Here you will find a window view of the projects I have worked on in the past. Please view these works as an informal portfolio. The + indicates that not all of the projects posted here are GIS related. I, Christie Rajtar, am the author of these works unless otherwise stated.
Proposing an Infrastructure Project
This was a project from one of my classes.
We were given a proposed infrastructure of Redstone near Rossland B.C. which we had to edit in AutoCAD, apply symbology, and import the .dwg into ArcGIS referenced with a coordinate system.
CAD.dwg
CAD Proposal
Final Product in ArcMAP 10.2:
ArcPY Programming Snipit to ceate quick and easy Study Area Boundaries
If you are interested in this project, view my blog post here.
Remote Sensing- Vegetation Changes
Visualizing Thailand’s land and vegetation changes following the 2012 Tsunami
Linear 2% Enhanced original images of an area in Thailand.
Before Tsunami
After Tsunami
Visualizing NDVI = (Near IR λ – Red λ)/ (Near IR λ + Red λ)
The brightest areas (white) represent the strongest vegetation, and black, weakes
NDVI Before Tsunami
Applying Differencing to NDVI images on the Green Band.
NDVI Differencing (before-after)
Adding a Rainbow Colour Table to the above image allows better interpretation of vegetation change.
Tsunami NDVI DIfferencing
Red represents the positively changed pixels (most change-loss in vegetation).
Blue represents the negatively changed pixels (minimal change in vegetation).
Remote Sensing: Data Sharpening (Gram Schmidt Transformations)
Two images with overlapping area of different pixel qualities can be fused to achieve higher resolution using Gram Schmidt Transformations.
Ikonos image Res: 4.0m
Quickbird image Res: 2.4m
Following Data Fusion, the cyan box highlights the fused/overlap of images and it is noticeably higher quality than both original images.
Data Fused Ikonos and Quickbird images using Gram Schmidt technique
GPS or Global Navigation Systems:
I had a class this semester for GPS where our final project was to compare the accuracy of two GPS devices, create 10 coordinates, and create a path. Allie Winter, Lauren Maluta and I worked together to complete this project over a two week period. The GPS devices used in this project were the Trimble TerraSync and GARMIN CX60. Before going out into the field, we used Trimble Office software for our mission planning. We found out the times of the maximum number of satellites visible to our specified location, and chose a time of day that maximized the satellite views which our group members were available.
Upon taking points and travel paths with our devices, we had to correct our data to an almanac for a specific date and time and corrected path data to complete gaps in our line feature, and remove background points of high standard deviations. This improved the smoothness of our line feature, and created a lower overall standard deviation for every point creating the line and point features. We referenced our point features to the closest tower and created an map to visualize our data in ArcMap 10.1. Our group came to a conclusion that the Trimble terrasync was a more accurate device than the CX60, and our points were more accurate in the x direction.
.
Automation -Finding a Suitable Habitat
This model was built in ArcMap 10.2 in order to identify suitable habitat for deer using the following parameters:
Slope < 80%
Southerly Aspect (112.5-247.5 degrees)
Elevation < 750m
Land
Further than 200m from a paved road
Model Builder
Final Map Product
Infographic
This infographic was created using Adobe’s Illustrator and Photoshop and ArcMap 10.1. This infographic displays the elevation of 7 peaks by symbol, and the distance of the peaks to the Seven Summit Trail in the West Kootenays, British Columbia.
Relief Map
Novice relief map designed in ArcMap and edited using Adobe Photoshop
Unsupervised Classification Methods
ISOData is a variant of the K-Mean method. The K-Mean method takes pixel seeds and classifies pixels based on the closest mean, where each new pixel is added into a class and a new mean is generated. It keeps iterating until there is no significant change in the mean of the pixel classification. The ISO DATA method adds a new pixel to a class and calculates the interclass distances. If there is not significant difference in the means, the means merge; if the standard deviation is larger than a predetermined threshold, a class is split in two, if there aren’t enough pixels to make a class, the pixels are dispersed into different classes.
Using the ISO Data technique, I will show you the difference between 5 and 10 iterations of classification. 10 iterations nicely classify the image compared to 5 iterations where it is very dark and difficult to see. ISOData image classification is much easier to see than K-Mean over only 10 iterations.
Original Image
5 Iterations with ISOData Method
10 iterations with ISOData Method
Related articles
- RedStone, Rossland BC (rdejonge1.wordpress.com)
- Our Local GIS (keenanfarq.wordpress.com)
- An Introduction to Working With LiDAR Data in ArcGIS 10.1 (camerona2650.wordpress.com)
rova news 