▸ Classification of space visuals:
▸ Classification of space visuals:
In the final stage of the controlled classification or target classification, the final product of the classification process is extracted with the preparation of the classification manual (similar to the map key) to determine the nature of the geographical phenomena represented on the space visual, with statistical tables prepared for each of these phenomena (e.g. area, number and percentages…. Etc.).
Aerial images are analyzed humanly based on high experience and intensive training and require time, but the analysis and interpretation of space visuals is often done computer-based on the use of specialized software.
Classification is a process designed to divide space visual into a number of categories or groups, each of which represents a specific geographical phenomenon on earth.
The classification process depends on the nature of the area (urban, desert, mountain or agricultural…. etc.) the spatial accuracy, spectral accuracy and radiometric accuracy of the space video used.
There are two ways to complete the classification or interpretation of space visual features:
1) Unattended classification or interpretation or automated classification
2) The most accurate classification, observer interpretation or classification.
Classification, undistingered interpretation, or automated classification is the process of interpreting space visualization based on a specialized program without user intervention.
As each space visual cell has been mentioned, a digital number represents the optical value or amount of radiation reflected from the surface area of the Earth represented by this cell, and the program then identifies cells that have the same digital number or fall into a specific category or period (e.g. the digital number between 55 and 90) and include these cells in one group.
There are some standard global standard classification systems (e.g. the U.S. Geological Survey’s Land Use Division System), which determine the type of geographical phenomenon based on digital number categories or cell photovoltaic values.
There are also some national or local classification systems, such as the Egyptian Dry and Semi-Dry Coastal Land Classification System.
The programme could therefore identify categories of classification of space visible phenomena based on one of these standard classification systems.
In this method, the user monitors or directs the software’s automated classification process by interfering with the identification of a numerical classification guide that represents the spectral characteristics of each feature pattern and geographical phenomena.
This human intervention is carried out through specific user information through its study of the geographical area and the knowledge of reliable information about its nature, geography and manifestations through old maps or space visuals already classified.
This human intervention process (called the training phase) takes place in parts of the visual where the user has real information about the nature of the phenomena or features of these parts and
then performs a human characterization process to identify the groups and characteristics of phenomena in those training areas, and then comes the classification phase of the entire space visual where the optical value of each cell is compared with the classification categories of the training phase.
In the final stage of the controlled classification or target classification, the final product of the classification process is extracted with the preparation of the classification manual (similar to the map key)
to determine the nature of the geographical phenomena represented on the space visual, with statistical tables prepared for each of these phenomena (e.g. area, number and percentages…. Etc.).