Land Use Classification Using Remote Sensing - Principles and General Procedures


A seven step procedure needs to be applied:

A detailed step-by-step description of how to use remote sensing data for land use planning would require numerous pages and would go well beyond the description that can be provided in this paper. Therefore, only the most important “major” steps have been highlighted below. If the method is to be applied then the user will have to refer to references and additional literature that provide sufficiently detailed descriptions on exactly how remote sensing data is to be used for generating specific maps and planning information.


  • Step 1:    Acquiring the satellite images: Suitable satellite images have to be purchased. It is important that the resolution that is required for the planning purposes is defined so that the correct satellite images are purchased. Balancing the cost of the images with the use to be made of them also has to be thought through. Detailed and high precision images are generally more costly and the planner has to consider whether the additional detail these images provide warrants the added cost or whether less detailed “cheaper” images would not suffice.
  • Step 2:    Rectifying the image. The images that have been acquired need to be rectified. This means that all the maps that will be used have to be put into the same coordinate system.
  • Step 3:    Classifying the images: Considerable amount of work is required by the operator in order classify the images. Classifying involves the operator determining different features on the computer screen. In order to be able to do this, the operator needs to have a sound knowledge of the features that he/she has to look for as well as good ground knowledge of the area that is being classified.
  • Step 4:    Field verification: Once the initial classification has been completed on the computer, the operator needs to go to the field to verify whether the classes that have been identified are correct. For example, if the operator identifies a plantation as being an olive grove then the field verification will determine whether this was a correct classification or not.
  • Step 5:    Reclassification of the area: In the event that the classification (i.e. wavelengths used to classify, for example olive groves compared to other forest features) proved incorrect, the classification will have to be repeated based upon the information derived from the field verification
  • Step 6:    Simplifying the classification: In view of the fact that too many classification of features may emerge it may be appropriate to merge some of the categories and land covers. The merging process should be done according to the exact purpose of the map. For example, if the map is designed to present certain specific land use features (i.e. primary forestry areas) then a detailed classification of this feature is recommended, while other features that are not essential for the map could be merged in order to reduce the overall detail being presented.


Something missing, unclear, misleading or a typo? Help us to make this page better!
Upon approval, the MethodFinder team will publish your comment here (* mandatory fields):