Monterey, CA, USA ABSTRACT 1. INTRODUCTION. phone ; fax ; nps.edu/rsc
|
|
- Blaise Chase
- 6 years ago
- Views:
Transcription
1 Application of image classification techniques to multispectral lidar point cloud data Chad I. Miller* a,b, Judson J. Thomas b, Angela M. Kim b, Jeremy P. Metcalf b, Richard C. Olsen b b SAIC, 1710 SAIC Drive, McLean, VA, USA 22102; b Naval Postgraduate School, 833 Dyer Road, Monterey, CA, USA ABSTRACT Data from Optech Titan are analyzed here for purposes of terrain classification, adding the spectral data component to the lidar point cloud analysis. Nearest-neighbor sorting techniques are used to create the merged point cloud from the three channels. The merged point cloud is analyzed using spectral analysis techniques that allow for the exploitation of color, derived spectral products (pseudo-ndvi), as well as lidar features such as height values, and return number. Standard spectral image classification techniques are used to train a classifier, and analysis was done with a Maximum Likelihood supervised classification. Terrain classification results show an overall accuracy improvement of 10% and a kappa coefficient increase of 0.07 over a raster-based approach. Keywords: lidar, image classification, multispectral lidar, point cloud 1. INTRODUCTION Classic terrain classification with lidar systems has generally been approached primarily from the basis of the topographic character of the data e.g. X, Y, and Z coordinates and their relationships. Somewhat more rarely, such analysis will take into account the intensity of the lidar returns from the monochromatic systems. There has been some work with multi-wavelength systems, such as the two-color bathymetric system from Airborne Hydrography AB, but this remains largely unexplored territory. Teledyne Optech has recently developed a 3-color lidar sensor, the Titan. Data from this system enable a new approach to terrain classification, entering into the domain traditionally held by passive, raster-based spectral imaging. The sensor collects lidar data in three different wavelengths simultaneously. Ensembles of data points are collected in three different bands. There are several approaches to the analysis of such data. It is possible to rasterize the intensity values from each of the Titan channels and combine them to create a multispectral image a 2- dimensional product. This approach gives up some of the fundamental information in the 3D point cloud, however. A different approach is followed here, where the 3D character of the data are preserved in the analysis phase. The technique described here was initially proposed by Thomas [1] to perform terrain classification using the multispectral point cloud rather than the lidar derived multispectral image. The proposed technique is not intended to ignore the spatial component of lidar data, rather determine if and how much the addition of the spectral component to the spatial component improves terrain classification results over a strictly spectral raster-based technique. Maximum Likelihood supervised classification results from a raster-based technique are compared to results from the proposed point cloud technique based on how well they agree with manually classified ground truth. The overall classification accuracy and kappa coefficient from the resulting confusion matrices are used as the comparison metric. *cimiller1@nps.edu; phone ; fax ; nps.edu/rsc Laser Radar Technology and Applications XXI, edited by Monte D. Turner, Gary W. Kamerman, Proc. of SPIE Vol. 9832, 98320X 2016 SPIE CCC code: X/16/$18 doi: / Proc. of SPIE Vol X-1
2 2. DATA 2.1 Sensor The Teledyne Optech Titan Multispectral Lidar System (Figure 1) takes the ability to seamlessly collect topography and bathymetry found in previous dual-wavelength sensors with near-infrared (NIR) and green lasers, and adds a third shortwave infrared (SWIR) laser. The three lasers are not co-aligned. Channel 1 is the SWIR channel operating at a wavelength of 1,550 nanometers (nm) with a 3.5 degree forward tilt. Channel 2 is the NIR channel operating at a wavelength of 1,064 nm with a 0 degree forward tilt. Channel 3 is the green channel operating at a wavelength of 532 nm with a 7 degree forward tilt. Intensities are recorded with a 12 bit dynamic range. um No_. lolu _=1e bite, Figure 1. Teledyne Optech Titan Multispectral Lidar System (from [2]). 2.2 Flight Figure 2 shows the area of the test flight that Optech performed using the Titan system near Toronto, Ontario Canada in October of Three flight lines covered a neighborhood with a mix of residential houses and vegetation. The flight lines extended over Lake Ontario for bathymetry collection. Data were collected at an altitude of 400 meters with a pulse repetition frequency (PRF) of 200 kilohertz (khz) per channel equaling 600 khz for the system resulting in a point density of 12 points per square meter (Table 1). The maximum PRF of the Titan sensor is 300 khz per channel for a total maximum of 900 khz for the system [2]. The data were processed using Optech Lidar Mapping Suite software with default parameters and raw, unaltered intensity values. A total of nine files in LASer (LAS) 1.2 format were provided, one LAS file per channel per flight line. Proc. of SPIE Vol X-2
3 Figure 2. Titan test flight area October Lake Ontario Toronto, Canada. Table 1. Titan test flight parameters. Parameter Value PRF (per channel) 200 KHz Field of view 30º Scan Frequency 40 Hz Altitude 400 m Speed 150 knots Point Spacing (DT/CT) 0.10 m/0.96 m Point Density 12 points/m 2 3. METHODOLOGY The analysis process depended primarily on a set of procedures designed to bring the 3 laser systems together into point collections that were linked together, with each point assigned a nearest-neighbor return from the other two laser systems. 3.1 Pre-Processing Duplicate points (i.e. points with the same X, Y, and Z values) were found within the point cloud files which caused errors with the nearest neighbor processing. The software tool LAStools lasduplicate was used to remove the duplicate points with the unique_xyz flag [3]. A unique point source ID was calculated for each point cloud by appending the set_point_source flag to the lasduplicate tool command specifying the corresponding flight line and channel. Individual channel point clouds were merged per flight line using the LAStools lasmerge tool (Figure 3). Proc. of SPIE Vol X-3
