Image Interpretation in Remote Sensing

Introduction

Image interpretation involves specific vocabulary in remote sensing.
Terminology may be challenging for beginners.
Remote sensing images differ from ground views:
- Capture data from an overhead perspective.
- Can include radiance outside the visible spectrum.
- Often present unfamiliar scales and resolutions.

Classification
- Assigning objects, features, or areas to classes based on imagery appearance.
- Includes:
  - Detection: Ability to determine presence or absence.
  - Recognition: Identifying specific classes or categories.
  - Identification: Confidence in specific details of classification.
Enumeration
- Counting discrete items in the imagery.
Measurement
- Finding numerical descriptions of size or amount.
Delineation
- Drawing outlines of regions or objects.
- Historically done by hand; now includes computer-driven digitization but hand digitization is still common using aerial imagery.

Image Tone
- Refers to the lightness and darkness of an image.
- Continuous gray scale from white to black.
- Example: Dry sand reflects white, wet sand reflects black.
Image Texture
- Roughness or smoothness of the image.
- Defined as a group of repeated small patterns.
- Example: Grasslands (smooth texture), forests (coarse texture).
Shadow
- Can be a visual obstacle but also provides height and shape information.
- Example: Shadows can indicate the height of structures and time of day.
Pattern
- Arrangement within the image, including geometric shapes.
- Example: Street grids, rice fields, orchards.
Association
- Contextual clues based on nearby recognizable landmarks.
- Example: Recognizing Hoover Dam suggests a reservoir and power station nearby.
Shape
- Specific shapes of objects as viewed from above.
- Example: Conifer tree (circular crown), deciduous tree (irregular shape).
Size
- Determining size based on known features (e.g., baseball field dimensions).
- Proper photo scale selection based on interpretation purpose.
Sight
- Topographic position details including elevation, slope, aspect, exposure, and adjacency.
- Importance varies with local conditions.

The eight elements of image interpretation are essential for analyzing remotely sensed images:
- Image tone, texture, shadow, pattern, association, shape, size, and sight.
Understanding these concepts aids in more accurate interpretation of remote sensing imagery.