CN108052914A - A kind of forest forest resource investigation method identified based on SLAM and image - Google Patents
A kind of forest forest resource investigation method identified based on SLAM and image Download PDFInfo
- Publication number
- CN108052914A CN108052914A CN201711391790.0A CN201711391790A CN108052914A CN 108052914 A CN108052914 A CN 108052914A CN 201711391790 A CN201711391790 A CN 201711391790A CN 108052914 A CN108052914 A CN 108052914A
- Authority
- CN
- China
- Prior art keywords
- forest
- data
- sample
- image
- acquisition
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/64—Three-dimensional objects
- G06V20/653—Three-dimensional objects by matching three-dimensional models, e.g. conformal mapping of Riemann surfaces
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Software Systems (AREA)
- Computer Graphics (AREA)
- Geometry (AREA)
- Image Processing (AREA)
Abstract
The present invention provides a kind of forest forest resource investigation methods identified based on SLAM and image, comprise the following steps：S1, image data and environmental data on the entire forest sample ground to be investigated of covering；S2 is handled the image data of acquisition using SLAM methods, with generating sample threedimensional model；S3 carries out singulation processing to threedimensional model, finds corresponding region on the image and be identified, obtain the three dimensional space coordinate of forest, height, the diameter of a cross-section of a tree trunk 1.3 meters above the ground, the parameters such as type and quantity；S4, scene original data and treated data import forest inventory control Single Component Management by the sample of acquisition.This method obtains forest data by vision camera and sensor, scale Forest Scene threedimensional model is built using SLAM methods, the identification and analysis of forest data are completed by point cloud segmentation and image identification, reduce the field work time, the efficiency of forest inventory investigation is improved, while ensure that the verifiability of forest inventory investigation result.
Description
Technical field
The present invention relates to a kind of forest forest resource investigation methods, belong to forest survey, computer vision and image procossing
Technical field.
Background technology
The forest reserves are the important components of natural resources, are the important substance bases of national economy and social development.
As forest resourceies are excessively cut down by the mankind using timber shortage and ecological environment is caused to be deteriorated, the mankind must not be without
Work is managed in forest inventory investigation and planning.And animal and plant and its environment item to be grown in the range of forest land, forest and forest zone
Part is that the forest inventory investigation of object is agricultural department, timber user and the reliable letter of personnel's offer for formulating and implementing national policy
Breath.
Stand' grade is the main contents of forest inventory investigation, the investigation factor of stand' grade mainly include seeds, the age,
The diameter of a cross-section of a tree trunk 1.3 meters above the ground, tree are high, and the acquisition of these data needs substantial amounts of survey tool and field measurement work.These survey tools at present
Use need manually to carry out, the accuracy of observed result dependent on observer work attainment and survey tool is used ripe
Practice degree, materially increase the error of survey data；Secondly as the complexity of forest environment and part measure
The volume and weight of instrument is bigger, further increases the difficulty of measurement work, influences the efficiency of measurement；Finally, acquisition
Investigation result is measured obtained data, lacks necessary method and the accuracy of data is verified.Therefore, reduce gloomy
Artificial action and work difficulty during woods resource investigation, it is gloomy to improve in the speed and accuracy of forest inventory investigation
Urgent problem to be solved during woods resource investigation.
The content of the invention
The present invention provides a kind of forest forest resource investigation methods identified based on SLAM and image, it is desirable to provide a kind of
To forest forest resource identification, forest forest resource investigation method that investigation factor obtains automatically solves current forest reserves tune
The problems such as manually dependence is big during looking into, of high cost, and investigation efficiency is low, and the cycle is long, and the accuracy of investigation result can not be verified,
So as to improve the efficiency and accuracy of forest forest resource investigation, the research cost of the forest reserves is reduced.