4 Point Source ID { hghtl.ne 1 Channel 1 Flghtline 1 Channel2 Flghtline 1 Channel 3 Flight line 2 Channel 1 Flghtllne 2 Channel 2 Flight line 2 Channel _ Flghtline 3 Channel _ F light line 3 Channel2 F Iightline 3 Channel 3 Figure 3. Titan point cloud with Point Source ID attributed. 3.2 Noise Processing There were problems found because the three Titan lasers are not co-aligned. In particular, there were issues found with points near the edges of the flight lines after merging the channels. Edge points became staggered between channels causing errors in the nearest neighbor calculation. The LAStools lasboundary tool was used to create a flight line boundary shapefile for each channel and an ArcGIS ArcPy script was developed to merge all of the shapefiles together and output a shapefile for each flight line depicting the coincident interior boundary for all three associated channels. The shapefile was then used with the LAStools lasclip tool to classify all edge points. Figure 4 shows the result of manual classification of all high and low noise points, along with the edge points, using Quick Terrain Modeler software. All noise points were removed from further analysis resulting in 39,456,691 total points within the point cloud. Figure 4. Titan point cloud with noise and boundary edge points classified. Proc. of SPIE Vol X-4
5 3.3 Spatial Classification Points were classified as ground and non-ground using the LAStools lasground tool. Height values were calculated by appending the compute_height flag to the lasground tool command. Noise points were removed by appending the keep_classification flag to the lasground tool command. Vegetation and buildings were classified by entering the lasground output into the LAStools lasclassify tool resulting in four classes: Unclassified, Ground, Vegetation, and Buildings (Figure 5). The entire Titan point cloud was manually classified within Quick Terrain Modeler using the following classes: Unclassified, Ground, Vegetation, Buildings, and Water (Figure 6). The manually classified point cloud was used as ground truth due to the lack of actual ground truth of the survey area. Figure 5. Titan point cloud classified using LAStools lasclassify tool. Unclassified Ground Vegetation _ Buildings Water Figure 6. Manually classified Titan point cloud classified using Quick Terrain Modeler software. Proc. of SPIE Vol X-5
6 3.4 Spectral Incorporation Spectral analysis of the Titan point cloud was made possible by finding the complimentary two channel nearest neighbors for each point and attributing the red, green, and blue (RGB) fields with their respective intensities. Channel 1 (1,550 nm) was set as the blue channel, channel 2 (1,064 nm) was set as the red channel, and channel 3 (532 nm) was set as the green channel as shown in Figure 7. The nearest neighbor calculation was implemented within the MATrix LABoratory (MATLAB) [4] environment using the OpenTSTOOL [5] approximate k-nearest neighbor algorithm and lasdata tools were utilized for reading and writing LAS files in MATLAB. The nearest neighbor routine was performed on single and multiple return points separately and merged together. Figure 7. Titan point cloud nearest neighbor RGB attribution. 3.5 Rasterization A lidar derived multispectral image was created from the three Titan channels by averaging the intensities in 0.5 meter 2 pixels using LAStools lasgrid tool with a value of 50 meters for the fill parameter to fill voids in the resulting rasters. The resulting rasters were clipped to the outer flight line boundary extent and the raster bands were integrated into a three band raster using ArcGIS (Figure 8). A raster classification image was created from the manually classified lidar point values by recording the majority of classification values in 0.5 meter 2 pixels using LAStools lasgrid tool (Figure 9). This ground truth raster was used to access the accuracy of the lidar derived multispectral image supervised classification result. Proc. of SPIE Vol X-6
7 Figure 8. Titan derived multispectral image. Red: Channel 2 (1,064 nm). Green: Channel 3 (532 nm). Blue: Channel 1 (1,550 nm). Figure 9. Classification image created by gridding the manually classified Titan point cloud using a 0.5 m 2 pixel area and recording the majority class of points found in each grid. 3.6 ENVI Preparation Spectral analysis of the Titan point cloud was performed within Environment for Visualizing Images (ENVI) software [6]. The problem with applying traditional spectral analysis techniques to multispectral lidar point cloud data within ENVI is that ENVI expects input data to be in traditional imagery formats. To accomplish the appropriate file format conversion, the processed Titan LAS files were converted to American Standard Code for Information Interchange (ASCII) format using the LAStools las2txt tool with the parse xyzirndecauptrgb flag to keep all the attributes. The ASCII files were then read into Interactive Data Language (IDL) variables and reformatted to a modified band sequential (BSQ) interleave (1 x # of points x # of attributes). A random subset of 20% of the points were flagged to be used as training data and vegetation indices were calculated [7] (Table 2). Table 3 is a complete list of the ASCII file attributes showing the supplied LAS 1.2 Data Record Format 1 attributes (1-13) with the User Data attribute (11) calculated as Height and the custom-calculated attributes (14-21) holding much of the spectral information. Finally, a standard ENVI data file and region of interest (ROI) file of the entire Titan point cloud were output for use in the n-dimensional Visualizer tool. Proc. of SPIE Vol X-7
