree: set if it is empty, then show there are not any TSDF data in this tree block, directly skip over the block that this tree block advances to lower one tree and search for; Set if not empty, then adopt the searching method of depth-first to search for this tree block, to find in this tree with the bottom leaf node that coordinate to be fetched data matches and the TSDF data read in its corresponding voxel; In search procedure, if the TSDF data in the corresponding voxel of present node non-bottom leaf node or present node are undesirable, then skip over this node, advance to next brother of node and continue search until search satisfactory TSDF data.
Hinge structure, the present invention has following Advantageous Effects:
(1) tool of the present invention has carried out tree construction compression to the voxel data evenly divided in original KinectFusion algorithm, and significantly reduce the storage resource consumption of space voxel data, compression efficiency reaches 90%.
(2) tool of the present invention devises the data search method based on Octree forest structure, can carry out efficiently reading fast to stored data.
(3) forest structure of tool design of the present invention makes the degree of depth of every Octree less, and search speed is faster.
Accompanying drawing explanation
Fig. 1 is the structural representation of body space.
Fig. 2 (a) and Fig. 2 (b) is respectively two kinds of situation schematic diagram that node needs segmentation.
Fig. 3 is the structural representation in the corresponding space of Octree and node thereof.
Fig. 4 is the storage resources comparison diagram of Octree forest compression method of the present invention and KinectFusion algorithm.
Fig. 5 (a) and Fig. 5 (b) is respectively KinectFusion algorithm surface reconstructed results and partial enlarged drawing thereof.
Fig. 6 (a) and Fig. 6 (b) are respectively the resurfacing result of the present invention after 4 layers of Octree forests compression and partial enlarged drawing thereof.
Fig. 7 (a) and Fig. 7 (b) are respectively the resurfacing result of the present invention after 5 layers of Octree forests compression and partial enlarged drawing thereof.
Embodiment
In order to more specifically describe the present invention, below in conjunction with the drawings and the specific embodiments, technical scheme of the present invention is described in detail.
The present embodiment will carry out the compression of affiliated Octree forest structure by reference to the accompanying drawings to the body space of a 3m × 3m × 3m size.
As shown in Figure 1, first whole body space is carried out rough even division 30 × 30 × 30 small cubes, each cube size is 0.1m, and then in each zonule, build an Octree, Octrees all in such body space constitutes an Octree forest.One piece of cubical area in the corresponding physical space of each Octree, and whether every tree belt has a tree label to be non-NULL tree to mark this tree.
Concrete forest builds with growing method as follows:
When the camera parameters of a given amplitude deepness image and its correspondence, can judge that every each node set is the need of growth segmentation according to three-dimensional mapping relations.The method judged is as follows:
Suppose the central point c=(X, Y, Z) of the corresponding area of space of Octree node
t, its region cube length of side is l, by coordinate transform, can be mapped in the projection plane of video camera the depth image vegetarian refreshments u=KR finding correspondence by selecting c
-1(c-t), wherein K is video camera internal reference matrix, R and t is respectively rotation matrix and the translation vector of video camera external parameter.
Then between this pixel being made the return trip empty according to its back projection that fathoms:
p＝R·π(K,u,Z(u))+t
Wherein: pixel u is converted to the 3D position under camera coordinate system by π (K, u, Z (u)) functional expression according to its depth value Z (u).When two kinds of situations below occurring any one time, then this node region needs further growth to segment:
1. when p point is in the encircle sphere of this node space, as shown in Fig. 2 (a).It is comparatively large that this situation generally appears at node space, and space includes surface data.Wherein ε is that expansion trace (being generally set to 0.01m) is asked in encirclement;
2. the distance of pixel projection point p distance node center c is less than or equal to and blocks distance T (being generally 0.03m), as shown in Fig. 2 (b).It is less that this situation generally appears at node space, and the situation that node center is nearer apart from surface;
||p-c||≤T
When node reaches the depth capacity D of Octree
maxtime (being generally 4 or 5), then no longer segment, and by TSDF data stored in depth capacity node.Therefore the leaf node being positioned at depth capacity in Octree is the Minimum Area unit of spatial surface data, corresponding to the voxel in former KinectFusion algorithm.What leaf node stored is TSDF data.
