Table of Content

20 February 2018, Volume 47 Issue 2

Previous Issue    Next Issue

Centroid Automatic Extraction of Spaceborne Laser Spot Image

YUAN Xiaoqi, LI Guoyuan, TANG Xinming, GAO Xiaoming, HUANG Genghua, LI Ye

2018, 47(2):  135-141.  doi:10.11947/j.AGCS.2018.20170517

Asbtract ( )   HTML   PDF (2730KB) ( )  

References | Related Articles | Metrics

The centroid extraction of laser spot images is an important part of satellite laser altimetry data processing,and has significance for obtaining accurate laser pointing angle.This paper summarizes the gray centroid method,Gauss fitting and elliptical fitting method,and compares their advantages and disadvantages.Based on this,a kind of elliptical fitting method combined with the Gaussian threshold is proposed.According to the simulation data test,the accuracy of the method is considerable.Moreover,the real laser spot image-LPA (laser profile array) of GLAS (geoscience laser altimeter system) data is implemented and validated.And the result shows that periodic variation of centroid of LPA is about for 1.5 hours and up to 2~3 pixels,which means the laser pointing angle maybe change 9 arc-seconds and it is necessary to improve the measurement accuracy of laser pointing angle by on-orbit change monitoring of laser centroid position.The conclusions can provide some valuable reference for the footprint images processing of domestic satellite laser altimeter.

Semi-analytical Model of the Waveform of Plantation Target for a Satellite Laser Altimeter

ZHANG Zhiyu, WANG Hong, ZHANG Wenhao, HUANG Ke, ZHOU Hui, MA Yue, LI Song

2018, 47(2):  142-152.  doi:10.11947/j.AGCS.2018.20170488

Asbtract ( )   HTML   PDF (4255KB) ( )  

References | Related Articles | Metrics

The space borne laser altimeter has been widely applied in the extraction of characteristic of plantation target,demonstrating LiDAR technology's huge potential in forestry.As the complexity of the formation of the waveform of plantation target,a waveform simulator basing on semi analytical model has been developed,which performs well in simulating the waveform of certain inputs.Using the GLAS waveforms captured on the area of Daxing'an Mountain,the correlation coefficient R² between the GLAS waveforms and the simulated waveforms is 0.91.The capacity of quantitative control of certain parameters like canopy shape,terrain slope,surface roughness and understory vegetation and producing a large number within limited time,allows it taking advantages in analyzing the influence of each parameter respectfully.Thus,the conclusion can provide guidance for data sources selection in inversion of vegetation target.

Hierarchical Threshold Adaptive for Point Cloud Filter Algorithm of Moving Surface Fitting

ZHU Xiaoxiao, WANG Cheng, XI Xiaohuan, WANG Pu, TIAN Xinguang, YANG Xuebo

2018, 47(2):  153-160.  doi:10.11947/j.AGCS.2018.20170491

Asbtract ( )   HTML   PDF (13413KB) ( )  

References | Related Articles | Metrics

In order to improve the accuracy,efficiency and adaptability of point cloud filtering algorithm,a hierarchical threshold adaptive for point cloud filter algorithm of moving surface fitting was proposed.Firstly,the noisy points are removed by using a statistic histogram method.Secondly,the grid index is established by grid segmentation,and the surface equation is set up through the lowest point among the neighborhood grids.The real height and fit are calculated.The difference between the elevation and the threshold can be determined.Finally,in order to improve the filtering accuracy,hierarchical filtering is used to change the grid size and automatically set the neighborhood size and threshold until the filtering result reaches the accuracy requirement.The test data provided by the International Photogrammetry and Remote Sensing Society (ISPRS) is used to verify the algorithm.The first and second error and the total error are 7.33%,10.64% and 6.34% respectively.The algorithm is compared with the eight classical filtering algorithms published by ISPRS.The experiment results show that the method has well-adapted and it has high accurate filtering result.

Comparison of Signal Extraction Method for Airborne LiDAR Bathymetry Based on Deconvolution

WANG Dandi, XU Qing, XING Shuai, LIN Yuzhun, LI Pengcheng

2018, 47(2):  161-169.  doi:10.11947/j.AGCS.2018.20170501

Asbtract ( )   HTML   PDF (1852KB) ( )  

References | Related Articles | Metrics

To improve the extraction accuracy for airborne LiDAR bathymetry,a signal extraction method based on deconvolution is introduced in waveform processing in this paper.The received waveform is preprocessed by deconvolution,and the accurate positions of the LiDAR signals are determined by the peak detection.For the deconvolution,the validity of four common algorithms,namely,Wiener filter deconvolution,nonnegative least squares,Richardson-Lucy deconvolution and blind deconvolution,are comparatively studied and the performance of the proposed method is assessed by the defined metrics.The experimental results show that the Richardson-Lucy deconvolution can effectively recover the signal resolution with wide adaptation and high success rate.The proposed method compared to the traditional peak detection methods offers a higher detection rate and accuracy and a wider range of bathymetry.