A kind of forest forest resource investigation method based on SLAM and deep learning of the present invention, comprises the following steps：
S1 carries out forest sample to be investigated data acquisition, obtains with covering entire the sample image data of scene and environment number
According to；
S2 is handled the image data of acquisition by SLAM methods, with generating sample the threedimensional model of scene；
S3 carries out singulation processing to threedimensional model, finds corresponding region on the image and be identified, to the result of identification into
One step is handled, with obtaining sample the type and quantity of forest, the investigation factors such as three dimensional space coordinate, height, the diameter of a cross-section of a tree trunk 1.3 meters above the ground of sample trees；
S4, scene original data and treated data import forest inventory control unit and are managed by the sample of acquisition.
A kind of forest forest resource investigation method based on SLAM and deep learning of the present invention, utilizes stereoscopic vision phase
Sensors such as machine, inertial navigation and alignment system etc. carry out forest sample to be investigated adopting for continuous image data and environmental information
Collection, the image data and the posture information of camera analyzed, using Spatial Mapping or Kintinuous etc.
SLAM algorithms generation three dimensions point cloud obtains the three-dimensional modeling data on sample ground to be investigated.By obtained three-dimensional modeling data into
Row identification and segmentation, are then handled the data after segmentation, and with reference to image data, generation data processing report obtains woods
The investigation factors such as species, quantity, the coordinate of sample trees, the height of wood.Data processing report, threedimensional model and initial data are imported
To forest inventory control unit, carry out the management of forest resource data, check, with the comparison of historical data and the various reports of generation
It accuses.
A kind of implementation of forest forest resource investigation method based on SLAM and deep learning of the present invention, bag
It includes：
Data acquisition unit carries out forest sample to be investigated the acquisition of image and environmental information, and acquisition is with covering forest sample
Complete image data and environmental information, and the data sending collected to model construction unit is handled；
Model construction unit, processing data acquisition unit obtained forest sample ground image data, by SLAM modeling techniques to figure
As data are handled, the three-dimensional point cloud on forest sample ground is obtained；
Data processing unit, receive three dimensional point cloud, three-dimensional point cloud is identified and is split, does singulation processing, into
Treated that data do further data analysis for row singulation, the parameters such as tree height, the diameter of a cross-section of a tree trunk 1.3 meters above the ground of sample trees is obtained, finally with reference to data
The environmental information that collecting unit obtains forms data processing report, and is conducted into Data Management Unit and is managed and ties up
Shield；
Data Management Unit, the data that the data and subsequent processing obtained to data acquisition unit obtain are stored and tieed up
Shield.
The advantages and positive effects of the present invention are：
（1）Equipment needed for data acquisition of the present invention is mainly vision camera, embedded platform and sensor, for example IMU, temperature pass
Sensor, humidity sensor etc., small compared to various forest measuring apparatus, easy to operate, artificial requirement and at low cost,
Influence of the complicated forest environment to data acquisition is reduced, data acquisition efficiency is high；
（2）The present invention automatically processes data using point cloud segmentation and image identification and generates mabage report, data processing
Efficient, the used time is short, reduces artificial input；
（3）The present invention is by initial data and treated that data are importing directly into Data Management Unit is managed, and by correlation
Data associate to use for inquiry, avoid manually entering, and solve current methodology for forest resource survey and be difficult to pair
The problem of investigation result is traced and verified, while be conducive to the comparison of current survey data and historical data.
Description of the drawings
Fig. 1 is a kind of work flow diagram of the forest forest investigation method identified based on SLAM and image；
Fig. 2 is a kind of implementation system framework of the forest forest resource investigation method identified based on SLAM and image.
Specific embodiment
As shown in Figure 1, a kind of forest forest resource investigation method based on SLAM and deep learning, comprises the following steps：
S1 carries out forest sample to be investigated data acquisition, obtains the image data and environmental data of with covering entire sample scene；
S2 is handled the image data of acquisition by SLAM methods, with generating sample the threedimensional model of scene；S3, to threedimensional model
Singulation processing is carried out, corresponding region is found on the image and is identified, the result of identification is further processed, with obtaining sample
The type and quantity of forest, the investigation factors such as three dimensional space coordinate, height, the diameter of a cross-section of a tree trunk 1.3 meters above the ground of sample trees；S4, scene is former by the sample of acquisition
Data and treated data import forest inventory control unit and are managed.Pass through the sensings such as stereo vision camera, inertial navigation
Scene image and environmental data are acquired device to sample, scene threedimensional model by SLAM method spanning forest samples, are passed through
The statistics of deep learning with completing forest sample forest tree resource, whole process do not depend on special hardware device, need not be complicated
Manual operation and professional knowledge, at the same gather and generation data for follow-up forest inventory investigation result verification, retrospect and
The analysis of dynamic changes of forest resources provides foundation, solves existing human cost during current forest inventory investigation
The problems such as height, poll cycle are long, investigation result error is big, investigation result is difficult to verify.