8 Table 2. Vegetation indices calculated. Index Green Normalized Difference Vegetation Index (GNDVI) Green Difference Vegetation Index (GDVI) Green Ratio Vegetation Index (GRVI) Formula GNDVI = (NIR - Green) / (NIR + Green) GDVI = NIR - Green GRVI = NIR / Green Table 3. List of ASCII point cloud attributes. ASCII Field Attribute ASCII Field Attribute 1 X 12 Point Source ID 2 Y 13 GPS Time 3 Z 14 Red 4 Intensity 15 Green 5 Return Number 16 Blue 6 Number of Returns 17 Manual Classification 7 Scan Direction 18 Green Normalized Difference Vegetation Index 8 Edge of Flight Line 19 Green Difference Vegetation Index 9 Classification 20 Green Ratio Vegetation Index 10 Scan Angle Rank 21 Reduction Flag 11 Height 3.7 N-D Visualizer Within ENVI software, the n-dimensional (n-d) Visualizer tool was used to identify endmember spectra and develop ROI training classes from the Titan point cloud for use in the Maximum Likelihood supervised classification. The n-d Visualizer tool allows for the viewing of multi-dimensional scatterplots and drawing of endmember clusters [8]. In the case of the Titan point cloud data, it is possible to view any number and combination of spatial and spectral attributes. For instance, turning on the X, Y, and Z attributes in n-d Visualizer will display the point cloud in typical spatial fashion and allow rotation of the point cloud about the three axes. The X, Y, and Z view was used to help validate endmember determination. By turning on the X attribute and the data reduction flag attribute, it was possible to display points used for training class determination. By turning on the X attribute and the lasclassify classification attribute, the Unclassified, Ground, Vegetation, and Building classes were easily separated. With the Z attribute and the Height attribute turned on, determination of the land/water/bathymetry interface was trivial (Figure 10). With the RGB attributes turned on, it was possible to determine 7 ground sub-classes: Railroad Tracks, Pavement Marking, Grass, Major Streets, Minor Streets, Sidewalks, and Soil and 4 building sub-classes: Roof 1-4. With the addition of the Vegetation and Water classes, 15 total training classes were identified. Proc. of SPIE Vol X-8
9 Figure 10. Two-dimensional scatterplot of the Titan point cloud in the ENVI n-dimensional Visualizer tool using the Z and Height attributes to classify the land/water/bathymetry interface. 3.8 Supervised Classification Maximum Likelihood supervised classification was performed on the Titan derived multispectral raster image (Figure 11) and on the RGB attributes of the entire Titan point cloud (Figure 12) using a probability threshold of 0.5 and the 13 training classes identified using the point cloud within the n-d Visualizer tool. Maximum Likelihood classification is the most common supervised classification method used with remote sensing image data [9] and calculates the probability that a given pixel, or lidar point, belongs to a specific class providing that the probability is above the probability threshold [10]. Training class ROIs used for the raster classification were drawn spatially from the training class areas identified using the point cloud, thus further sub-setting the training class areas used in the raster classification. The 11 sub-classes (7 ground sub-classes and 4 building sub-classes) from the Maximum Likelihood result for both the raster and the point cloud were merged into Ground and Building classes respectively to match the manually classified ground truth classes. Proc. of SPIE Vol X-9
10 Classification Unclassified _ Ground Vegetation _ Buildings Water 4... ' =. Figure 11. Titan derived multispectral raster image Maximum Likelihood supervised classification result. Unclassified Ground Vegetation Buildings Water Figure 12. Titan point cloud Maximum Likelihood supervised classification result. 4. RESULTS The Titan derived multispectral raster image supervised classification result was compared to the manually classified ground truth classification raster image and the Titan point cloud supervised classification result was compared to the manually classified ground truth point cloud using the final merged Unclassified, Ground, Vegetation, Buildings, and Water classes. Table 4 is the confusion matrix comparing the raster result to the raster ground truth and resulted in an overall classification accuracy of 41% and kappa coefficient of Table 5 is the confusion matrix comparing the point cloud result to the point cloud ground truth and resulted in an overall classification accuracy of 51% and kappa coefficient of Proc. of SPIE Vol X-10
11 Table 4. Confusion matrix comparing the Titan derived multispectral raster image supervised classification result to the manually classified ground truth classification raster image. Ground Truth (Percent) Class Unclassified Ground Vegetation Buildings Water Total Unclassified Ground Vegetation Buildings Water Total Table 5. Confusion matrix comparing the Titan point cloud supervised classification result to the manually classified ground truth point cloud. Ground Truth (Percent) Class Unclassified Ground Vegetation Buildings Water Total Unclassified Ground Vegetation Buildings Water Total CONCLUSIONS The point cloud result showed a 10% improvement in overall classification accuracy and a 0.07 increase in the kappa coefficient over the raster result. The kappa coefficient measures the agreement between classification and ground truth pixels, or lidar points. A kappa value of 1 represents perfect agreement while a value of 0 represents no agreement [11]. Therefore, an increase in kappa value shows an improvement in agreement between the point cloud result and the ground truth. Comparison of the class percent totals between the two confusion matrices shows that the raster result performed better with the Ground (+ 7.5%) and Water (+ 4.86%) classes over the point cloud result whereas the point cloud result performed better on the Vegetation (+ 7.35%) and Buildings (+ 4.81%) classes over the raster result. The additional spatial information of the Z dimension provided by the lidar data plays a more important role in discriminating Vegetation and Buildings than the Ground and Water classes. Spectrally, the Vegetation and Water classes proved difficult to discriminate due to the similarity of spectra within those classes owing to the low spectral resolution of the Titan system. Furthermore, the selection of Titan channels is not ideal for discrimination of vegetation. Calculation of vegetation indices using the green Titan channel rather than a commonly used red wavelength to isolate the red edge of vegetation did not aid in sub-classifying the Vegetation class. Lack of actual ground truth of the survey area also complicated the sub-classification of the Vegetation class. Green light is reflected by water and infrared energy is absorbed by water causing spectrally flat, vegetation-like water signatures and classifier confusion between the Vegetation and Water classes. Proc. of SPIE Vol X-11