When obtaining a new two field picture, by the method, all nodes in traversal forest, successively grow renewal from root node to leaf node, upgrade the TSDF data in all leaf nodes, the Octree topological structure built and space structure corresponding to node thereof are as shown in Figure 3 simultaneously.
As shown in Figure 4, the cycle tests that we have chosen one section of 300 frame has carried out record to consumed storage resources, and after the compression of Octree forest, storage resources occupancy reduces greatly.Former KinectFusion algorithm is constant takies 512M internal memory, and the maximum occupancy under same volume resolution after compression also only has 52M internal memory.
Data search; Need the data of searching leaf node when carrying out ray cast, present embodiment adopts a kind of method of hierarchical search to carry out fast finding, and detailed process is as follows:
Step one: for any point g=(X, Y, Z) in space
t, obtain according to the length of side l in the corresponding space of every the root vertex evenly divided which tree that this point is arranged in forest:
Wherein:
for rounding downwards.
Step 2: judge whether this tree is empty tree, if be empty, so illustrate that this root vertex corresponding region does not exist surface data, directly can skip over the region that this tree region advances to lower one tree according to the tree label of tree r; If this tree is not empty, then carries out step 3 and carry out meticulousr search.
Step 3: when setting as non-NULL, then carry out the searching method of depth-first, according to the Octree search of having built up to leaf node, reads corresponding TSDF data.If in search procedure, certain node do not have leaf node or TSDF data undesirable, then skip over this node, advance to next brother of node continue search until search satisfactory leaf data.
As shown in Fig. 5 ~ 7, Kinfu512 is the surface model that original KinectFusion light renders; Oct_f4 is the reconstructed results of 4 layers of Octree forest, and theoretical resolution is identical with Kinfu512; Oct_f5 is the reconstructed results of 5 layers of Octree forest, and theoretical resolution is the twice of Kinfu512.Can find out that the reconstructed results after the compression of Octree forest is suitable with former arithmetic result, not occur data degradation, and higher reconstruction precision can be reached.
Above-mentioned is can understand and apply the invention for ease of those skilled in the art to the description of embodiment.Person skilled in the art obviously easily can make various amendment to above-described embodiment, and General Principle described herein is applied in other embodiments and need not through performing creative labour.Therefore, the invention is not restricted to above-described embodiment, those skilled in the art are according to announcement of the present invention, and the improvement made for the present invention and amendment all should within protection scope of the present invention.
Claims (8)
1. based on a three-dimensional voxel access method for Octree forest compression, as follows:
Storage is operated: first, rough even partition is carried out to three-dimensional space, obtain the block of multiple identical square; Then, in described block, build Octree, namely utilize depth camera back projection dynamic tree correspondence carries out block segmentation from root node; Finally, TSDF data are stored in the voxel corresponding to Octree bottom leaf node;
For read operation: first, according to the coordinate waiting to fetch data in three-dimensional space, determine the block belonging to it; Then, adopt the searching method of depth-first to search for this block Octree, to search the bottom leaf node matched with coordinate to be fetched data in tree; Finally, from voxel corresponding to the leaf node searched, corresponding TSDF data are read.
2. three-dimensional voxel access method according to claim 1, it is characterized in that: the concrete grammar of described structure Octree is: using corresponding for the block root node as Octree, if by judging to divide block, then block is evenly divided into the voxel of eight identical squares, eight child nodes under eight corresponding root nodes of voxel difference; In like manner, if by judging to divide voxel, then further voxel is evenly divided into eight little voxels, eight little voxels, eight child nodes respectively under corresponding large voxel node; When the Octree number of plies reaches the default maximum number of plies, even if voxel meets the requirement of Further Division, also it is not split again.