The step S1 described in forest forest resource investigation method that this kind is identified based on SLAM and image comprises the following steps：
The general path of S11, with determining forest sample data acquisition carry out data acquisition according to path, and ensureing the data of acquisition can cover
With covering entire sample scene；S12 is configured the initiation parameter of data acquisition device, main to gather picture including camera
Resolution ratio and frame per second, the contrast of camera acquisition picture color, the starting point coordinate of data acquisition, the form of storage data and position
It puts, data acquisition modes etc.；S13 carries out data acquisition along the data acquisition path of planning, and record is obtained by vision camera
Forest sample ground image data and by environmental sensor obtain environmental information；S14 can lead in data acquisition
The browsing data function of crossing data acquisition device checks the quality of gathered data.
The step S2 described in forest forest resource investigation method that this kind is identified based on SLAM and image comprises the following steps：
S21 obtains the depth map of each frame of gathered data and the pose of camera by the step S1 data gathered；S22, by the depth of reading
Degree figure is converted into three-dimensional point cloud and calculates the normal vector of every bit；According to the pose of camera, the point cloud of present frame is merged by S23
Into grid model；S24 is obtained from model projection under present frame visual angle according to present frame camera pose using ray casting algorithm
Point cloud, and its normal vector is calculated, for the input picture registration to next frame；S25 cycles S22 to S24, in cycling
In the process, the point cloud under scene different visual angles is obtained by mobile camera, rebuilds complete scene surface.
The step S3 described in forest forest resource investigation method that this kind is identified based on SLAM and image comprises the following steps：
S31 uses the threedimensional model of S2 generations the point cloud segmentations algorithms such as RanSaC to realize the extraction of single target in point cloud；S32 leads to
Imaging relations and RGBD registration informations are crossed, find the region corresponding to single target on the image；It is not of the same race to establish forest by S33
The data set of class forest；Based on the data set of forest forest, single target forest is identified by image-recognizing method by S34
Information；S35 is handled single target to what S31 was obtained, obtain the stand' grades such as tree height, the diameter of a cross-section of a tree trunk 1.3 meters above the ground of forest because
Son, the environmental information that combining environmental sensor obtains export data with defined form, obtain data processing report；
The step S4 described in forest forest resource investigation method that this kind is identified based on SLAM and image comprises the following steps：S41,
Contextual data and treated that three-dimensional modeling data is arranged in the form of file to the sample of acquisition；S42 will be gathered
Sample the position of contextual data, the position of three-dimensional modeling data, sample trees data and environmental information be expressed as XML format；S43,
Into the data import modul of forest inventory control unit, XML file is loaded, forest sample is obtained by Data acquisition and issuance
To structural data and unstructured data imported into forest inventory control unit；S44 into forest management unit, is checked
Forest identification information and forest environment information, the integrality that detection data import；S45 passes through data analysis and comparison, generation
The various reports of forest forest tree resource, such as forest tree resource information report, temperature change report.