12 This study presents a novel method to apply traditional spectral analysis techniques to a multispectral lidar point cloud and shows that the technique can improve terrain classification results over a standard raster technique. Future work on radiometric correction of Titan intensities and actual ground truth collection is envisioned which may further improve terrain classification results using the proposed method. REFERENCES [1] Thomas, J., Terrain Classification Using Multi-Wavelength LiDAR Data, (master s thesis, Naval Postgraduate School, 2015). Retrieved from Calhoun [2] Optech Titan Multispectral LiDAR System: High Precision Environmental Mapping, Teledyne Optech, n.d., accessed online at on 8 April [3] Isenburg, M., LAStools efficient tools for LiDAR processing (version , academic), obtained from [4] MATLAB and Statistics Toolbox Release 2015b, The MathWorks, Inc., Natick, Massachusetts, United States. [5] Merkwirth, C., Parlitz, U., Wedekind, I., Engster, D., and Lauterborn, W., OpenTSTOOL User Manual, [6] ENVI version IDL 8.5 (64-bit) classic mode (Harris Geospatial Solutions, Boulder, Colorado). [7] Broadband Greeness, Harris Geospatial Solutions, n.d., accessed online at on 4 April [8] The n-d Visualizer, Harris Geospatial Solutions, n.d., accessed online at on 30 March [9] Richards, J. A., [Remote Sensing Digital Image Analysis: An Introduction], Springer, Berlin, Germany, 194 (2006). [10] Maximum Likelihood, Harris Geospatial Solutions, n.d., accessed online at on 30 March [11] Calculate Confusion Matrices, Harris Geospatial Solutions, n.d., accessed online at on 30 March Proc. of SPIE Vol X-12
ABSTRACT 1. INTRODUCTION 2. DATA
Spectral LiDAR Analysis for Terrain Classification Charles A. McIver, Jeremy P. Metcalf, Richard C. Olsen* Naval Postgraduate School, 833 Dyer Road, Monterey, CA, USA 93943 ABSTRACT Data from the Optech
More informationMethods for LiDAR point cloud classification using local neighborhood statistics
Methods for LiDAR point cloud classification using local neighborhood statistics Angela M. Kim, Richard C. Olsen, Fred A. Kruse Naval Postgraduate School, Remote Sensing Center and Physics Department,
More informationABSTRACT 1. INTRODUCTION
Correlation between lidar-derived intensity and passive optical imagery Jeremy P. Metcalf, Angela M. Kim, Fred A. Kruse, and Richard C. Olsen Physics Department and Remote Sensing Center, Naval Postgraduate
More informationTerrain categorization using LIDAR and multi-spectral data
Terrain categorization using LIDAR and multi-spectral data Angela M. Puetz, R. C. Olsen, Michael A. Helt U.S. Naval Postgraduate School, 833 Dyer Road, Monterey, CA 93943 ampuetz@nps.edu, olsen@nps.edu
More informationAPPLICATION OF SOFTMAX REGRESSION AND ITS VALIDATION FOR SPECTRAL-BASED LAND COVER MAPPING
APPLICATION OF SOFTMAX REGRESSION AND ITS VALIDATION FOR SPECTRAL-BASED LAND COVER MAPPING J. Wolfe a, X. Jin a, T. Bahr b, N. Holzer b, * a Harris Corporation, Broomfield, Colorado, U.S.A. (jwolfe05,
More informationIMPROVED TARGET DETECTION IN URBAN AREA USING COMBINED LIDAR AND APEX DATA
IMPROVED TARGET DETECTION IN URBAN AREA USING COMBINED LIDAR AND APEX DATA Michal Shimoni 1 and Koen Meuleman 2 1 Signal and Image Centre, Dept. of Electrical Engineering (SIC-RMA), Belgium; 2 Flemish
More informationDetecting trails in LiDAR point cloud data
Detecting trails in LiDAR point cloud data Angela M. Kim and Richard C. Olsen Naval Postgraduate School, Remote Sensing Center, 1 University Circle, Monterey, CA, USA ABSTRACT The goal of this work is
More informationAn Introduction to Lidar & Forestry May 2013
An Introduction to Lidar & Forestry May 2013 Introduction to Lidar & Forestry Lidar technology Derivatives from point clouds Applied to forestry Publish & Share Futures Lidar Light Detection And Ranging
More informationFigure 1: Workflow of object-based classification
Technical Specifications Object Analyst Object Analyst is an add-on package for Geomatica that provides tools for segmentation, classification, and feature extraction. Object Analyst includes an all-in-one
More informationClassify Multi-Spectral Data Classify Geologic Terrains on Venus Apply Multi-Variate Statistics
Classify Multi-Spectral Data Classify Geologic Terrains on Venus Apply Multi-Variate Statistics Operations What Do I Need? Classify Merge Combine Cross Scan Score Warp Respace Cover Subscene Rotate Translators
More informationAPPENDIX E2. Vernal Pool Watershed Mapping
APPENDIX E2 Vernal Pool Watershed Mapping MEMORANDUM To: U.S. Fish and Wildlife Service From: Tyler Friesen, Dudek Subject: SSHCP Vernal Pool Watershed Analysis Using LIDAR Data Date: February 6, 2014
More informationIntegrated analysis of Light Detection and Ranging (LiDAR) and Hyperspectral Imagery (HSI) data
Integrated analysis of Light Detection and Ranging (LiDAR) and Hyperspectral Imagery (HSI) data Angela M. Kim, Fred A. Kruse, and Richard C. Olsen Naval Postgraduate School, Remote Sensing Center and Physics
More informationGround LiDAR fuel measurements of the Prescribed Fire Combustion and Atmospheric Dynamics Research Experiment
Ground LiDAR fuel measurements of the Prescribed Fire Combustion and Atmospheric Dynamics Research Experiment Eric Rowell, Erik Apland and Carl Seielstad IAWF 4 th Fire Behavior and Fuels Conference, Raleigh,
More informationCrop Counting and Metrics Tutorial
Crop Counting and Metrics Tutorial The ENVI Crop Science platform contains remote sensing analytic tools for precision agriculture and agronomy. In this tutorial you will go through a typical workflow
More informationHYPERSPECTRAL REMOTE SENSING
HYPERSPECTRAL REMOTE SENSING By Samuel Rosario Overview The Electromagnetic Spectrum Radiation Types MSI vs HIS Sensors Applications Image Analysis Software Feature Extraction Information Extraction 1
More informationCLASSIFICATION OF NONPHOTOGRAPHIC REMOTE SENSORS
CLASSIFICATION OF NONPHOTOGRAPHIC REMOTE SENSORS PASSIVE ACTIVE DIGITAL CAMERA THERMAL (e.g. TIMS) VIDEO CAMERA MULTI- SPECTRAL SCANNERS VISIBLE & NIR MICROWAVE HYPERSPECTRAL (e.g. AVIRIS) SLAR Real Aperture
More informationFiles Used in This Tutorial. Background. Feature Extraction with Example-Based Classification Tutorial