3. three-dimensional voxel access method according to claim 2, is characterized in that: judge that the conditioning process whether block or voxel need to carry out dividing is as follows:
First, by coordinate transform, the central point C of block or voxel is mapped in the projection plane of depth camera, obtains corresponding depth image vegetarian refreshments U;
Then, depth image vegetarian refreshments U is returned in three-dimensional space according to its depth value back projection, obtain the subpoint P of central point C in space on body surface;
Finally, need be divided block or voxel by following judging whether according to the three-dimensional coordinate of central point C and subpoint P:
If block or voxel do not exist with body surface in three-dimensional space intersection, then meet following relation and then block or voxel divided:
||p-c||≤T
If block or voxel exist with body surface in three-dimensional space intersection, then meet following relation and then block or voxel divided:
Wherein: p and c is respectively subpoint P and the coordinate vector of central point C in three-dimensional space, and l is the length of side of block or voxel, T and ε is the intercept parameter of setting.
4. three-dimensional voxel access method according to claim 2, is characterized in that: if block does not make Further Division, then the Octree of its correspondence is empty tree, does not have any TSDF data in this block.
5. three-dimensional voxel access method according to claim 3, is characterized in that: the expression formula of described coordinate vector p is as follows:
p＝R·π(K,u,Z(u))+t
Wherein: u is the coordinate vector of depth image vegetarian refreshments U in three-dimensional space, K is the internal reference matrix of depth camera, Z (u) is the depth value of depth image vegetarian refreshments U, R and t is respectively rotation matrix and the translation vector of depth camera external parameter; π (K, u, Z (u)) is the coordinate vector be converted to according to its depth value Z (u) by depth image vegetarian refreshments U under depth camera coordinate system.
6. three-dimensional voxel access method according to claim 5, is characterized in that: the expression formula of described coordinate vector u is as follows:
u＝KR
-1(c-t)。
7. three-dimensional voxel access method according to claim 1, it is characterized in that: determine that the method waiting to fetch data affiliated block is: the position vector r being calculated affiliated block of waiting to fetch data by following formula, can be found by the arrangement sequence number in position vector r and wait the particular location of affiliated block in three-dimensional space that fetch data;
Wherein: x, y, z corresponds to the coordinate figure waiting to fetch data on three-dimensional space X-axis, Y-axis, Z axis, L is the length of side of block,
for downward bracket function, Num
x, Num
y, Num
zcorrespond to and wait the arrangement sequence number of affiliated block in three-dimensional space X-direction, Y direction, Z-direction of fetching data.
8. three-dimensional voxel access method according to claim 1, it is characterized in that: when block Octree is searched for, first judge whether this tree is empty tree: set if it is empty, then show there are not any TSDF data in this tree block, directly skip over the block that this tree block advances to lower one tree and search for; Set if not empty, then adopt the searching method of depth-first to search for this tree block, to find in this tree with the bottom leaf node that coordinate to be fetched data matches and the TSDF data read in its corresponding voxel; In search procedure, if the TSDF data in the corresponding voxel of present node non-bottom leaf node or present node are undesirable, then skip over this node, advance to next brother of node and continue search until search satisfactory TSDF data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410765462.2A CN104616345B (en) | 2014-12-12 | 2014-12-12 | Octree forest compression based three-dimensional voxel access method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410765462.2A CN104616345B (en) | 2014-12-12 | 2014-12-12 | Octree forest compression based three-dimensional voxel access method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104616345A true CN104616345A (en) | 2015-05-13 |
