Marshall Kirchner's UWF GIS Blog

  Supervised Classification Map of Germantown, Maryland This lab was one of my favorites thus far, and involved a variety of new concepts/tools that are definitely very useful and powerful. I initially ran into troubles renaming my recoded classes, but I found that one must write the name of the recoded classes with their new class names before recoding. Other than this minor discrepancy, I enjoyed working through this assignment. I found it very important to use a band combination that has a large spectral difference to ensure there is little confusion within the classes. Secondly, I found it very important to use colors for the classes that are noticeably different from one another so that one may distinguish between classes in the image. I enjoyed using the distance image to find areas of the map that needed more signature to be added to ensure overall classification accuracy. One other minor issue I had was not having the 'Unclassified' class exist or appear in the legend,

  Feature Identification Using Spatial Enhancement Techniques I am becoming very confident in my GIS abilities and am much more comfortable with ERDAS Imagine software after this assignment. It was very enjoyable learning how to identify features within ERDAS and I will certainly use the techniques learned in this lab in future assignments. The histogram was crucial to my understanding of pixel values and what they mean. Utilizing grayscale/panchromatic images allowed me to notice trends that I would not otherwise see. I found that altering the band combinations allowed me to distinguish features from their surroundings, and I enjoyed changing these combinations to best suit the purposes of my final map layout. For my final layout, I decided to get creative and utilize a series of lines and polygons to create more unique subset maps that would showcase the features in a larger view. This final map I am very proud of and I enjoyed making it every step of the way. 

  Classified and Measured Washington Forested Land This week's lab involved utilizing a brand new software, ERDAS Imagine, to alter and express data. I greatly enjoyed being able to work outside of ArcGIS and appreciated learning the features and functionalities of this new software. I did encounter a few errors throughout the assignment, as a multitude of steps were either outdated or have changed since the creation of the lab. For example, the creation of the 'Area' field on ERDAS imagine did not properly adjust the areas of the classified lands as it should have. Also, there was no plausible way to adjust the Legend to display the areas of the land swaths, without manually having to do so via the labeling function. The second part of this lab (Part b) was very useful in learning more about the functionality of ERDAS Imagine, and I enjoyed learning how to find the spatial resolution on the software. It was interesting comparing how many of the functions of this app are si

  LULC Map of Pascagoula, Mississippi This week's laboratory assignment involved the creation of a Land Use/Land Cover map of a certain city in Mississippi. I was able to practice my editing and classification skills to create the map using a series of polygons in a created shape file. Upon creating these polygons, I classified them based on their unique land use/land cover characteristics and then labeled them. Then, I used the random point generator tool to create a series of 30 points that I would use to assess the accuracy of my classifications with Google Maps' Streetview. I really enjoyed this assignment, because I felt that my own unique perspectives and classifications were able to be used and it feels like the maps I am producing are becoming increasingly complex. It was difficult at first to distinguish between the different LULC classes, but miniscule differences (such as building shape/size, presence of parking lots, vegetation coverage, etc.) began to make all the