Feature Extraction with Example-Based Classification Tutorial In this tutorial, you will use Feature Extraction to extract rooftops from a multispectral QuickBird scene of a residential area in Boulder,
More informationRaster Classification with ArcGIS Desktop. Rebecca Richman Andy Shoemaker
Raster Classification with ArcGIS Desktop Rebecca Richman Andy Shoemaker Raster Classification What is it? - Classifying imagery into different land use/ land cover classes based on the pixel values of
More informationSIMULATED LIDAR WAVEFORMS FOR THE ANALYSIS OF LIGHT PROPAGATION THROUGH A TREE CANOPY
SIMULATED LIDAR WAVEFORMS FOR THE ANALYSIS OF LIGHT PROPAGATION THROUGH A TREE CANOPY Angela M. Kim and Richard C. Olsen Remote Sensing Center Naval Postgraduate School 1 University Circle Monterey, CA
More informationAutomated Extraction of Buildings from Aerial LiDAR Point Cloud and Digital Imaging Datasets for 3D Cadastre - Preliminary Results
Automated Extraction of Buildings from Aerial LiDAR Point Cloud and Digital Imaging Datasets for 3D Pankaj Kumar 1*, Alias Abdul Rahman 1 and Gurcan Buyuksalih 2 ¹Department of Geoinformation Universiti
More informationA Method to Create a Single Photon LiDAR based Hydro-flattened DEM
A Method to Create a Single Photon LiDAR based Hydro-flattened DEM Sagar Deshpande 1 and Alper Yilmaz 2 1 Surveying Engineering, Ferris State University 2 Department of Civil, Environmental, and Geodetic
More informationLab 9. Julia Janicki. Introduction
Lab 9 Julia Janicki Introduction My goal for this project is to map a general land cover in the area of Alexandria in Egypt using supervised classification, specifically the Maximum Likelihood and Support
More informationHydrocarbon Index an algorithm for hyperspectral detection of hydrocarbons
INT. J. REMOTE SENSING, 20 JUNE, 2004, VOL. 25, NO. 12, 2467 2473 Hydrocarbon Index an algorithm for hyperspectral detection of hydrocarbons F. KÜHN*, K. OPPERMANN and B. HÖRIG Federal Institute for Geosciences
More informationDefining Remote Sensing
Defining Remote Sensing Remote Sensing is a technology for sampling electromagnetic radiation to acquire and interpret non-immediate geospatial data from which to extract information about features, objects,
More informationAirborne Laser Scanning: Remote Sensing with LiDAR
Airborne Laser Scanning: Remote Sensing with LiDAR ALS / LIDAR OUTLINE Laser remote sensing background Basic components of an ALS/LIDAR system Two distinct families of ALS systems Waveform Discrete Return
More informationMULTISPECTRAL MAPPING
VOLUME 5 ISSUE 1 JAN/FEB 2015 MULTISPECTRAL MAPPING 8 DRONE TECH REVOLUTION Forthcoming order of magnitude reduction in the price of close-range aerial scanning 16 HANDHELD SCANNING TECH 32 MAX MATERIAL,
More informationAlaska Department of Transportation Roads to Resources Project LiDAR & Imagery Quality Assurance Report Juneau Access South Corridor
Alaska Department of Transportation Roads to Resources Project LiDAR & Imagery Quality Assurance Report Juneau Access South Corridor Written by Rick Guritz Alaska Satellite Facility Nov. 24, 2015 Contents
More informationENVI Classic Tutorial: Multispectral Analysis of MASTER HDF Data 2
ENVI Classic Tutorial: Multispectral Analysis of MASTER HDF Data Multispectral Analysis of MASTER HDF Data 2 Files Used in This Tutorial 2 Background 2 Shortwave Infrared (SWIR) Analysis 3 Opening the
More informationGEOBIA for ArcGIS (presentation) Jacek Urbanski
GEOBIA for ArcGIS (presentation) Jacek Urbanski INTEGRATION OF GEOBIA WITH GIS FOR SEMI-AUTOMATIC LAND COVER MAPPING FROM LANDSAT 8 IMAGERY Presented at 5th GEOBIA conference 21 24 May in Thessaloniki.
More informationSpatial Density Distribution
GeoCue Group Support Team 5/28/2015 Quality control and quality assurance checks for LIDAR data continue to evolve as the industry identifies new ways to help ensure that data collections meet desired
More informationMunicipal Projects in Cambridge Using a LiDAR Dataset. NEURISA Day 2012 Sturbridge, MA
Municipal Projects in Cambridge Using a LiDAR Dataset NEURISA Day 2012 Sturbridge, MA October 15, 2012 Jeff Amero, GIS Manager, City of Cambridge Presentation Overview Background on the LiDAR dataset Solar
More informationCOMBINING HIGH SPATIAL RESOLUTION OPTICAL AND LIDAR DATA FOR OBJECT-BASED IMAGE CLASSIFICATION
COMBINING HIGH SPATIAL RESOLUTION OPTICAL AND LIDAR DATA FOR OBJECT-BASED IMAGE CLASSIFICATION Ruonan Li 1, Tianyi Zhang 1, Ruozheng Geng 1, Leiguang Wang 2, * 1 School of Forestry, Southwest Forestry
More informationABSTRACT 1. INTRODUCTION 2. OBSERVATIONS
Visualization and analysis of LiDAR waveform data Richard C. Olsen, Jeremy P. Metcalf, Remote Sensing Center, Naval Postgraduate School, Monterey, CA 93943 ABSTRACT LiDAR waveform analysis is a relatively
More informationFOOTPRINTS EXTRACTION
Building Footprints Extraction of Dense Residential Areas from LiDAR data KyoHyouk Kim and Jie Shan Purdue University School of Civil Engineering 550 Stadium Mall Drive West Lafayette, IN 47907, USA {kim458,
More informationTools, Tips and Workflows Geiger-Mode LIDAR Workflow Review GeoCue, TerraScan, versions and above
GeoCue, TerraScan, versions 015.005 and above Martin Flood August 8, 2016 Geiger-mode lidar data is getting a lot of press lately as the next big thing in airborne data collection. Unlike traditional lidar
More informationReality Check: Processing LiDAR Data. A story of data, more data and some more data
Reality Check: Processing LiDAR Data A story of data, more data and some more data Red River of the North Red River of the North Red River of the North Red River of the North Introduction and Background
More informationLand Cover Classification Techniques
Land Cover Classification Techniques supervised classification and random forests Developed by remote sensing specialists at the USFS Geospatial Technology and Applications Center (GTAC), located in Salt
More informationAerial and Mobile LiDAR Data Fusion
Creating Value Delivering Solutions Aerial and Mobile LiDAR Data Fusion Dr. Srini Dharmapuri, CP, PMP What You Will Learn About LiDAR Fusion Mobile and Aerial LiDAR Technology Components & Parameters Project
More informationThe 2014 IEEE GRSS Data Fusion Contest
The 2014 IEEE GRSS Data Fusion Contest Description of the datasets Presented to Image Analysis and Data Fusion Technical Committee IEEE Geoscience and Remote Sensing Society (GRSS) February 17 th, 2014
More information[Youn *, 5(11): November 2018] ISSN DOI /zenodo Impact Factor
GLOBAL JOURNAL OF ENGINEERING SCIENCE AND RESEARCHES AUTOMATIC EXTRACTING DEM FROM DSM WITH CONSECUTIVE MORPHOLOGICAL FILTERING Junhee Youn *1 & Tae-Hoon Kim 2 *1,2 Korea Institute of Civil Engineering
More informationRemote Sensing Introduction to the course
Remote Sensing Introduction to the course Remote Sensing (Prof. L. Biagi) Exploitation of remotely assessed data for information retrieval Data: Digital images of the Earth, obtained by sensors recording
More informationAardobservatie en Data-analyse Image processing
Aardobservatie en Data-analyse Image processing 1 Image processing: Processing of digital images aiming at: - image correction (geometry, dropped lines, etc) - image calibration: DN into radiance or into
More informationA DATA DRIVEN METHOD FOR FLAT ROOF BUILDING RECONSTRUCTION FROM LiDAR POINT CLOUDS
A DATA DRIVEN METHOD FOR FLAT ROOF BUILDING RECONSTRUCTION FROM LiDAR POINT CLOUDS A. Mahphood, H. Arefi *, School of Surveying and Geospatial Engineering, College of Engineering, University of Tehran,
More informationSpectral Classification
Spectral Classification Spectral Classification Supervised versus Unsupervised Classification n Unsupervised Classes are determined by the computer. Also referred to as clustering n Supervised Classes
More informationTerrestrial GPS setup Fundamentals of Airborne LiDAR Systems, Collection and Calibration. JAMIE YOUNG Senior Manager LiDAR Solutions
Terrestrial GPS setup Fundamentals of Airborne LiDAR Systems, Collection and Calibration JAMIE YOUNG Senior Manager LiDAR Solutions Topics Terrestrial GPS reference Planning and Collection Considerations
More informationLocating the Shadow Regions in LiDAR Data: Results on the SHARE 2012 Dataset
Locating the Shadow Regions in LiDAR Data: Results on the SHARE 22 Dataset Mustafa BOYACI, Seniha Esen YUKSEL* Hacettepe University, Department of Electrical and Electronics Engineering Beytepe, Ankara,
More informationN.J.P.L.S. An Introduction to LiDAR Concepts and Applications
N.J.P.L.S. An Introduction to LiDAR Concepts and Applications Presentation Outline LIDAR Data Capture Advantages of Lidar Technology Basics Intensity and Multiple Returns Lidar Accuracy Airborne Laser
More informationIntroduction to digital image classification
Introduction to digital image classification Dr. Norman Kerle, Wan Bakx MSc a.o. INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION Purpose of lecture Main lecture topics Review
More informationHAWAII KAUAI Survey Report. LIDAR System Description and Specifications
HAWAII KAUAI Survey Report LIDAR System Description and Specifications This survey used an Optech GEMINI Airborne Laser Terrain Mapper (ALTM) serial number 06SEN195 mounted in a twin-engine Navajo Piper
More informationTopographic Lidar Data Employed to Map, Preserve U.S. History
OCTOBER 11, 2016 Topographic Lidar Data Employed to Map, Preserve U.S. History In August 2015, the National Park Service (NPS) contracted Woolpert for the Little Bighorn National Monument Mapping Project
More informationTools, Tips and Workflows Colorized Point Clouds take on Role of 3D Image
l Karrie-Sue Simmers, Darrick Wagg 1-15-2013 Revision 1.0 In a visual world an image is easily recognizable to viewers when compared to a monochrome point cloud and can be one of the biggest challenges
More informationDIGITAL IMAGE ANALYSIS. Image Classification: Object-based Classification
DIGITAL IMAGE ANALYSIS Image Classification: Object-based Classification Image classification Quantitative analysis used to automate the identification of features Spectral pattern recognition Unsupervised
More informationIntroduction to Remote Sensing
Introduction to Remote Sensing Spatial, spectral, temporal resolutions Image display alternatives Vegetation Indices Image classifications Image change detections Accuracy assessment Satellites & Air-Photos
More informationSubmerged Aquatic Vegetation Mapping using Object-Based Image Analysis with Lidar and RGB Imagery
Submerged Aquatic Vegetation Mapping using Object-Based Image Analysis with Lidar and RGB Imagery Victoria Price Version 1, April 16 2015 Submerged Aquatic Vegetation Mapping using Object-Based Image Analysis
More informationMultisensoral UAV-Based Reference Measurements for Forestry Applications
Multisensoral UAV-Based Reference Measurements for Forestry Applications Research Manager D.Sc. Anttoni Jaakkola Centre of Excellence in Laser Scanning Research 2 Outline UAV applications Reference level
More informationLiDAR Data Processing:
LiDAR Data Processing: Concepts and Methods for LEFI Production Gordon W. Frazer GWF LiDAR Analytics Outline of Presentation Data pre-processing Data quality checking and options for repair Data post-processing
More informationAirborne LiDAR Data Acquisition for Forestry Applications. Mischa Hey WSI (Corvallis, OR)
Airborne LiDAR Data Acquisition for Forestry Applications Mischa Hey WSI (Corvallis, OR) WSI Services Corvallis, OR Airborne Mapping: Light Detection and Ranging (LiDAR) Thermal Infrared Imagery 4-Band
More informationENVI. Get the Information You Need from Imagery.
Visual Information Solutions ENVI. Get the Information You Need from Imagery. ENVI is the premier software solution to quickly, easily, and accurately extract information from geospatial imagery. Easy
More informationHigh Resolution Laserscanning, not only for 3D-City Models
Lohr 133 High Resolution Laserscanning, not only for 3D-City Models UWE LOHR, Ravensburg ABSTRACT The TopoSys laserscanner system is designed to produce digital elevation models (DEMs) of the environment
More informationThe Reference Library Generating Low Confidence Polygons
GeoCue Support Team In the new ASPRS Positional Accuracy Standards for Digital Geospatial Data, low confidence areas within LIDAR data are defined to be where the bare earth model might not meet the overall
More informationINCREASING CLASSIFICATION QUALITY BY USING FUZZY LOGIC
JOURNAL OF APPLIED ENGINEERING SCIENCES VOL. 1(14), issue 4_2011 ISSN 2247-3769 ISSN-L 2247-3769 (Print) / e-issn:2284-7197 INCREASING CLASSIFICATION QUALITY BY USING FUZZY LOGIC DROJ Gabriela, University
More informationOverview. Image Geometric Correction. LA502 Special Studies Remote Sensing. Why Geometric Correction?
LA502 Special Studies Remote Sensing Image Geometric Correction Department of Landscape Architecture Faculty of Environmental Design King AbdulAziz University Room 103 Overview Image rectification Geometric
More informationHyperspectral Remote Sensing
Hyperspectral Remote Sensing Multi-spectral: Several comparatively wide spectral bands Hyperspectral: Many (could be hundreds) very narrow spectral bands GEOG 4110/5100 30 AVIRIS: Airborne Visible/Infrared
More informationTHE USE OF ANISOTROPIC HEIGHT TEXTURE MEASURES FOR THE SEGMENTATION OF AIRBORNE LASER SCANNER DATA
THE USE OF ANISOTROPIC HEIGHT TEXTURE MEASURES FOR THE SEGMENTATION OF AIRBORNE LASER SCANNER DATA Sander Oude Elberink* and Hans-Gerd Maas** *Faculty of Civil Engineering and Geosciences Department of
More informationVisual Information Solutions. E3De. The interactive software environment for extracting 3D information from LiDAR data.
Visual Information Solutions E3De. The interactive software environment for extracting 3D information from LiDAR data. Photorealistic Visualizations. 3D Feature Extraction. Versatile Geospatial Products.
More informationGeospatial Computer Vision Based on Multi-Modal Data How Valuable Is Shape Information for the Extraction of Semantic Information?
remote sensing Article Geospatial Computer Vision Based on Multi-Modal Data How Valuable Is Shape Information for the Extraction of Semantic Information? Martin Weinmann 1, * and Michael Weinmann 2 1 Institute
More informationBy Colin Childs, ESRI Education Services. Catalog
s resolve many traditional raster management issues By Colin Childs, ESRI Education Services Source images ArcGIS 10 introduces Catalog Mosaicked images Sources, mosaic methods, and functions are used
More informationTerraScan New Features
www.terrasolid.com TerraScan New Features Arttu Soininen 03.02.2016 32 & Various Improvements Compute normal vectors action on project also without trajectory information Multiple source classes in Classify
More informationLeica - Airborne Digital Sensors (ADS80, ALS60) Update / News in the context of Remote Sensing applications
Luzern, Switzerland, acquired with GSD=5 cm, 2008. Leica - Airborne Digital Sensors (ADS80, ALS60) Update / News in the context of Remote Sensing applications Arthur Rohrbach, Sensor Sales Dir Europe,
More informationThe Gain setting for Landsat 7 (High or Low Gain) depends on: Sensor Calibration - Application. the surface cover types of the earth and the sun angle
Sensor Calibration - Application Station Identifier ASN Scene Center atitude 34.840 (34 3'0.64"N) Day Night DAY Scene Center ongitude 33.03270 (33 0'7.72"E) WRS Path WRS Row 76 036 Corner Upper eft atitude
More informationSeabed Mapping with LiDAR
Seabed Mapping with LiDAR 2011 Jakarta David Jonas Lt Cdr Rupert Forester-Bennett RN (ret( ret d) October 18 th 2011 Mapping in South East Asia Field Survey Aerial Photography LiDAR Pleased to Introduce
More informationRemote Sensing Sensor Integration
Remote Sensing Sensor Integration Erica Tharp LiDAR Supervisor Table of Contents About 3001 International Inc Remote Sensing Platforms Why Sensor Integration? Technical Aspects of Sensor Integration Limitations
More informationQuick Start. Color wheel (8 bit BSQ, 440 pixels by 290 lines by 3 bands), 373 KB (binary).
The various sample data files after expansion (use Zip): Quick Start Cc Cc.hdr Cc.wav URBAN URBAN.hdr URBAN.wav TERRAIN TERRAIN.hdr Color wheel (8 bit BSQ, 440 pixels by 290 lines by 3 bands), 373 KB (binary).
More informationEVOLUTION OF POINT CLOUD
Figure 1: Left and right images of a stereo pair and the disparity map (right) showing the differences of each pixel in the right and left image. (source: https://stackoverflow.com/questions/17607312/difference-between-disparity-map-and-disparity-image-in-stereo-matching)
More informationSYNERGY BETWEEN AERIAL IMAGERY AND LOW DENSITY POINT CLOUD FOR AUTOMATED IMAGE CLASSIFICATION AND POINT CLOUD DENSIFICATION
SYNERGY BETWEEN AERIAL IMAGERY AND LOW DENSITY POINT CLOUD FOR AUTOMATED IMAGE CLASSIFICATION AND POINT CLOUD DENSIFICATION Hani Mohammed Badawy a,*, Adel Moussa a,b, Naser El-Sheimy a a Dept. of Geomatics
More informationAutomatic Segmentation of Semantic Classes in Raster Map Images
Automatic Segmentation of Semantic Classes in Raster Map Images Thomas C. Henderson, Trevor Linton, Sergey Potupchik and Andrei Ostanin School of Computing, University of Utah, Salt Lake City, UT 84112
More informationPresented at the FIG Congress 2018, May 6-11, 2018 in Istanbul, Turkey
Presented at the FIG Congress 2018, May 6-11, 2018 in Istanbul, Turkey Evangelos MALTEZOS, Charalabos IOANNIDIS, Anastasios DOULAMIS and Nikolaos DOULAMIS Laboratory of Photogrammetry, School of Rural
More informationMATLAB Tools for LIDAR Data Conversion, Visualization, and Processing
MATLAB Tools for LIDAR Data Conversion, Visualization, and Processing Xiao Wang a, Kaijing Zhou a, Jie Yang a, Yilong Lu *a a Nanyang Technological University, 50 Nanyang Avenue, Singapore 639798 ABSTRACT
More informationUnsupervised and Self-taught Learning for Remote Sensing Image Analysis
Unsupervised and Self-taught Learning for Remote Sensing Image Analysis Ribana Roscher Institute of Geodesy and Geoinformation, Remote Sensing Group, University of Bonn 1 The Changing Earth https://earthengine.google.com/timelapse/
More informationLecture 06. Raster and Vector Data Models. Part (1) Common Data Models. Raster. Vector. Points. Points. ( x,y ) Area. Area Line.