| CN104616345B CN104616345B (en) | 2017-05-24 |
Family
ID=53150777
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410765462.2A Active CN104616345B (en) | 2014-12-12 | 2014-12-12 | Octree forest compression based three-dimensional voxel access method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104616345B (en) |
Cited By (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106548507A (en) * | 2015-09-16 | 2017-03-29 | 富士通株式会社 | The method and apparatus of three-dimensional reconstruction object |
| CN106548466A (en) * | 2015-09-16 | 2017-03-29 | 富士通株式会社 | The method and apparatus of three-dimensional reconstruction object |
| CN106558076A (en) * | 2015-09-16 | 2017-04-05 | 富士通株式会社 | The method and apparatus of three-dimensional reconstruction object |
| CN107507276A (en) * | 2016-06-14 | 2017-12-22 | 中国石油化工股份有限公司 | The 3-dimensional digital rock core storage method that slabbed core for any direction is shown |
| CN108010126A (en) * | 2017-12-11 | 2018-05-08 | 苏州蜗牛数字科技股份有限公司 | Method and system based on voxel structure large-scale complex landform |
| CN108090404A (en) * | 2016-11-23 | 2018-05-29 | 法乐第(北京)网络科技有限公司 | Three dimensional image processing method |
| CN108196266A (en) * | 2017-12-29 | 2018-06-22 | 中山大学 | Vegetation canopy three-dimensional radiative transfer analogy method based on Lidar data |
| CN108268879A (en) * | 2016-12-30 | 2018-07-10 | 乐视汽车（北京）有限公司 | Three-dimensional full convolutional network implementation method, equipment and electronic equipment |
| CN108805261A (en) * | 2017-04-28 | 2018-11-13 | 微软技术许可有限责任公司 | Convolutional neural networks based on Octree |
| CN108921774A (en) * | 2018-06-08 | 2018-11-30 | 广州虎牙信息科技有限公司 | The storage organization and related edit method, apparatus, equipment and storage medium of model |
| CN109155846A (en) * | 2018-08-14 | 2019-01-04 | 深圳前海达闼云端智能科技有限公司 | Three-dimensional reconstruction method and device of scene, electronic equipment and storage medium |
| CN109861965A (en) * | 2018-12-05 | 2019-06-07 | 苏州蜗牛数字科技股份有限公司 | A kind of voxel Terrain Data Compression method and system |
| CN110892725A (en) * | 2017-07-13 | 2020-03-17 | 交互数字Vc控股公司 | Method and apparatus for encoding/decoding a point cloud representing a 3D object |
| US10853998B2 (en) | 2016-10-12 | 2020-12-01 | Hewlett-Packard Development Company, L.P.C | Serialising a representation of a three dimensional object |
| CN112372632A (en) * | 2020-10-12 | 2021-02-19 | 达闼机器人有限公司 | Mechanical arm motion control method and device and computer readable storage medium |
| CN112734902A (en) * | 2020-12-24 | 2021-04-30 | 清华大学天津高端装备研究院 | Method for rapid modeling by using depth camera |
| CN113487690A (en) * | 2021-09-08 | 2021-10-08 | 北京医百科技有限公司 | Depth map processing method, system and storage medium |
| CN113713381A (en) * | 2021-09-09 | 2021-11-30 | 腾讯科技（深圳）有限公司 | Object management method, device, equipment, storage medium and system |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030117402A1 (en) * | 2001-12-21 | 2003-06-26 | Hubrecht Alain Yves Nestor | Systems and methods for simulating frames of complex virtual environments |
| CN1632834A (en) * | 2003-12-22 | 2005-06-29 | 中国科学院自动化研究所 | Point reconstruction based very large-scale medical image 3D visualization method |
| CN101763652A (en) * | 2009-06-03 | 2010-06-30 | 中国科学院自动化研究所 | Three-dimensional framework fast extraction method based on branch feathers |
| EP2339481A1 (en) * | 2009-12-03 | 2011-06-29 | National Digital Research Centre | Enablement of three-dimensional hosting, indexing, analysing and querying structure for spatial systems |
