So long as you get the GIST of it
  1. Classes
  2. /
  3. Remote Sensing -- GEOG526
  4. /
  5. Labs
  6. /
  7. Lab - Data Acquisition and Image Preprocessing

Lab - Data Acquisition and Image Preprocessing

Introduction This lab serves as an introduction to downloading remotely sensed data from one, of the several, online sources and includes some common remote sensing image preprocessing functions. This lab allows you to become familiar with the process of searching for and downloading a Landsat scene, and how to stack layers from the original products. You will also use multiple SPOT images to learn the image mosaic and subset processes and to subset the image to an area of interest. You need to submit your results to BB this lab. Because you only have one chance to submit your work, submit all the results after you finish the whole lab. There are totally 9 snapshots you need to submit, please name your file as #_your last name (# represents 1-9). Readings Textbook: Chapter 6 and Lab08 Guide Materials: SPOT6 MS images; ERDAS IMAGINE 2014

Part1: Landsat Image Acquisition, Download and Import

In this portion of the lab you will search Earth Explorer to identify “suitable” Landsat 8 imagery, download a scene and import it into ERDAS Imagine.
A suitable image is defined by the user’s needs. This includes the spatial extent, the spatial resolution, temporal resolution and the level of cloud contamination. You may only need a portion of a scene to be cloud-free. Or perhaps your application requires a very specific time stamp (data from a specific year, season or both), so you may be able to accept some cloud contamination. You may need to acquire imagery from different seasons to capture vegetation phenology and that may require the use of data spanning multiple years. Go to http://earthexplorer.usgs.gov/. Follow the instructions to create an online account. Once your account is created, log in to the website using your credentials.
You can search for data using various spatial parameters. For example, you can define using an Address/Place (street address, City, State, County), a Path/Row (such as the WRS-2, the unique scene identifier for Landsat discussed in Lab 6) or X, Y coordinates (which can be the location of one point, an intersect between two pairs of x,y coordinates or a box defined by x,y coordinates). You can also specify a date range for the search. Leave this part blank for now. Under the Search Criteria tab type in the location of an area you are interested in acquiring Landsat imagery. Click the Address/Place tab to define your location. Type in one or more of the following to your location: street address, city, county, state, country. Click the Show button. Then Click on an Address/Place in the list to show the location on the map and add coordinates to the Area of Interest Control.

  1. Where is your location of interest? ___________________________¬ (1 point)¬ Click the Data Sets tab and scroll through the list to see all the data available for download. Now Click on Landsat Archive. Then check the box L8 OLI/TIRS (i.e. you are searching Landsat 8 imagery only).
  2. How many datasets are available through Earth Explorer? _____________________ (1 point) Click the Additional Criteria tab. This tab allows you to specify additional search criteria including % of acceptable cloud cover, Data processing type (i.e. what level of processing has occurred on the data), path/row or even scene identification. Look at the options. Do not select anything under this tab, for this exercise we want to see all data available for your selected area regardless of cloud cover. Click the Results tab to view the list of images available for the area you specified.

In the Search Results window, Click on Show Browse Overlay, the second option in this toolbar . It will display your image over the imagery in the main window. It will highlight the icon in green like shown above, when the image is displayed. Scroll through the available imagery and display images that look the most cloud free. Find the most cloud free image for the year and take a screen shot and submit it to BB.

  1. What is the path/row and the date of the Landsat 8 image you found with the least amount of cloud cover? (2 points) Path/row: ______________ Date: ___________________________

  2. What is the estimated percent cloud cover over the entire scene (circle one)? (1 point)
    Cloud-free <10% 10-20% 20-30% 30-40% 50-60% 60-70% 70-80% 80-90% 90-100%

IF you were to download the image, click the Download icon on the image toolbar. Click to download the Level 1 GeoTiff Data Product and save the image locally. (this could take a few minutes) We are not downloading these data for this lab.

Pretend you downloaded your image. You would need to use 7 zip or another software package to uncompress the file. Each band would be in a separate TIFF file. Navigate to G:/Fall-2015/G526/Lab08/Landsat8/ and copy the individual bands to your computer. You now need to stack the bands to produce one output image file. To create a layer stacked image, open ERDAS Imagine 2014. Under the Raster tab click Spectral (located under the Resolution subtab) and select Layer Stack. For the input file, navigate to the location where you copied the TIFFs and select band 2 (*_B2). Click the Add button. Repeat the process for bands 3-7. Make sure and add the bands in sequential order. Once all bands are loaded in sequential order, specify an output file name and location and use an *.img file format for output. Then click OK to run the process. Once the process is complete, view the layer stack in ERDAS Imagine. Close ERDAS Imagine and delete the files you copied and created. Now we will do another search on Earth Explorer. You will use path/row and year information to identify the most suitable scenes for crop type mapping in Iowa. You will need to identify the best spring date, summer date and fall date to represent the seasonality of the growing season. For the purposes of this lab, spring is defined as March 1st – May 31st, summer as June 1st- August 31st and fall as September 1st – November 30th. Use this Search Criteria: Path/Row 25/31; Data set: Landsat 4-5; Dates: March 01, 2010 through November 1, 2010. 5. How many images are there to choose from? ____________ (1 point)