As described in Figure 2, a kind of implementation of the forest forest resource investigation method identified based on SLAM and image, bag
It includes：Data acquisition unit, carries out forest sample to be investigated the acquisition of image and environmental information, and acquisition is with covering forest sample complete
Whole image data and environmental information, and the data sending collected to model construction unit is handled；Model construction list
Member, image data, is handled image data by SLAM modeling techniques with handling the forest sample that data acquisition unit obtains,
Obtain the three-dimensional point cloud on forest sample ground；Data processing unit receives three dimensional point cloud, three-dimensional point cloud is identified and is divided
It cuts, does singulation processing, treated that data do further data analysis to carrying out singulation, and the tree for obtaining sample trees is high, chest
The parameters such as footpath, the environmental information finally obtained with reference to data acquisition unit form data processing report, and are conducted into data
Administrative unit is managed and safeguards；Data Management Unit, the data and subsequent processing obtained to data acquisition unit obtain
Data stored and safeguarded.
Data acquisition described in the implementation for the forest forest resource investigation method that this kind is identified based on SLAM and image
Unit includes：Parameter setting module sets the sample frequency of camera, the resolution ratio for gathering image, color contrast, storage data
Form etc.；Data disaply moudle, to the environment that camera acquired image information, sensor are gathered in data acquisition
Information and the routing information of data acquisition are shown in real time；Browsing data module checks and is gathered and stored by camera
Image data.
Model construction described in the implementation for the forest forest resource investigation method that this kind is identified based on SLAM and image
Unit includes：The data of acquisition are converted to PNG coloured pictures and depth map needed for subsequent algorithm by data conversion module；Algorithm selects
Module is selected, selection is for the threedimensional model algorithm built and the parameter that threedimensional model is set to build, threedimensional model developing algorithm bag
It includes：Elastic Fussion, kintinuous Spatial Mapping etc. are, it is necessary to which the parameter set includes：Point cloud
Density, the form for exporting threedimensional model etc.；Model browsing module browses the threedimensional model of generation, can show a cloud,
Grid, wire frame etc..
Data processing described in the implementation for the forest forest resource investigation method that this kind is identified based on SLAM and image
Unit includes：Threedimensional model and initial data are imported into data processing unit and handled by data import modul；Tree information
Module by the collection of forest information, is entered into forest information module, in data handling procedure
Image identifies；Model splits module, may be selected different point cloud segmentation algorithm, for example, edge detection method, scan-line algorithm,
RanSaC algorithms etc. carry out the singulation processing of threedimensional model；Picture recognition module selects monomer model, by imaging relations and
RGBD registration informations find the region corresponding to single target on the image, then in conjunction with the woods stored in tree information module
Wooden information obtains the corresponding forest species of monomer model；Forest parameter identification module, it is right according to the demand of stand' grade parameter
Monomer model such as measures at the processing, obtains forest parameter；Report generation module is stored with various data processings reports in module
Masterplate, select different data report masterplates, generate different data processing reports.
Data management described in the implementation for the forest forest resource investigation method that this kind is identified based on SLAM and image
Unit includes：Including user management, rights management etc., the control with operating right is accessed for user for system management module；Number
According to import modul, can initial data, threedimensional model and data processing being reported to, importeding into Data Management Unit is managed；Number
It is investigated that see module, can temporally, place, classification etc. check the data for being deposited into Data Management Unit；Report generation
Module such as compares the data of acquisition, is merged at the spanning forests resource investigation report.
In conclusion compared with existing forest forest resource investigation method, the present invention in forest forest fact-finding process only
The data acquisition unit that need to be made of vision camera, embedded platform and sensor, small, at low cost, easy to operate, forest
Complex environment influences data acquisition small, data acquisition efficiency height, drastically reduces the time of artificial field work,
Data are automatically processed using point cloud segmentation and image identification and generate mabage report, data-handling efficiency is high, and the used time is short,
The data for finally gathering and handling can be importing directly into Data Management Unit and be managed, and relevant data are associated
So that inquiry uses, avoid manually entering, while solve current methodology for forest resource survey and be difficult to chase after investigation result
The problem of tracing back and verifying is conducive to the comparison of current survey data and historical data and the generation of survey report.