| CN103236058A (en) * | 2013-04-25 | 2013-08-07 | 内蒙古科技大学 | Method for obtaining volume of interest of four-dimensional heart image |
-
2014
- 2014-12-12 CN CN201410765462.2A patent/CN104616345B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030117402A1 (en) * | 2001-12-21 | 2003-06-26 | Hubrecht Alain Yves Nestor | Systems and methods for simulating frames of complex virtual environments |
| CN1632834A (en) * | 2003-12-22 | 2005-06-29 | 中国科学院自动化研究所 | Point reconstruction based very large-scale medical image 3D visualization method |
| CN101763652A (en) * | 2009-06-03 | 2010-06-30 | 中国科学院自动化研究所 | Three-dimensional framework fast extraction method based on branch feathers |
| EP2339481A1 (en) * | 2009-12-03 | 2011-06-29 | National Digital Research Centre | Enablement of three-dimensional hosting, indexing, analysing and querying structure for spatial systems |
| CN103236058A (en) * | 2013-04-25 | 2013-08-07 | 内蒙古科技大学 | Method for obtaining volume of interest of four-dimensional heart image |
Non-Patent Citations (4)
| Title |
|---|
| FRISKEN S 等: "Simple and effi cient t raversal methods for quadt rees and oct rees", 《GRAPHICS TOOLS》 * |
| 吴晓军 等: "基于八叉树的三维网格模型体素化方法", 《工程图学学报》 * |
| 毛先成: "三维数字矿床与隐伏矿体立体定量预测研究", 《中国博士学位论文全文数据库》 * |
| 穆斌 等: "基于投影体积与八叉树的三维网格模型体素化方法", 《地理与地理信息科学》 * |
Cited By (29)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106548466B (en) * | 2015-09-16 | 2019-03-29 | 富士通株式会社 | The method and apparatus of three-dimensional reconstruction object |
| CN106548466A (en) * | 2015-09-16 | 2017-03-29 | 富士通株式会社 | The method and apparatus of three-dimensional reconstruction object |
| CN106558076A (en) * | 2015-09-16 | 2017-04-05 | 富士通株式会社 | The method and apparatus of three-dimensional reconstruction object |
| CN106548507A (en) * | 2015-09-16 | 2017-03-29 | 富士通株式会社 | The method and apparatus of three-dimensional reconstruction object |
| CN106558076B (en) * | 2015-09-16 | 2019-06-18 | 富士通株式会社 | The method and apparatus of three-dimensional reconstruction object |
| CN107507276A (en) * | 2016-06-14 | 2017-12-22 | 中国石油化工股份有限公司 | The 3-dimensional digital rock core storage method that slabbed core for any direction is shown |
| CN107507276B (en) * | 2016-06-14 | 2020-06-23 | 中国石油化工股份有限公司 | Three-dimensional digital core storage method for core slice display in any direction |
| US10853998B2 (en) | 2016-10-12 | 2020-12-01 | Hewlett-Packard Development Company, L.P.C | Serialising a representation of a three dimensional object |
| CN108090404A (en) * | 2016-11-23 | 2018-05-29 | 法乐第(北京)网络科技有限公司 | Three dimensional image processing method |
| CN108268879A (en) * | 2016-12-30 | 2018-07-10 | 乐视汽车（北京）有限公司 | Three-dimensional full convolutional network implementation method, equipment and electronic equipment |
| CN108805261B (en) * | 2017-04-28 | 2021-11-12 | 微软技术许可有限责任公司 | Convolutional neural network based on octree |
| US11704537B2 (en) | 2017-04-28 | 2023-07-18 | Microsoft Technology Licensing, Llc | Octree-based convolutional neural network |
| CN108805261A (en) * | 2017-04-28 | 2018-11-13 | 微软技术许可有限责任公司 | Convolutional neural networks based on Octree |
| CN110892725A (en) * | 2017-07-13 | 2020-03-17 | 交互数字Vc控股公司 | Method and apparatus for encoding/decoding a point cloud representing a 3D object |
| CN108010126A (en) * | 2017-12-11 | 2018-05-08 | 苏州蜗牛数字科技股份有限公司 | Method and system based on voxel structure large-scale complex landform |
| CN108196266B (en) * | 2017-12-29 | 2021-07-02 | 中山大学 | Vegetation canopy three-dimensional radiation transmission simulation method based on Lidar data |
| CN108196266A (en) * | 2017-12-29 | 2018-06-22 | 中山大学 | Vegetation canopy three-dimensional radiative transfer analogy method based on Lidar data |
| CN108921774A (en) * | 2018-06-08 | 2018-11-30 | 广州虎牙信息科技有限公司 | The storage organization and related edit method, apparatus, equipment and storage medium of model |
| WO2020034086A1 (en) * | 2018-08-14 | 2020-02-20 | 深圳前海达闼云端智能科技有限公司 | Three-dimensional reconstruction method and apparatus for scene, and electronic device and storage medium |
