Dalton Inman

  For this lab, it was asked to create a map of supervised classification of land use in Germantown, Maryland. This was completed by creating spectral signatures specifying what type of land class an area was at given coordinates points. The main map and inset were created in ERDAS Imagine using the Thematic Recoding tool. These were then exported to ArcGIS Pro to create the final map.

For this lab, four types of skills were learned to create this map. The first was examining pixel value histograms for shapes and patterns in the data. This allowed for understanding how many bright and dark features were within an image. The second was viewing an image as grayscale to identify darker and lighter shapes and areas was. This was used to find extremely dark or light features, but could also be used to identify features that had a transition of dark and light, called edge detecting. The third was adjusting multispectral bands to make specific features stand out from one another. By adjusting the red, green, and blue bands of an image, certain features that would not be able to be seen in a True Color or grayscale, would appear. Lastly, the fourth was implementing an inquire cursor to find and identify the specific brightness of features. The inquire cursor in ERDAS Imagine can identify which layer of an image is the brightest or darkest, and can also be used to identify th...

  In this Lab, ERDAS Imagine was explored to view two types of satellite images (AVHRR and Landsat TM). Landsat TM aerial imagery was chosen to make a map of as it is much more detailed than AVHRR imagery. The image is a subset of a larger image that was in ERDAS, which was turned into a map in ArcGIS Pro. Specific areas can be pointed out and the size of each can be analyzed as well (i.e. hectares), as seen in the legend. Digital data can also be interpreted on ERDAS as well as different resolution types, which define how well an image can be seen or interpreted. 

  This map that was created in ArcGIS pro, showcases outlined regions of Pascagoula, Mississippi. These regions are given codes based on the aerial view of the image (i.e. collection of small houses is a residential area given the code number 11). The areas on this map were then compared to Google Earths street view and aerial imagery, to find whether the accuracy of the codes placed by the user were correct. Areas that have a green dot matched with Google Earth, and red dots were incorrect on this map compared to Google Earth.

  In this first aerial image, areas were mapped out based on their tones and textures. For tones (red outline), differing areas are marked by: Very Light, Light, Medium, Dark, and Very Dark. Tones represent the amount of darkness or lightness, or brightness, in an image. For textures (purple outline), differing areas marked by: Very Fine, Fine, Mottled, Coarse, and Very Coarse. Textures represent the tonal variation in an image or particular areas of an image. Texture is seen by the mixing of multiple tones, i.e. very course is a combination of all types of tones congregated in one area. In this second aerial image, points of reference were marked by order of: Shape/Size, Shadows, Patterns, and Associations. Shape/Size of an object was determined to understand what the object could be. These will typically be very simple things.  In this image, a Pier (the long stretch of area, moving away from sand), Sand (makes up the majority of areas within image), and Road (Outlined Sand ...