Effects of Different LiDAR Intensity Normalization Methods on Scotch Pine Forest Leaf Area Index Estimation

YOU Haotian, XING Yanqiu, PENG Tao, DING Jianhua

2018, 47(2):  170-179.  doi:10.11947/j.AGCS.2018.20170515

Asbtract ( )   HTML   PDF (1171KB) ( )  

References | Related Articles | Metrics

The intensity data of airborne light detection and ranging (LiDAR) are affected by many factors during the acquisition process. It is of great significance for the normalization and application of LiDAR intensity data to study the effective quantification and normalization of the effect from each factor. In this paper, the LiDAR data were normalized with range, angel of incidence, range and angle of incidence based on radar equation, respectively. Then two metrics, including canopy intensity sum and ratio of intensity, were extracted and used to estimate forest LAI, which was aimed at quantifying the effects of intensity normalization on forest LAI estimation. It was found that the range intensity normalization could improve the accuracy of forest LAI estimation. While the angle of incidence intensity normalization did not improve the accuracy and made the results worse. Although the range and incidence angle normalized intensity data could improve the accuracy, the improvement was less than the result of range intensity normalization. Meanwhile, the differences between the results of forest LAI estimation from raw intensity data and normalized intensity data were relatively big for canopy intensity sum metrics. However, the differences were relatively small for the ratio of intensity metrics. The results demonstrated that the effects of intensity normalization on forest LAI estimation were depended on the choice of affecting factor, and the influential level is closely related to the characteristics of metrics used. Therefore, the appropriate method of intensity normalization should be chosen according to the characteristics of metrics used in the future research, which could avoid the waste of cost and the reduction of estimation accuracy caused by the introduction of inappropriate affecting factors into intensity normalization.

Evaluation of Airborne LiDAR Bathymetric Parameters on the Northern South China Sea Based on MODIS Data

DING Kai, LI Qingquan, ZHU Jiasong, WANG Chisheng, GUAN Minglei, CUI Yang, YANG Chao, XU Tian

2018, 47(2):  180-187.  doi:10.11947/j.AGCS.2018.20170536

Asbtract ( )   HTML   PDF (4924KB) ( )  

References | Related Articles | Metrics

In order to study the spatial distribution of maximum coastal zone mapping and imaging LiDAR(CZMIL)detectable depth in the northern South China Sea,we firstly research the existing K_d(490) inversion algorithm in the northern South China Sea.The relationship between the diffuse attenuation coefficient K_d(490) and K_d(532) is established based on the optical profile data measured,and the relationship between the diffuse attenuation coefficient K_d(532) and the maximum CZMIL detectable depth is summarized.Then,using the remote measurement data of the Aqua-MODIS,we obtain the spatial distribution of diffuse attenuation coefficient at 532 nm in the northern South China Sea in January,June and October,2014.It shows that June is more suitable for bathymetry operation than in October and January.Finally,we obtain the spatial distribution of maximum CZMIL detectable depth in June in the northern South China Sea.The results show that the CZMIL detectable water depths in the northern South China Sea are about 0~71.18 m.The study provides a reference for the time selection and flight scheme of LiDAR bathymetry operation in the northern South China Sea.

Multi-scale Features and Markov Random Field Model for Powerline Scene Classification

YANG Juntao, KANG Zhizhong

2018, 47(2):  188-197.  doi:10.11947/j.AGCS.2018.20170556

Asbtract ( )   HTML   PDF (1595KB) ( )  

References | Related Articles | Metrics

Timely and accurate monitoring the safety of power line can prevent dangerous situations effectively. It is proposed that a Markov random field(MRF) model, into which a random forest classifier being integrated, to classify airborne LiDAR point cloud for power line scene. First, it is extracted that multi-scale visual features according to spatial pyramid theory to represent geometry information of the point and its neighborhood. And then a random forest classifier is used to describe the probability distribution of observed data. Meanwhile, contextual prior probability is established using MRF model, which is formulated as a multi-label energy function. Finally, the multi-label graph-cut technique is used to minimize energy function for optimizing the labels. It is validated the proposed method with LiDAR point cloud acquired by helicopter and mini-UAV power line inspection system. Experimental results demonstrate that the model can effectively classify pylon, power line and vegetation, with the overall accuracy of over 98%. Moreover, compared with other methods, the proposed model shows higher classification accuracy, particularly for the classification of the pylon.

