Image Enhancement

Goals: 

The purpose of this lab was to develop skills in image enhancement, mosaic, model building, image enhancement for visual interpretation and processing techniques. The lab work on altering the spatial resolution of images for analysis as well as enhancing the radiometric quality of images and linking Google Earth Pro and Erdas viewers in order to analyze and interpret features and their associations to the surrounding landscape. 

Methods: 

In this lab we used the inquiry box tools and raster tools. To enhance the spatial resolution of the satellite images we pan sharpened them using the resolution merge tool. The image was further enhanced by re sampling the pixels by using both the nearest neighbor method and the bi linear interpolation method. We also reduced the haze on an image by using radiometric enhancement using the haze reduction tool provided on the toolbar.

Results:

Part 1: Image Sub Setting

This part looked at two sub setting methods. The first sub setting method was the Inquire Box in Erdas. This method helps create an image with a more regular shape.  

Figure 1. Image subset by using the Inquire box

The second sub setting method used a shapefile to process the image. The shapefile was overlayed over the original image. This method is more suitable for processing irregular images than the first Inquire Box method.

Figure 2. Image subset by shapefile

Part 2: Resolution Merge

This part focused on pan sharpening to enhance the spatial resolution of a satellite image.We used the Resolution Merge tool to complete the pan sharpening process.  The resulting pan sharpened image more clear than the original and had a higher spatial resolution. The features were much more defined in this second pan sharpened image.

Part 3: Radiometric Enhancement

When haze occurs in images radiometric enhancement techniques can help increase image quality. We used the Haze Reduction tool to reduce the amount of haze to improve the image interpretation.

Part 4: Linking Image Viewers to Google Earth Pro

We linked an image viewer with Google Earth Pro to improve image interpretation. This method allows you to view both viewer on a synced view which can help with image association and recognition. 

Part 5: Image Re sampling

Image re sampling changes the size of pixels of an image. We worked with the Nearest Neighbor method and the Bilinear Interpolation method. These methods have different advantages and disadvantages and are used in different circumstances.  

Part 6: Mosaicking Images

Mosaicking is the combination and stitching together of multiple images into one seamless image. In this section we utilized Mosaic Express and MosaicPro. Mosaic Express faster and simpler but not as smooth. MosaicPro had much better results and better color correction and was smoother overall.

Figure 3. Mosaic Express

Figure 4. Mosaic Pro

Part 7: Image Difference Interpretation

This part was dealing with binary change detection. It focuses on analyzing the brightness of the pixels to determine changes in the land forms. The pixel brightness values are subtracted from the other. However these two images need to have the same radiometric characteristics and resolution. The information needed is provided on the histogram. This section focuses on change between 1991 - 2011. Using the model maker and the Conditional equation we were able to see highlighted areas of change. Using ArcMap we were able to create a map that displayed these changes with a legend and scale.

Figure 5. Histogram used to detect change in Eau Claire area

Figure 6. Map created in ArcMap of change between 1991 and 2011