Visual interpretation is a crucial step in reviewing aerial photos or satellite imagery to identify and assess various objects on the Earth's surface. Through remote sensing technology, image interpretation becomes the primary method for obtaining accurate spatial information about geographic landscapes.
This process involves analyzing geographic reality based on the detection, identification, and spatial localization of objects captured by sensors. Each object has unique visual characteristics, known as interpretation elements, that allow analysts to distinguish one object from another precisely
 |
| Object characteristics (soonet.ca/eliris/remotesensing). |
8 Object Characteristics in Visual Interpretation
To conduct an in-depth interpretation, remote sensing standards utilize eight basic elements of object recognition. Below are the explanations along with applicative examples:
- Tone and Color
Tone is the level of brightness or darkness of an object found on an image (especially on black and white images), while color applies to multispectral imagery. Tone is the primary recognition element; without tone differences, boundaries between objects would not be visible. Tone depends on how much the object reflects, emits, or absorbs electromagnetic waves.
Example: Wet asphalt surfaces, moist soil, or deep waters will absorb light, giving them a dark tone (black). Conversely, a new zinc roof or beach sand reflects light, giving them a bright tone (white). - Shape
Shape shows the general configuration, outline, or two-dimensional boundary of an object. Shape is a highly distinct attribute, allowing many man-made objects and natural features to be identified directly.
Example: School buildings generally have an I, L, or U shape. Volcanoes are recognized by their cone shape. A football field is easily recognized by its rectangular shape with specific ratios.
Shape of a football field (https://earth.google.com/). - Size
Size includes the spatial dimensions of an object such as distance, area, length, and volume. Size must always be related to the image scale. This attribute is highly useful for distinguishing two objects that have similar shapes but different functions.
Example: Distinguishing between residential houses and industrial factories. Both might be rectangular, but a factory building has a much more massive size (area) compared to the roofs of residential homes.
Size comparison of buildings on satellite imagery (www.google.com/earth/). Texture is the frequency of tonal changes on the image, produced by the aggregation of small objects that cannot be distinguished individually. Texture is generally classified into three categories: coarse, medium, and fine.
Example: A tropical rainforest with an uneven tree canopy will produce a coarse texture. Agricultural areas (like sugarcane or palm oil) have a medium/mottled texture. Meanwhile, grasslands, empty land, or the surface of a calm lake will produce a fine texture.- Pattern
Pattern is the spatial arrangement or layout of repeated shapes recorded on the image. Pattern is highly effective for distinguishing natural phenomena from man-made objects, which tend to be much more regular.
Example: Natural river flow patterns (like dendritic or radial patterns), transmigration settlement patterns that form a regular grid, or the neat rows of rubber plantation areas - Shadow
Although shadows can hide object details located in dark areas, the shape of a shadow actually helps identify the vertical profile or relative height of an object.
Example: Telecommunication towers, skyscrapers, or suspension bridges are often easier to identify and have their relative heights calculated from their elongated shadows cast on the ground. - Site
Site is the geographic location of an object relative to its surrounding environment. Object recognition based on site relies on landscape characteristics such as topography, slope, or proximity to specific natural features.
Example: Mangrove forests are always situated in muddy coastal areas or brackish river estuaries. Tea plantations are situated in highlands or mountainous areas. - Association
Association is the strong relationship between one object and another. If one object is found, its appearance is often a sure indicator of the presence of another object nearby.
Example: A train station is strongly associated with branching and converging railway tracks. A university or school building is marked by the presence of sports fields and large parking areas nearby.
The integrated application of these eight visual interpretation elements is crucial in remote sensing data extraction, which is subsequently widely used in urban spatial planning, disaster mitigation management, deforestation monitoring, and global environmental analysis.
Referensi
Aljoufie, M., Brussel, M., Zuidgeest, M., & van Maarseveen, M. (2012). Urban growth and transport infrastructure interaction in Jeddah between 1980 and 2007. International Journal of Applied Earth Observation and Geoinformation, 21(1), 493–505.
Bell, D. M., Cohen, W. B., Reilly, M., & Yang, Z. (2018). Visual interpretation and time series modeling of Landsat imagery highlight drought’s role in forest canopy declines. Ecosphere, 9(6), e02195.
Farghaly, D., Elba, E., & Urban, B. (2016). Towards Sustainable Land Uses within the Elbe River Biosphere Reserve in Lower Saxony, Germany by Means of TerraSAR-X Images, (March), 97–121.
Kucsicsa, G., & Grigorescu, I. (2018). Urban Growth in the Bucharest Metropolitan Area: Spatial and Temporal Assessment Using Logistic Regression. Journal of Urban Planning and Development, 144(1), 05017013.
Svatonova, H. (2016). Analysis of Visual Interpretation of Satellite Data. ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, XLI-B2(July), 675–681.
Tiede, D. (2014). A new geospatial overlay method for the analysis and visualization of spatial change patterns using object-oriented data modeling concepts. Cartography and Geographic Information Science, 41(3), 227–234.