| CN109155846B (en) * | 2018-08-14 | 2020-12-18 | 深圳前海达闼云端智能科技有限公司 | Three-dimensional reconstruction method and device of scene, electronic equipment and storage medium |
| CN109155846A (en) * | 2018-08-14 | 2019-01-04 | 深圳前海达闼云端智能科技有限公司 | Three-dimensional reconstruction method and device of scene, electronic equipment and storage medium |
| CN109861965B (en) * | 2018-12-05 | 2021-06-18 | 苏州蜗牛数字科技股份有限公司 | Voxel terrain data compression method and system |
| CN109861965A (en) * | 2018-12-05 | 2019-06-07 | 苏州蜗牛数字科技股份有限公司 | A kind of voxel Terrain Data Compression method and system |
| CN112372632A (en) * | 2020-10-12 | 2021-02-19 | 达闼机器人有限公司 | Mechanical arm motion control method and device and computer readable storage medium |
| WO2022078234A1 (en) * | 2020-10-12 | 2022-04-21 | 达闼机器人有限公司 | Mechanical arm movement control method and device, and computer-readable storage medium |
| CN112734902A (en) * | 2020-12-24 | 2021-04-30 | 清华大学天津高端装备研究院 | Method for rapid modeling by using depth camera |
| CN113487690A (en) * | 2021-09-08 | 2021-10-08 | 北京医百科技有限公司 | Depth map processing method, system and storage medium |
| CN113713381A (en) * | 2021-09-09 | 2021-11-30 | 腾讯科技（深圳）有限公司 | Object management method, device, equipment, storage medium and system |
| CN113713381B (en) * | 2021-09-09 | 2023-06-20 | 腾讯科技（深圳）有限公司 | Object management method, device, equipment, storage medium and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104616345B (en) | 2017-05-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104616345A (en) | Octree forest compression based three-dimensional voxel access method | |
| Wang et al. | Lidar point clouds to 3-D urban models $: $ A review | |
| CN101533529B (en) | Range image-based 3D spatial data processing method and device | |
| CN105378796B (en) | Scalable volume 3D reconstruct | |
| Zeng et al. | Octree-based fusion for realtime 3D reconstruction | |
| Xu et al. | Combining IFC and 3D tiles to create 3D visualization for building information modeling | |
| CN103021017B (en) | Three-dimensional scene rebuilding method based on GPU acceleration | |
| CN110675496B (en) | Grid subdivision and visualization method and system based on three-dimensional urban geological model | |
| CN108919944A (en) | A kind of virtual roaming method carrying out data lossless interaction in display end based on digital city model realization | |
| CN107833273B (en) | Oblique photography three-dimensional model objectification application method based on three-dimensional simulation model | |
| CN112530005B (en) | Three-dimensional model linear structure recognition and automatic restoration method | |
| CN102831634B (en) | Efficient accurate general soft shadow generation method | |
| CN105825471A (en) | Unity-3D-based three-dimensional surface reconstruction and rendering method | |
| CN104318605A (en) | Parallel lamination rendering method of vector solid line and three-dimensional terrain | |
| CN115221580A (en) | Godot-based park digital twin building model construction method | |
| Ioannidis et al. | 5D Multi-Purpose Land Information System. | |
| CN115631317B (en) | Tunnel lining ortho-image generation method and device, storage medium and terminal | |
| CN109472862A (en) | A kind of substation's 3 d modeling system | |
| Pan et al. | A visibility-based surface reconstruction method on the GPU | |
| Wang et al. | Visual-perception-driven Urban Three-dimensional Scene Data Scheduling Method. | |
| Amiraghdam et al. | LOOPS: LOcally Optimized Polygon Simplification | |
| CN103985150B (en) | A kind of image processing method and electronic equipment | |
| Zhou et al. | Progress and review of 3D image feature reconstruction | |
| CN117456550B (en) | MR-based CAD file viewing method, device, medium and equipment | |
| Liu et al. | Fusing multiscale charts into 3D ENC systems based on underwater topography and remote sensing image |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |