OPTIONAL EXERCISE 1: CREATING A FUSION PROJECT PART A

Transcription

1 Exercise Objec ves In the previous exercises, you were provided a full Fusion LIDAR dataset. In this exercise, you will begin with raw LIDAR data and create a new Fusion project one that will be as complete as the previous example dataset. In subsequent exercises (Crea ng a Fusion Project Parts 2 and 3), you will con nue to create a rich Fusion project by: crea ng the Bare Earth surface, a Top of Canopy Model (TCM) and a Canopy Height Model (CHM). You will also work with Fusion command line executables in the DOS environment and create batch files to allow you to work more efficiently. Prerequisites Exercise Data: lubrecht.zip, includes nine les of LIDAR data(lub_ le1.las...lub_ le9.las). Overview of Major Steps 1. Download and Organize Data. 2. Create and Load the Reference Image. 3. Explore the Catalog Command. Op onal Exercise 1 page 1

2 1. Download and Organize Data a. Determine the format. The LAS (laser) file format is a public file format for sharing 3 dimensional point cloud data between users. Although the LAS format is the preferred format for lidar data, vendors will o en provide lidar data in other formats such as ASCII text or ASCII CSV. In order for these files to be used by Fusion s LDV, they must first be converted to either an LAS or LDA file. Fusion can read LAS files directly the only step required is to generate index files (.ldi and.ldx). This happens automa cally the first me an LAS file is loaded. The files that we will use for this exercise are a set of nine LAS files (lub_ le1.las, lub_ le2.las...lub_ le9.las). Delivering LIDAR data in les is a common delivery method to split the data into a more manageable file size. For more informa on about the LAS format, click the following link: h p:// ee General/LASer LAS File Format Exchange Ac vi es.html. A er you ve downloaded the data file (lubrecht.zip), you must extract the enclosed.las files to c:\lidar\lubrecht (you ll have to create the lubrecht folder first). b. Consider the various methods to inspect the content. 1. Use Fusion (Tools Miscellaneous u li es Examine LAS Files) to inspect the LAS file headers. 2. Use a text editor (e.g. Wordpad or Notepad) to open an ASCII text file, or Use Excel to open an ASCII CSV file. 3. Ask yourself ques ons such as: Does the data include all returns (requires filtering)? Are addi onal a ributes available: intensity (can be used to create a geolocated image), scan angle, return number, other? c. Inspect the LAS file headers. 1. Open Fusion. 2. Click Tools l Miscellaneous U li es l Examine LAS file headers. 3. Navigate to and Select one of the LAS files. First Return (pulse) density image created from the lidar data. Fusion will read and write LAS files, but some of the fields for each return are not populated (although the data is s ll there). All returns are counted as first returns in the LAS header file (right) This missing informa on is important to determine the number of returns that exist in the data, i.e.:1st and last or mul ple returns (beneficial for forest structure modeling). To view more informa on about each data point, the LAS file can be converted to an ASCII csv file. LAS files are ready to use in Fusion and conver ng them to CSV files will give us addi onal informa on about the data (which may help us in our data analysis). Op onal Exercise 1 page 2

3 4. Take a moment to inspect. d. Convert one of the LAS files to a CSV file format. This is an op onal step to learn more about your LIDAR data. 1. Click Tools l Data Conversion l Export data from LAS or LDA files to other formats 2. Select one of your LAS files for the input file. 3. Navigate to an output folder of your choice and create an output file with an.csv extension 4. Ensure that you use the output format drop down box to select (ASCII CSV format (XYZIPRS)). 5. Click Convert. When the process finishes close the window. 6. Navigate to and load the CSV file into Excel a er the data conversion is complete. Excel will only open a por on of the file since it is limited to 65,536 rows (records) but that is more then sufficient for our purpose. Do you no ce addi onal informa on available in this data format? (for example the return #, nadir angle or intensity). Even though your LAS data is already compa ble with Fusion you can t view it in Fusion un l you load an associated geo referenced image. We ll create one now Create and Load the Reference Image A georeferenced image is a cri cal requirement for exploring LIDAR data in Fusion it provides the coordinate system and visual reference for the LIDAR data. There are at least three ways to input a useable reference image: 1) find an exis ng georeferenced image (e.g. NAIP; see note below) remembering that you may have to reproject the image to match the LIDAR projec on; 2) Use the Create image using LIDAR data u lity found within FU SION; or 3) you can create a reference image automa cally by running the catalog.exe DOS command line u lity. In sec on 3 of this exercise, we ll explore the catalog command for now, let s use the Create image using LIDAR data u lity to create our reference image. a. Open Fusion (if it is not open already) 1. Click Tools Miscellaneous U li es Create an image using LIDAR point data Navigate to and Select all nine LAS files (lub_ le1.las...lub_ le9.las) 3. Click Open this will open the Image Creator dialog: 4. Click the Browse bu on. 5. Specify an output loca on and file name (suggested: lub_ref.bmp) and Click Save. 6. Change the Pixel size to Click the Create image bu on. 8. Once the image is created click the Close bu on. b. In Fusion s main bu on panel, click the Image bu on. 1. Navigate to and select lub_ref.bmp and click Open. 2. Ensure that the checkbox is checked next to the Image bu on. The intensity image will be displayed in Fusion (see Appendix A for more details on improving reference imagery). You can also use an exis ng image (e.g. NAIP or an orthocorrected photo). The image has to be in the same projec on and same units as the LIDAR data but does not have to have the same geographical extent as the LIDAR data it can be smaller (as small as a single pixel) or it can be larger. Op onal Exercise 1 page 3

4 3. Explore the Catalog Command The Catalog u lity has a number of useful func ons: 1) it can automa cally generate index files, data density images, return intensity images and more; 2) it provides a quick assessment of the file footprint (coverage) and point density sta s cs within the footprint; and 3) it provides an ini al QA/QC of vendor provided data. The catalog u lity is most useful for a large number of vendor delivered LIDAR data files (Nine in our case. See bo om right graphic), but can also be used for single data files. We re exploring it now since it provides an alternate way to create reference images from the data. Catalog.exe is a Fusion command line program thus, it can be used in batch mode (to be explained later in this exercise) or directly at the command line. In addi on, it can be used directly at the command line on a large numbers of files. a. Data Prepara on 1. Create a text file containing the file paths to the nine LAS data files. (Similar to the example in the side bar). Suggested Name: lub_ les.txt. 2. Save lub_ les.txt to an intui ve loca on as you will call on the file repeatedly in the following exercises. b. Let s explore the Catalog command. 1. Open a DOS command prompt window. 2. Type catalog <enter> and inspect the syntax quickly. 3. Open the Fusion Manual and read the Catalog sec on. Remember, the more you understand the capabili es of the FUSION commands, the more efficient you will be at extrac ng useful informa on from lidar data. c. Now that you are familiar with the Catalog command, let s dissect an example below: Input LAS data files compiled into one text file. This makes it easier to run the Fusion command line executables with mul ple les (LAS files) and create one output file. catalog /image /index /coverage /firstdensity:2,1,3 /intensity:2,0,255 /outlier c:\lidar\sampledata\lub_tiles.txt c:\lidar\sampledata\lub_cat Catalog is the FUSION executable command /image creates image files showing the coverage area for each LIDAR file. /index creates LIDAR data file indexes if they don t already exist /coverage creates one image that shows nominal coverage area for all data files included in catalog /firstdensity:2,1,3 creates an image for all data files that shows the density of first returns for the area represented by each pixel. The three parameter values represent: Overall Data Coverage JPEG. pixel area, minimum point density, and maximum point density /intensity:2,0,255 creates an intensity image for all data files using the average intensity for all first returns within each pixel. The three parameter values represent: pixel area, minimum intensity value, and maximum intensity value. /outlier iden fies data les that might contain eleva on outliers. c:\lidar\sampledata\lub_ les.txt specifies the input data file. Op onal Exercise 1 page 4

