Green Team Summary of Findings: Imagery Analysis (4 out of 5 stars)

Note: This post represents the synthesis of the thoughts, procedures and experiences of others as represented in the 16 articles read in advance (see previous posts) and the discussion among the students and instructor during the Advanced Analytic Techniques class at Mercyhurst College on 15 April 2010 regarding Imagery Analysis specifically. This technique was evaluated based on its overall validity, simplicity, flexibility and its ability to effectively use unstructured data.

Description:

Imagery analysis is the extraction of meaningful information from images. It is used by both the military and the civilian sectors. Imagery analysis is highly useful in state-related matters as well as natural resource related businesses and scientific research institutes. The groups that use imagery analysis run the gamut from NGO's to multinational organizations.

There are many different types of imagery that can be used alone or in combination with others. This list can include satellite, infrared, electro-optical, and multispectral imagery. One of the most well known types of imagery is satellite imagery. Since the early days of satellite imagery there have been dramatic improvements in availability and quality that have lead to increased accessibility. Some of the largest commercial vendors are GeoEye, DigitalGlobe, ImageSat International, and Satellite Imaging Corporation. Google Earth is one of the most popular sources for images because it is commercially available to anyone with Internet access.

STRENGTHS AND WEAKNESSES:

Strengths

• Capable of providing a detailed overview of large area

• More manageable than on-the-ground classification efforts

• Accessibility and ability of systems continues to improve

• Data provided is clear and highly detailed

• Can be used to corroborate other collected data

• Can be applied to a variety of problems, and is very adaptable

• Allows for data collection in remote areas

• Relatively easy to perform cue to commercial availability

• Allows for a retrospective view for an area

Weaknesses

• Software and systems can be expensive

• Limited by resolution, image quality, atmospheric haze, and contrast

• Hard to compare images taken from different angles and at different resolutions

• This methodology still requires a trained analyst to go over the tentative findings and discriminate between objects

• Need for trained professionals (level of training debatable)

• Chance for error if preparatory steps not followed such as accounting for clouds, resolution, etc.

• Does not account for other factors such as change between the images

• Can be subject deception

How To:

There appear to be no standard procedures applicable to all images at all times or for all purposes. However, several standard processes for specific purposes were identified through the literature reviewed in advance of this class. For example, with respect to classifying types of vegetation:

1.Identify landmarks

2 Align landmarks (if using photos from different time periods)

3.Standardize levels of contrast within an image (necessary when images were taken by different sensors)

4.Image Classification- divide the photograph into "clusters", spatial areas with the same characteristics then decide what those are with respect to the region and culture of an area.

5.Accuracy Assessment- compare classification results to what the ground areas truly are as revealed by aerial photographs or teams on the ground.

Personal Application:

The class looked at ten designated images; five different cities at two different heights. The class had three minutes to look at the two images of the city and identify the following: Density, Bodies of Water, Building Shadows, Landmarks, and Vegetation. These five criteria helped to narrow the range of possible cities. For example, when looking at New York City, the identification was made possible by looking for vegetation (Central Park), building shadows (consistently tall buildings), and density (buildings very close together over a wide area).