Claims (6)
- A kind of 1. forest forest resource investigation method identified based on SLAM and image, which is characterized in that comprise the following steps：S1 carries out forest sample to be investigated data acquisition, obtains with covering entire the sample image data of scene and environment number According to；S2 is handled the image data of acquisition by SLAM methods, with generating sample the threedimensional model of scene；S3 carries out singulation processing to threedimensional model, finds corresponding region on the image and be identified, to the result of identification into One step is handled, with obtaining sample the type and quantity of forest, the investigation factors such as three dimensional space coordinate, height, the diameter of a cross-section of a tree trunk 1.3 meters above the ground of sample trees；S4, scene original data and treated data import forest inventory control unit and are managed by the sample of acquisition.
- 2. a kind of forest forest resource investigation method identified based on SLAM and image according to claim 1, feature It is, the step S1 comprises the following steps：S11, data acquisition path, data acquisition is carried out according to path with determining forest sample, and ensureing the data of acquisition can cover Entire sample ground scene；S12 is configured the initiation parameter of data acquisition device, the main resolution ratio and frame for including camera acquisition picture Rate, the contrast of camera acquisition picture color, the starting point coordinate of data acquisition, the storage form of data and position, data acquisition Mode etc.；S13 carries out data acquisition, the figure on the forest sample ground that record is obtained by vision camera along the data acquisition path of planning Sheet data and the environmental information obtained by environmental sensor；S14 can check the quality of gathered data in data acquisition by the browsing data function of data acquisition device.
- 3. a kind of forest forest resource investigation method identified based on SLAM and image according to claim 1, feature It is, the step S2 comprises the following steps：S21 obtains the depth map of each frame of gathered data and the pose of camera by the step S1 data gathered；The depth map of reading is converted into three-dimensional point cloud and calculates the normal vector of every bit by S22；According to the pose of camera, the point cloud of present frame is fused in grid model by S23；S24 obtains the point cloud under present frame visual angle from model projection according to present frame camera pose using ray casting algorithm, and And its normal vector is calculated, for the input picture registration to next frame；S25 cycles S22 to S24, during cycling, the point cloud under scene different visual angles is obtained by mobile camera, is rebuild Complete scene surface.
- 4. a kind of forest forest resource investigation method identified based on SLAM and image according to claim 1, feature It is, the step S3 comprises the following steps：S31 uses the threedimensional model of S2 generations deep neural network to realize the extraction of single target in point cloud；S32 by imaging relations and RGBD registration informations, finds the region corresponding to single target on the image；S33 establishes the data set of forest variety classes forest；Based on the data set of forest forest, the information of single target forest is identified by image-recognizing method by S34；S35 is handled single target to what S31 was obtained, obtains the stand' grades factors such as tree height, the diameter of a cross-section of a tree trunk 1.3 meters above the ground of forest, with reference to The environmental information that environmental sensor obtains exports data with defined form, obtains data processing report.
- 5. a kind of forest forest resource investigation method identified based on SLAM and image according to claim 1, feature It is, the step S4 comprises the following steps：S41, contextual data and treated that three-dimensional modeling data is arranged in the form of file to the sample of acquisition；S42, the position of contextual data, the position of three-dimensional modeling data, sample trees data and environmental information are represented by the sample of acquisition For XML format；S43 into the data import modul of forest inventory control unit, loads XML file, forest sample is passed through data acquisition The structural data and unstructured data obtained with analysis imported into forest inventory control unit；S44 into forest management unit, checks forest identification information and forest environment information, and detection data import complete Property；S45, by data analysis and comparison, the various reports of spanning forest forest tree resource, as forest tree resource information report, Temperature change report etc..