  1. a. The best spring image date is: ______________________. (2 points)

b. The best summer image date is: ____________________. (2 points)

c. The best fall image date is: ________________________. (2 points) Part 2. SPOT Data import and format conversion We will find variety of data format when you download the remote sensing products. For example, “hdf (Hierarchical Data Format)” is the format for MODIS, ASTER. The data format of SPOT6 is JPEG 2000 (with the file extension as .jp2). In most of time, we want to keep all of the dataset in one format, which is convenient to manage. In this part, we will learn how to use ERDAS 2014 to impart the data and export data to the format you want. ERDAS IMAGINE 2014 has the default format: IMAGINE image (.img). Open the ERDAS 2014, navigate to “Manage Data” tab, click “Import Data” and then you will see the window-“Import”. In this exercise, we will import the SPOT data (in folder include exact directory location SPOT6_MS) to the ERDAS format.

Change the “format “ to JPEG 2000, select your SPOT data and then specify the output name and location (The default format is *.img). After you press OK button, it will show a new window, and click OK.

Repeat the process to import all the jp2 files in the folder. Or you can do batch by coding and import all the images at once. Part 3. Image mosaic Load all six SPOT6 images (*.img) into the same 2D view. Set the image “fit to frame” (we have talked about several ways to do it in previous labs), and then you should see a view like this.

Navigate to “Raster” tab, click “Mosaic” and choose MosaicPro from 2D View. It will show a new window (MosiacPro). You also can add and delete the images. You can try different function buttons in the tool bar. For example, you can set the resample method by clicking Edit-> Image Resample Options… Then you click Process->Run Mosaic…

  1. After the mosaic finishes, open the image in a new 2D View. Compare the views, what’s the differences between unmosaic view and mosaic view? (2 points)

  2. If you are a data manager of large database, discuss the advantages and disadvantages of storing your data as mosaiced images from the view of data management (you may start from the aspects of updating data, storage, retrieval and distribution of data, processing of data, etc.) and data applications. (3 points)

  3. Reload all the six original SPOT6 image in a new 2D view. And this time, you can try to play with create/edit/delete hole in the mosaic process . Create one hole (or several) in the image mosaic process, and then redo the mosaic process. Snapshot your final results and submit it to BB. (2 points) Typically adjacent images have overlap areas, e.g. the aerial photos (required about 30% side-overlap). In these cases, you may need to create a seam line or seam polygon. We will not cover this part in this lab. Part 3. Subset Imagery (see section 11.8 in your textbook for a complete description) Now make the mosaic result visible. Go to “Raster”, and click “Subset and Chip”. ERDAS provides several ways to subset images.

A) Choose “Create Subset Image”.

  1. Try to subset the following features and save and submit the snapshots to BB. (hint: this is SPOT6, with blue band) (8 points)
  1. Ocean ( with true color combination)
  2. Airport ( with true color combination)
  3. Agricultural area (with False color combination(color infrared))
  4. One feature you are interested in (with your interpretation result).

B) Choose “Dice image” then the “Dice an image” window shows up. This function is to cut the image into several similar pieces, just like the original SPOT6 images.

  1. Dice the mosaic image into 3 *2 sub-images (3 columns * 2 rows). Submit the snapshot of results to BB (true color combination) to the assignment. (4 points)

The values you set for “Dimension in x dir.”________________ “Dimension in y dir.” _______________ (2 points)

The following figure is an example of the results in false color combination.

C) Create an irregular shape of subset. a) Create an “AOI” (Area of Interest) layer. Right click on the image, and choose “New AOI Layer”. It will have a new layer in your active 2D view.

b) Draw an AOI. Navigate to “Drawing” tab, and in the section of “Insert Geometry”. You can choose different shapes to draw an AOI.

Choose “Polygon” button (see where the mouse is pointing above), then you can create an AOI over the area of interest you want to subset. To outline your AOI, position your mouse at the starting location and click, then move to the next area and click, repeat until you have delineated your subset area. After digitizing the AOI, go to File-> Save as -> AOI layer as… to save the AOI as a file. Save the aoi file to the Lab08 directory.

c) Then go to the Raster->Subset & Chip -> Create Subset Image. Now You DON’T need to define the “subset Definition” we set before, which was a rectangle shape. Click the “AOI” button. In the Window of “Choose AOI”, click AOI file and go to the Lab08 directory and select the AOI file you created before. After you give an output file name, click OK. You will get a subset output file like below.

  1. Now create the irregular shapes of AOI, one polygon, one ellipse (press shift when you draw an ellipse to create a circle), and subset the features you like. Submit two snapshots to BB. (6 points)

You are done!!!

Wrapping up

There is no need to save anything from this lab, so when done you can simply close without saving. Submit your answers to the questions on blackboard.