5 c:\lidar\sampledata\lub_cat defines the output data file. 1. When you have finished entering the script, Save catalog.bat and Close. 2. Navigate to your batch folder and right click on catalog.bat, choose edit. 3. Double check your syntax. Note: if you le click twice on the batch file it will run, even if you are not ready!! e. There is only one more step: execute the batch file from the command line prompt: 1. Open your DOS Command Line window, if needed. 2. Change the directory to the batch folder (>cd \fusion\batch), if needed. 3. Type lub_catalog (you don t need to type in the.bat part) followed by the enter key and let it run. 4. Navigate to and open the output lub_cat.html file (it will open in a web browser). Note the File Summary Sta s cs (includes Extent and Eleva on Mins and Maxes, ~5.2 Million returns, nominal return density of 1.23 returns/default unit area). 5. Examine the First Return (Pulse) Density Image and legend pay special a en on to the red and yellow pixels in the image. These areas that have a low number of first returns (pulses). Can you explain the large yellow area in the upper right quadrant of the image? 6. Finally, inspect the Intensity Image how does it compare with the Intensity image you created earlier in this exercise? 7. Save the current project as Op onalexer03.dvz. We stated earlier that the catalog u lity provides a means to perform a quick QA/QC of vendor provided data. What is your assessment of this data? End of Exercise 3 Op onal Exercise 1 page 5

6 APPENDIX 1: IMPROVING REFERENCE IMAGE RESULTS 1. Adjus ng Contrast and Brightness a. If the resul ng image lacks contrast or is too bright, extra steps can be added to improve results. 1. Before clicking the Create Image bu on, click the Scan for data ranges bu on. This will display a histogram of the data. 2. If, for example, the data range is but there are very few intensity values greater than 200 you can click the Clamp data range bu on and enter a min value of 0 and a max value of 200 (not 255). This will improve the contrast and reduce the brightness of the image. 2. Se ng Color of No Data Value a. This reference image (right graphic) was created from the LIDAR data thus, the reference image will have the exact same geographical extent as the point data file and be perfectly co registered to the point data file. Also note: red pixels represent the background of the image (no data) This is a fast way to check the distribu on of the data. 1. You can re create the reference image (or create another) that has a less dis nc ve background color (a medium grey works well) 2. Another op on for re crea ng the reference image is by increasing the pixel size (lower pixel size values give a more resolute image but with more gaps (red pixels) in the data, higher pixel values give fewer gaps but coarser resolu on). The intensity image lub_ref.bmp. Any image created will have an associated world file (for bmp files the associated world file has a bmpw extension as shown below). The World File gives pixel dimensions and coordinate data for the intensity image. Appendix 1 page 6

7 APPENDIX 2: IMPORTING GENERIC ASCII.CSV TO A LAS FORMAT 1. Convert the Raw Data a. While examining the raw data near the beginning of this exercise, there was an op onal step to create a CSV file to examine the data more closely. In some cases you may receive data in this format. The following steps outline the process of conver ng a CSV file to an LAS format and generate an index file (an index file makes working with the data faster): 1. Open FUSION 2. Click Tools Data Conversion Select the op on that is appropriate for the file format your CSV file is a generic ASCII file so choose the Import generic ASCII lidar data... bu on. This will display the Import ASCII data dialog as shown in the sidebar graphic. 4. Within the Import ASCII data dialog click the upper Browse bu on to select the CSV file you created earlier (note: you will have to change the Files of type to All files to see the csv file). 5. Click the lower Browse bu on to specify the output binary LAS file loca on and name. (Suggested name: lubrechtxyziprs.las (lubrecht experimental forest and xyziprs represents the first le er of the imported fields)). 6. Start import on row The default Column assignments are correct through Intensity (Field 4), but you will need to change the last four column assignments: 8. Select Field 5 for Pulse Number 9. Select Field 6 for Return Number 10. Select none for Returns/pulse (the default) 11. Select Field 7 for Scan Angle 12. Ensure that the Parsed columns sec on appears as expected. 13. Click the Import data bu on. Note: Although they re not labeled, the upper input box (with associated browse bu on) is for the input file, and the lower input box is for the output LAS file. Appendix 2 page 7