Object Classification Using Airborne Multispectral LiDAR Data

PAN Suoyan, GUAN Haiyan

2018, 47(2):  198-207.  doi:10.11947/j.AGCS.2018.20170512

Asbtract ( )   HTML   PDF (18494KB) ( )  

References | Related Articles | Metrics

Airborne multispectral LiDAR system,which obtains surface geometry and spectral data of objects,simultaneously,has become a fast effective,large-scale spatial data acquisition method.Multispectral LiDAR data are characteristics of completeness and consistency of spectrum and spatial geometric information.Support vector machine (SVM),a machine learning method,is capable of classifying objects based on small samples.Therefore,by means of SVM,this paper performs land cover classification using multispectral LiDAR data. First,all independent point cloud with different wavelengths are merged into a single point cloud,where each pixel contains the three-wavelength spectral information.Next,the merged point cloud is converted into range and intensity images.Finally,land-cover classification is performed by means of SVM.All experiments were conducted on the Optech Titan multispectral LiDAR data,containing three individual point cloud collected by 532 nm,1024 nm,and 1550 nm laser beams.Experimental results demonstrate that ①compared to traditional single-wavelength LiDAR data,multispectral LiDAR data provide a promising solution to land use and land cover applications;②SVM is a feasible method for land cover classification of multispectral LiDAR data.

Calibration of Mounted Parameter for Ship-borne 3D Laser Scanning System

XU Wenxue, TIAN Ziwen, ZHOU Zhimin, ZANG Yufu, GUO Kai, LIU Yanxiong

2018, 47(2):  208-214.  doi:10.11947/j.AGCS.2018.20170505

Asbtract ( )   HTML   PDF (2041KB) ( )  

References | Related Articles | Metrics

Ship-borne 3D laser scanning technology has vital theoretical significance and practical value in marine surveying and mapping.As one of the key steps of ship-borne 3D laser scanning, calibration of mounted parameter is urgent to be solved.This paper proposes a calibration method of mounted parameter without control points for ship-borne 3D laser scanning system.Based on the corresponding points in overlapping area, mounted parameter calibration model of scan system is built by time and spatia registration model of ship-borne 3D laser scanning data.Finally, differential least squares are applied to obtain optimum mounted parameters.Experiments demonstrate the reasonable and effectiveness of this method,the quality of scanning data can be significantly improved.

Automatic Registration of Vehicle-borne Mobile Mapping Laser Point Cloud and Sequent Panoramas

CHEN Chi, YANG Bisheng, TIAN Mao, LI Jianping, ZOU Xianghong, WU Weitong, SONG Yiheng

2018, 47(2):  215-224.  doi:10.11947/j.AGCS.2018.20170520

Asbtract ( )   HTML   PDF (4792KB) ( )  

References | Related Articles | Metrics

An automatic registration method of mobile mapping system laser point cloud and sequence panoramic image is proposed in this paper.Firstly,hierarchical object extraction method is applied on LiDAR data to extract the building façade and outline polygons are generated to construct the skyline vectors.A virtual imaging method is proposed to solve the distortion on panoramas and corners on skylines are further detected on the virtual images combining segmentation and corner detection results.Secondly,the detected skyline vectors are taken as the registration primitives.Registration graphs are built according to the extracted skyline vector and further matched under graph edit distance minimization criteria.The matched conjugate primitives are utilized to solve the 2D-3D rough registration model to obtain the initial transformation between the sequence panoramic image coordinate system and the LiDAR point cloud coordinate system.Finally,to reduce the impact of registration primitives extraction and matching error on the registration results,the optimal transformation between the multi view stereo matching dens point cloud generated from the virtual imaging of the sequent panoramas and the LiDAR point cloud are solved by a 3D-3D ICP registration algorithm variant,thus,refine the exterior orientation parameters of panoramas indirectly.Experiments are undertaken to validate the proposed method and the results show that 1.5 pixel level registration results are achieved on the experiment dataset.The registration results can be applied to point cloud and panoramas fusion applications such as true color point cloud generation.