Lecture 06 Raster and Vector Data Models Part (1) 1 Common Data Models Vector Raster Y Points Points ( x,y ) Line Area Line Area 2 X 1 3 Raster uses a grid cell structure Vector is more like a drawn map
More informationGeometric Rectification of Remote Sensing Images
Geometric Rectification of Remote Sensing Images Airborne TerrestriaL Applications Sensor (ATLAS) Nine flight paths were recorded over the city of Providence. 1 True color ATLAS image (bands 4, 2, 1 in
More informationBRIEF EXAMPLES OF PRACTICAL USES OF LIDAR
BRIEF EXAMPLES OF PRACTICAL USES OF LIDAR PURDUE ROAD SCHOOL - 3/9/2016 CHRIS MORSE USDA-NRCS, STATE GIS COORDINATOR LIDAR/DEM SOURCE DATES LiDAR and its derivatives (DEMs) have a collection date for data
More informationDigital Photogrammetric System. Version 6.3 USER MANUAL. LIDAR Data processing
Digital Photogrammetric System Version 6.3 USER MANUAL Table of Contents 1. About... 3 2. Import of LIDAR data... 3 3. Load LIDAR data window... 4 4. LIDAR data loading and displaying... 6 5. Splitting
More informationFiles Used in this Tutorial
Generate Point Clouds and DSM Tutorial This tutorial shows how to generate point clouds and a digital surface model (DSM) from IKONOS satellite stereo imagery. You will view the resulting point clouds
More information2. POINT CLOUD DATA PROCESSING
Point Cloud Generation from suas-mounted iphone Imagery: Performance Analysis A. D. Ladai, J. Miller Towill, Inc., 2300 Clayton Road, Suite 1200, Concord, CA 94520-2176, USA - (andras.ladai, jeffrey.miller)@towill.com
More informationLIDAR MAPPING FACT SHEET
1. LIDAR THEORY What is lidar? Lidar is an acronym for light detection and ranging. In the mapping industry, this term is used to describe an airborne laser profiling system that produces location and
More information2014 Google Earth Engine Research Award Report
2014 Google Earth Engine Research Award Report Title: Mapping Invasive Vegetation using Hyperspectral Data, Spectral Angle Mapping, and Mixture Tuned Matched Filtering Section I: Section II: Section III:
More informationLight Detection and Ranging (LiDAR)
Light Detection and Ranging (LiDAR) http://code.google.com/creative/radiohead/ Types of aerial sensors passive active 1 Active sensors for mapping terrain Radar transmits microwaves in pulses determines
More informationOutline of Presentation. Introduction to Overwatch Geospatial Software Feature Analyst and LIDAR Analyst Software
Outline of Presentation Automated Feature Extraction from Terrestrial and Airborne LIDAR Presented By: Stuart Blundell Overwatch Geospatial - VLS Ops Co-Author: David W. Opitz Overwatch Geospatial - VLS
More informationADVANCED TERRAIN PROCESSING: ANALYTICAL RESULTS OF FILLING VOIDS IN REMOTELY SENSED DATA TERRAIN INPAINTING
ADVANCED TERRAIN PROCESSING: ANALYTICAL RESULTS OF FILLING VOIDS IN REMOTELY SENSED DATA J. Harlan Yates Patrick Kelley Josef Allen Mark Rahmes Harris Corporation Government Communications Systems Division
More informationIncluding the Size of Regions in Image Segmentation by Region Based Graph
International Journal of Emerging Engineering Research and Technology Volume 3, Issue 4, April 2015, PP 81-85 ISSN 2349-4395 (Print) & ISSN 2349-4409 (Online) Including the Size of Regions in Image Segmentation
More informationClassification of urban feature from unmanned aerial vehicle images using GASVM integration and multi-scale segmentation
Classification of urban feature from unmanned aerial vehicle images using GASVM integration and multi-scale segmentation M.Modiri, A.Salehabadi, M.Mohebbi, A.M.Hashemi, M.Masumi National Geographical Organization
More informationData: a collection of numbers or facts that require further processing before they are meaningful
Digital Image Classification Data vs. Information Data: a collection of numbers or facts that require further processing before they are meaningful Information: Derived knowledge from raw data. Something
More informationLiDAR Technical Report NE Washington LiDAR Production 2017
LiDAR Technical Report NE Washington LiDAR Production 2017 Presented to: Washington DNR 1111 Washington Street SE Olympia, Washington 98504 Submitted by: 860 McKinley St Eugene, OR 97402 July 26, 2017
More informationMARS v Release Notes Revised: May 23, 2018 (Builds and )
MARS v2018.0 Release Notes Revised: May 23, 2018 (Builds 8302.01 8302.18 and 8350.00 8352.00) Contents New Features:... 2 Enhancements:... 6 List of Bug Fixes... 13 1 New Features: LAS Up-Conversion prompts
More informationE3De. E3De Discover the Next Dimension of Your Data.
International Support Exelis Visual Information Solutions is a global company with direct offices in North America, Europe, and Asia. Combined with our extensive, worldwide distributor network, we can
More informationUTILIZACIÓN DE DATOS LIDAR Y SU INTEGRACIÓN CON SISTEMAS DE INFORMACIÓN GEOGRÁFICA
UTILIZACIÓN DE DATOS LIDAR Y SU INTEGRACIÓN CON SISTEMAS DE INFORMACIÓN GEOGRÁFICA Aurelio Castro Cesar Piovanetti Geographic Mapping Technologies Corp. (GMT) Consultores en GIS info@gmtgis.com Geographic
More informationIntroducing ArcScan for ArcGIS
Introducing ArcScan for ArcGIS An ESRI White Paper August 2003 ESRI 380 New York St., Redlands, CA 92373-8100, USA TEL 909-793-2853 FAX 909-793-5953 E-MAIL info@esri.com WEB www.esri.com Copyright 2003
More informationGenerating passive NIR images from active LIDAR
Generating passive NIR images from active LIDAR Shea Hagstrom and Joshua Broadwater Johns Hopkins University Applied Physics Lab, Laurel, MD ABSTRACT Many modern LIDAR platforms contain an integrated RGB
More informationEXTRACTING SURFACE FEATURES OF THE NUECES RIVER DELTA USING LIDAR POINTS INTRODUCTION
EXTRACTING SURFACE FEATURES OF THE NUECES RIVER DELTA USING LIDAR POINTS Lihong Su, Post-Doctoral Research Associate James Gibeaut, Associate Research Professor Harte Research Institute for Gulf of Mexico
More information