- 6. being used to implement described in claim 1-5 any one is based on SLAM and the forest forest resource investigation of image identification The implementation of method, which is characterized in that including：Data acquisition unit carries out forest sample to be investigated the acquisition of image and environmental information, and acquisition is with covering forest sample Complete image data and environmental information, and the data sending collected to model construction unit is handled；Model construction unit, processing data acquisition unit obtained forest sample ground image data, by SLAM modeling techniques to figure As data are handled, the three-dimensional point cloud on forest sample ground is obtained；Data processing unit, receive three dimensional point cloud, three-dimensional point cloud is identified and is split, does singulation processing, into Treated that data do further data analysis for row singulation, the parameters such as tree height, the diameter of a cross-section of a tree trunk 1.3 meters above the ground of sample trees is obtained, finally with reference to data The environmental information that collecting unit obtains forms data processing report, and is conducted into Data Management Unit and is managed and ties up Shield；Data Management Unit, the data that the data and subsequent processing obtained to data acquisition unit obtain are stored and tieed up Shield.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711391790.0A CN108052914A (en) | 2017-12-21 | 2017-12-21 | A kind of forest forest resource investigation method identified based on SLAM and image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711391790.0A CN108052914A (en) | 2017-12-21 | 2017-12-21 | A kind of forest forest resource investigation method identified based on SLAM and image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108052914A true CN108052914A (en) | 2018-05-18 |
Family
ID=62131075
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711391790.0A Pending CN108052914A (en) | 2017-12-21 | 2017-12-21 | A kind of forest forest resource investigation method identified based on SLAM and image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108052914A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110110687A (en) * | 2019-05-15 | 2019-08-09 | 江南大学 | Fruit automatic identifying method on tree based on colouring information and three-D profile information |
| CN110443197A (en) * | 2019-08-05 | 2019-11-12 | 珠海格力电器股份有限公司 | A kind of visual scene intelligent Understanding method and system |
| CN111134112A (en) * | 2019-10-28 | 2020-05-12 | 山东省林木种质资源中心 | Forest germplasm resource field collection method and system |
| CN111179335A (en) * | 2019-12-28 | 2020-05-19 | 东北林业大学 | Standing tree measuring method based on binocular vision |
| WO2020211427A1 (en) * | 2019-04-16 | 2020-10-22 | 广东康云科技有限公司 | Segmentation and recognition method, system, and storage medium based on scanning point cloud data |
| CN113723224A (en) * | 2021-08-12 | 2021-11-30 | 新疆爱华盈通信息技术有限公司 | Method, device and medium for remotely and intelligently monitoring tree felling |
| CN115032607A (en) * | 2022-05-26 | 2022-09-09 | 季华实验室 | LiDAR SLAM data-based stumpage position and breast diameter estimation method and system |
| CN116205394A (en) * | 2023-05-05 | 2023-06-02 | 浙江茂源林业工程有限公司 | Forest resource investigation and monitoring method and system based on radio navigation |
| CN116311010A (en) * | 2023-03-06 | 2023-06-23 | 中国科学院空天信息创新研究院 | Method and system for woodland resource investigation and carbon sink metering |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP3865764B1 (en) * | 2006-08-15 | 2007-01-10 | アルスマエヤ株式会社 | Forest resource survey method and forest resource survey apparatus |
| CN101488226A (en) * | 2008-01-16 | 2009-07-22 | 中国科学院自动化研究所 | Tree measurement and reconstruction method based on single three-dimensional laser scanning |
| KR20110093707A (en) * | 2010-02-10 | 2011-08-18 | 대한민국(관리부서 : 산림청 국립산림과학원장) | A method for forest inventory field survey system using gps and rfid |
| CN102881039A (en) * | 2012-07-30 | 2013-01-16 | 中国林业科学研究院资源信息研究所 | Method for building three-dimensional vector models of trees based on three-dimensional laser scanning data |