Façade Solar Potential Analysis Using Multisource Point Cloud

LIANG Fuxun, YANG Bisheng, HUANG Ronggang, DONG Zhen, LI Jianping

2018, 47(2):  225-233.  doi:10.11947/j.AGCS.2018.20170521

Asbtract ( )   HTML   PDF (8506KB) ( )  

References | Related Articles | Metrics

An efficient method for façade solar potential analysis using multisource point cloud is proposed.Firstly,the terrestrial laser scanning data and the UAV images of study area are fused.Then,a ray-based sunlight assessment algorithm suitable for the fused data is proposed.With the simplified solar irradiance model,the sunlight duration and solar potential on façade can be finally estimated.Two buildings with different orientations and types were selected to test the method.The results show that our method is efficient in solar potential estimation of façade,and can be easily applied for useful applications,like sunlight duration test in architecture,and estimation of solar irradiance on windows.

Roadside Multiple Objects Extraction from Mobile Laser Scanning Point Cloud Based on DBN

LUO Haifeng, FANG Lina, CHEN Chongcheng, Huang Zhiwen

2018, 47(2):  234-246.  doi:10.11947/j.AGCS.2018.20170524

Asbtract ( )   HTML   PDF (18677KB) ( )  

References | Related Articles | Metrics

This paper proposed an novel algorithm for exploring deep belief network (DBN) architectures to extract and recognize roadside facilities (trees,cars and traffic poles) from mobile laser scanning (MLS) point cloud.The proposed methods firstly partitioned the raw MLS point cloud into blocks and then removed the ground and building points.In order to partition the off-ground objects into individual objects,off-ground points were organized into an Octree structure and clustered into candidate objects based on connected component.To improve segmentation performance on clusters containing overlapped objects,a refining processing using a voxel-based normalized cut was then implemented.In addition,multi-view features descriptor was generated for each independent roadside facilities based on binary images.Finally,a deep belief network (DBN) was trained to extract trees,cars and traffic pole objects.Experiments are undertaken to evaluate the validities of the proposed method with two datasets acquired by Lynx Mobile Mapper System.The precision of trees,cars and traffic poles objects extraction results respectively was 97.31%,97.79% and 92.78%.The recall was 98.30%,98.75% and 96.77% respectively.The quality is 95.70%,93.81% and 90.00%.And the F1 measure was 97.80%,96.81% and 94.73%.

Point Cloud Information Extraction for Streetlights with Vehicle-borne LiDAR

LI Yongqiang, DONG Yahan, ZHANG Xitong, LI Pengpeng

2018, 47(2):  247-259.  doi:10.11947/j.AGCS.2018.20170527

Asbtract ( )   HTML   PDF (2271KB) ( )  

References | Related Articles | Metrics

The acquisition of detailed information for the streetlights in a large scene remains a tough task since the streetlights are of great number and types. In this paper, a method is proposed to extract and classify the streetlights, with the aid of prior sample sets on the basis of skeleton-line-buffer discriminant algorithm. First, a model and a priori sample set for streetlights are established according to the expression characteristics of streetlights in vehicle-borne LiDAR point cloud. Secondly, with the theory and method of mathematical morphology, the rod-shaped objects are extracted in vehicle LiDAR point cloud scene, and the candidate streetlights are chosen under the constraint of streetlight model and semantic rules. Then, the candidate samples are selected from the sample sets according to the parameter information and the statistical information obtained from the selected streetlights. Finally, based on the matching algorithm of least squares theory, we select and match the priori samples of streetlights and the candidate streetlights. Based on the double buffer of streetlight skeleton information, we discriminate and analyze the candidate streetlights to achieve the extraction and identification of street lights. Finally, the priori samples of streetlights and the point cloud of the candidate streetlights are matched and screened with the matching algorithm of least square theory; and based on the double buffer of streetlight skeleton information, the candidate streetlights are discriminated and analyzed to achieve the extraction and identification of streetlights. Our experiment shows that the algorithm is efficient and robust for the extraction of detailed information of streetlights. For the streetlights with less occlusion and relatively complete data, the extraction accuracy is 0.952, and for those with serous occlusion, low point cloud density and poor data integrity, the extraction accuracy is 0.780. And the above results validate the robustness of the proposed algorithm for the extraction of intermediate streetlights from large scenes. The detailed information extracted by the algorithm can be used to serve the fine and dynamic management of streetlights in large scenes.

