By Yong Sun Yong
GIS Officer, ASEAN Institute of Forest Management, Kuala Lumpur, Malaysia
At ASEAN Institute of Forest Management, we use GIS technology extensively for forest management applications and provide GIS education for people from all over the world. We have been using R2V for Windows as the data acquisition tool to automatically digitize maps and acquire spatial data for our GIS applications and found R2V highly efficient in automated digitizing of large size maps. With all the raster and vector data processing and editing tools provided, we are able to digitize maps from various sources very accurately and fast.
Following is a 30 minute seven step R2V training I prepared based R2V's user's manual and my communication with Dr. Yecheng Wu. I have given this training to my students and colleagues at my institute several times and it has helped them to get a quick start on using R2V. I'd like to share it here to show you how easily you can use R2V to digitize your maps and reduce the cost of your GIS applications.
STEP 1: ACTIVATE R2V
Start the program by double clicking the R2V for Windows Icon.
STEP 2: OPEN IMAGE
Open image by selecting File/Open Image. Specify the file name in the file open dialogue box. Several images windows can be opened at the same time. R2V supports most image types saved in TIFF formats, including 1-bit monochrome, 4 and 8-bit grayscale, 8-bit color and 24-bit color. We normally scan our maps at 200 - 800 DPI (dots per inch) depending on the original map quality.
Once an image is displayed, it can be zoomed in and out to any scale. For grayscale images, the contrast can be adjusted interactively using the View/Contrast command. The View/Overlay command is very useful to control the display items within the image window. All display layers, such as the raster image layer, line layer, point layer, control points, text area markers can be toggled on and off with this command.
STEP 3: IMAGE PROCESSING AND EDITING
Image processing functions are provided under the Image Menu. Commands such as Horizontal and Vertical Flip, Rotation, Transpose are used to correct image orientation. The Pixel Tool provides a very powerful raster image editing capability to remove defects in the scanned image. Select Image/Pixel Tool/Draw Pixel and then define the new pixel value and pen size. Use the mouse as a paint brush to edit the image pixels.
STEP 4: AUTOMATIC VECTORIZATION
Extract all vectors by selecting Vector/Extract Lines. A dialogue box appears to allow selections of vectorization parameters and select START to start the vectorization process. The cursor becomes an hourglass when processing and returns to arrow when done. The extracted lines will be displayed in the image window. All 1-bit monochrome and 8-bit grayscale images are automatically vectorized using this command.
R2V has several functions for color image processing and normally requires some combination of functions to extract vector data properly from color images due to the complexity of color image.
For 8-bit palette color image, because image pixels are color code and displayed with its defined color map, you need to select a training or feature region to indicate the color or colors to be vectorized. For example, if you want to get vector data for all the water boundaries, first draw a selection rectangle within a water area, such as river or lake, and while the rectangle is displayed on screen, select Vector/Extract Lines command. Then select "Boundary" option to vectorize the region boundaries. You can repeat this process to vectorize another color feature and add the vector data to the current data set.
For 24-bit true color images, it's normally the best to classify (Image/Classification) it first to clean up the image and then vectorize it. The classification is based on a clustering algorithm and is very fast. When selecting number of classes for classification, do not make it too many as regions will get fragmented too much. Do some experiment with the classification to see how many classes should be used for a certain type of images.
After the classification, if you want to get boundary for all the classes, simply select Vector/Extract command. If only one class boundary is to be generated, you need to draw a selection rectangle within a class region to indicate the class to be vectorized. When the selection rectangle is still displayed, select Vector/Extract Lines to get the boundary. Repeat this process to add more class boundaries to the current dataset.
STEP 5: LINE EDITING AND LABELING
Edit lines by selecting Edit/Lines and use the Right Mouse Button to bring up the popup menu for editing options. Within the editor, you can add new lines, add nodes, move nodes, delete nodes, split line, delete lines, delete all lines within a selected region or entire image. Lines can be labeled using Assign ID option after Set the ID value. Various vector data post processing and display functions are available under the Vector menu.
STEP 6: PROJECT TO REAL-WORLD COORDINATE
To convert the generated vector data to a certain projection system, such as UTM, start the Control Point selection tool by selecting Define Control Point option under Vector menu. The cursor changes to cross cursor when it's moved into the image window. Now you can bring up the control point editing options using the right mouse button.
Point and click at a known location to select. The control point dialogue box will appear to ask you to enter the coordinate values for the registration. In other words, if you want to map image location (10, 10) to a location under the new projection system, say, (2000, 3000), then the source coordinates are (10, 10) while the destination coordinates entered should be (2000, 3000).
Normally, 4 control points on the four corners are enough to make the registration although more can be selected. Note: the accuracy of the registration is totally dependent on how good the control points are selected, not on how many.
There are some editing options available, such as move, delete, add, change value, etc. Make the selection in the float pop up menu by pressing the right mouse button. Select Done from the menu when finishing the selection.
Note that since the vector data is geometrically transformed to the new coordinate system, they will no longer be aligned with your source image. So make sure save a copy of the vector data without registration applied so you can edit the data later using the source image as backdrop. The raster image can be geo-referenced by creating an Image World File using the selected control points.
STEP 7: EXPORT TO A GIS SYSTEM
Save your vector data by selecting File/Export Vector. The generated vector data can be saved to Arc/Info (ARC), ArcView Shapefile (SHP), MapInfo (MIF), XYZ (3D points file), SVG ( Scalable Vector Graphics) and DXF formats. Now, you have extracted the vector data from the scanned image and saved to a file other mapping or GIS software can use.
R2V will ask you whether you want to apply the registration if vector data is geometrically transformed to the new coordinate system. Select Yes and provide a file name to save the registered vector data. The vector file is now available for your GIS or mapping applications.