| CN103256895A (en) * | 2013-05-09 | 2013-08-21 | 四川九洲电器集团有限责任公司 | Method for carrying out forestry investigation with three-dimensional laser scanner utilized |
| CN105913016A (en) * | 2016-04-08 | 2016-08-31 | 南京林业大学 | Strip LiDAR data upscaling-based forest biomass estimating method |
| CN106097456A (en) * | 2016-06-06 | 2016-11-09 | 王洪峰 | Oblique photograph outdoor scene three dimensional monolithic model method based on self-adapting cluster algorithm |
| CN106251399A (en) * | 2016-08-30 | 2016-12-21 | 广州市绯影信息科技有限公司 | A kind of outdoor scene three-dimensional rebuilding method based on lsd slam |
| CN106803267A (en) * | 2017-01-10 | 2017-06-06 | 西安电子科技大学 | Indoor scene three-dimensional rebuilding method based on Kinect |
| CN106815850A (en) * | 2017-01-22 | 2017-06-09 | 武汉地普三维科技有限公司 | The method that canopy density forest reserves very high is obtained based on laser radar technique |
| CN107172360A (en) * | 2017-07-06 | 2017-09-15 | 杨顺伟 | Unmanned plane is with shooting method and device |
| CN107179086A (en) * | 2017-05-24 | 2017-09-19 | 北京数字绿土科技有限公司 | A kind of drafting method based on laser radar, apparatus and system |
| CN107292965A (en) * | 2017-08-03 | 2017-10-24 | 北京航空航天大学青岛研究院 | A kind of mutual occlusion processing method based on depth image data stream |
-
2017
- 2017-12-21 CN CN201711391790.0A patent/CN108052914A/en active Pending
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP3865764B1 (en) * | 2006-08-15 | 2007-01-10 | アルスマエヤ株式会社 | Forest resource survey method and forest resource survey apparatus |
| CN101488226A (en) * | 2008-01-16 | 2009-07-22 | 中国科学院自动化研究所 | Tree measurement and reconstruction method based on single three-dimensional laser scanning |
| KR20110093707A (en) * | 2010-02-10 | 2011-08-18 | 대한민국(관리부서 : 산림청 국립산림과학원장) | A method for forest inventory field survey system using gps and rfid |
| CN102881039A (en) * | 2012-07-30 | 2013-01-16 | 中国林业科学研究院资源信息研究所 | Method for building three-dimensional vector models of trees based on three-dimensional laser scanning data |
| CN103256895A (en) * | 2013-05-09 | 2013-08-21 | 四川九洲电器集团有限责任公司 | Method for carrying out forestry investigation with three-dimensional laser scanner utilized |
| CN105913016A (en) * | 2016-04-08 | 2016-08-31 | 南京林业大学 | Strip LiDAR data upscaling-based forest biomass estimating method |
| CN106097456A (en) * | 2016-06-06 | 2016-11-09 | 王洪峰 | Oblique photograph outdoor scene three dimensional monolithic model method based on self-adapting cluster algorithm |
| CN106251399A (en) * | 2016-08-30 | 2016-12-21 | 广州市绯影信息科技有限公司 | A kind of outdoor scene three-dimensional rebuilding method based on lsd slam |
| CN106803267A (en) * | 2017-01-10 | 2017-06-06 | 西安电子科技大学 | Indoor scene three-dimensional rebuilding method based on Kinect |
| CN106815850A (en) * | 2017-01-22 | 2017-06-09 | 武汉地普三维科技有限公司 | The method that canopy density forest reserves very high is obtained based on laser radar technique |
| CN107179086A (en) * | 2017-05-24 | 2017-09-19 | 北京数字绿土科技有限公司 | A kind of drafting method based on laser radar, apparatus and system |
| CN107172360A (en) * | 2017-07-06 | 2017-09-15 | 杨顺伟 | Unmanned plane is with shooting method and device |
| CN107292965A (en) * | 2017-08-03 | 2017-10-24 | 北京航空航天大学青岛研究院 | A kind of mutual occlusion processing method based on depth image data stream |
Non-Patent Citations (3)
| Title |
|---|
| CHUANG QIAN 等: "An Integrated GNSS/INS/LiDAR-SLAM Positioning Method for Highly Accurate Forest Stem Mapping", 《REMOTE SENSING》 * |
| MATTI OHMAN 等: "Tree Measurement and Simultaneous Localization and Mapping System for Forest Harvesters", 《6TH INTERNATIONAL CONFERENCE ON FIELD AND SERVICE ROBOTICS - FSR 2007》 * |
| 李旺 等: "机载激光雷达数据分析与反演青海云杉林结构信息", 《遥感学报》 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020211427A1 (en) * | 2019-04-16 | 2020-10-22 | 广东康云科技有限公司 | Segmentation and recognition method, system, and storage medium based on scanning point cloud data |
| CN110110687A (en) * | 2019-05-15 | 2019-08-09 | 江南大学 | Fruit automatic identifying method on tree based on colouring information and three-D profile information |