Defect Detecting Technology for over 100-meter Shaft

TANG Luliang, ZI Chenbo, LI Qingquan, CHU Xu, LIU Haibo, CHEN Xi, SUN Fei

2018, 47(2):  260-268.  doi:10.11947/j.AGCS.2018.20170522

Asbtract ( )   HTML   PDF (4393KB) ( )  

References | Related Articles | Metrics

Concrete defect detection of diversion shafts in large hydropower station has a direct impact on the safe operation of the station,endangering the lives of cities and millions of people downstream.However,the research on the detection of the 100-meter-level shafts is still a blank in the world.It is a "blind spots" in detection field.In this paper,a method of integrating multiple mapping sensors based on unmanned airship is proposed to resolve the international problem of data acquisition and disease detection in vertical profile the diversion shafts in hydropower stations; We have designed a brand new device to acquire data and detect the disease of the shafts:a cylindrical unmanned airship adapted to the shafts environment is the floating loading platform,on which multi-sensors such as panoramic CCD camera,3D laser scanner,inertial measurement unit,barometric altimeter,illumination module and control module etc is integrated.Trial implemented inside the No.2,No.3 and No.4 shafts of Nuozhadu hydropower station from May 2017 to July,indicates that the detection device can be well applied for the overhaul of similar vertical profile.The detecting device fills the gap of 100-meter-level shaft detection field and presents a wide application prospect.

Bilevel Optimization for Scene Segmentation of LiDAR Point Cloud

LI Minglei, LIU Shaochuang, YANG Huan, QI Chen

2018, 47(2):  269-274.  doi:10.11947/j.AGCS.2018.20170493

Asbtract ( )   HTML   PDF (4656KB) ( )  

References | Related Articles | Metrics

The segmentation of point clouds obtained by light detection and ranging (LiDAR) systems is a critical step for many tasks,such as data organization,reconstruction and information extraction.In this paper,we propose a bilevel progressive optimization algorithm based on the local differentiability.First,we define the topological relation and distance metric of points in the framework of Riemannian geometry,and in the point-based level using k-means method generates over-segmentation results,e.g.super voxels.Then these voxels are formulated as nodes which consist a minimal spanning tree.High level features are extracted from voxel structures,and a graph-based optimization method is designed to yield the final adaptive segmentation results.The implementation experiments on real data demonstrate that our method is efficient and superior to state-of-the-art methods.

On the Consistent Normal Vector Adjustment of Point Cloud Using Surface Variation

HE Hua, LI Zongchun, YAN Rongxin, YANG Zaihua, RUAN Huanli, FU Yongjian

2018, 47(2):  275-280.  doi:10.11947/j.AGCS.2018.20170494

Asbtract ( )   HTML   PDF (5831KB) ( )  

References | Related Articles | Metrics

In order to improve the efficiency and accuracy of existing normal vector adjustment algorithms,a consistent normal vector adjustment algorithm using surface variation is proposed.Firstly,the normal vector and surface variation of point cloud are calculated using principal component analysis.Then,the points on flat or uneven area are distinguished based on surface variation.In the process of adjusting normal vector,the search scope is narrowed to k-nearest neighbors and the number of adjusted normal vector is increased to improve efficiency.The propagating direction of normal vector is restrained to insure accuracy.The experiments show that the proposed algorithm can always receive accurate result on flat region,feature condition and high curvature area,meanwhile,the proposed algorithm is more efficient than the existing algorithms.

Surface Reconstruction Algorithm Based on 3D Delaunay Triangulation

JIA Junhui, HUANG Ming, LIU Xianglei

2018, 47(2):  281-290.  doi:10.11947/j.AGCS.2018.20170490

Asbtract ( )   HTML   PDF (4283KB) ( )  

References | Related Articles | Metrics

With the development of 3D laser scanning technology,the demand for 3D modeling of point cloud data is increasing.As one of the core technologies of 3D modeling,surface reconstruction technology has an extremely widespread application prospect in reverse engineering,computer vision,computer graphics and virtual reality technology.This paper presents a surface reconstruction algorithm based on three-dimensional Delaunay triangulation,it is essentially a greedy algorithm,and combined with the idea of surface region growing algorithm.Manifold surface composed of a selection of explicit triangles can be reconstructed by this algorithm under a certain topological limit,the best triangles are selected from the preconstructed 3D Delaunay triangulation according to the appropriate triangle selection criteria by this algorithm,and added to the growing surface one by one.Compared with the current mainstream implicit surface reconstruction algorithm,this method has the advantages of less parameter dependency and no need to calculate normal line,and it can reconstruct the point cloud model of terrain scanning,building scanning and fine scanning.Using this algorithm to reconstruct the surface of a variety of point cloud models,experimental results show that the quality of the surface generated by the algorithm is choiceness,and the efficiency of reconstruction is faster,this surface reconstruction algorithm has stronger practicability and can be better applied in the field of 3D modeling.