| CN110443197A (en) * | 2019-08-05 | 2019-11-12 | 珠海格力电器股份有限公司 | A kind of visual scene intelligent Understanding method and system |
| CN111134112A (en) * | 2019-10-28 | 2020-05-12 | 山东省林木种质资源中心 | Forest germplasm resource field collection method and system |
| CN111179335A (en) * | 2019-12-28 | 2020-05-19 | 东北林业大学 | Standing tree measuring method based on binocular vision |
| CN113723224A (en) * | 2021-08-12 | 2021-11-30 | 新疆爱华盈通信息技术有限公司 | Method, device and medium for remotely and intelligently monitoring tree felling |
| CN115032607A (en) * | 2022-05-26 | 2022-09-09 | 季华实验室 | LiDAR SLAM data-based stumpage position and breast diameter estimation method and system |
| CN116311010A (en) * | 2023-03-06 | 2023-06-23 | 中国科学院空天信息创新研究院 | Method and system for woodland resource investigation and carbon sink metering |
| CN116205394A (en) * | 2023-05-05 | 2023-06-02 | 浙江茂源林业工程有限公司 | Forest resource investigation and monitoring method and system based on radio navigation |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108052914A (en) | A kind of forest forest resource investigation method identified based on SLAM and image | |
| Giannetti et al. | Integrating terrestrial and airborne laser scanning for the assessment of single-tree attributes in Mediterranean forest stands | |
| Gené-Mola et al. | Fruit detection and 3D location using instance segmentation neural networks and structure-from-motion photogrammetry | |
| Koukoulas et al. | Mapping individual tree location, height and species in broadleaved deciduous forest using airborne LIDAR and multi‐spectral remotely sensed data | |
| CN102959354B (en) | Method and apparatus for for analyzing tree canopies with LiDAR data | |
| CN109447169A (en) | The training method of image processing method and its model, device and electronic system | |
| CN105761150B (en) | Crop information and sample acquisition method and system | |
| CN108224895B (en) | Article information input method and device based on deep learning, refrigerator and medium | |
| CN110033484A (en) | Set high extracting method to the high closed forest sample of a kind of combination UAV image and TLS point cloud | |
| CN113435282B (en) | Unmanned aerial vehicle image ear recognition method based on deep learning | |
| CN106528513A (en) | An unmanned aerial vehicle work-based work report generating method and device | |
| CN108256116A (en) | A kind of farming land as-is data Quick Acquisition method | |
| Vastaranta et al. | Forest stand age classification using time series of photogrammetrically derived digital surface models | |
| CN110443862A (en) | Lithologic map filling method and system based on unmanned aerial vehicle and electronic equipment | |
| Vatandaşlar et al. | Extraction of forest inventory parameters using handheld mobile laser scanning: A case study from Trabzon, Turkey | |
| CN109446983A (en) | A kind of coniferous forest felling accumulation evaluation method based on two phase unmanned plane images | |
| Westling et al. | Graph-based methods for analyzing orchard tree structure using noisy point cloud data | |
| CN116071424A (en) | Fruit space coordinate positioning method based on monocular vision | |
| Chiappini et al. | Comparing Mobile Laser Scanner and manual measurements for dendrometric variables estimation in a black pine (Pinus nigra Arn.) plantation | |
| CN109684910A (en) | A kind of method and system of network detection transmission line of electricity ground surface environment variation | |
| CN113532424B (en) | Integrated equipment for acquiring multidimensional information and cooperative measurement method | |
| Fol et al. | Evaluating state-of-the-art 3D scanning methods for stem-level biodiversity inventories in forests | |
| Husin et al. | Study of the oil palm crown characteristics associated with basal stem rot (BSR) disease using stratification method of point cloud data | |
| Saeed et al. | Cotton plant part 3D segmentation and architectural trait extraction using point voxel convolutional neural networks | |
| CN108109125A (en) | Information extracting method and device based on remote sensing images |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180518 |