The term change detection is part of data integration process, which has to do with the comparision of spatial information recorded over the same area at different times to detect changes in the features of interest. It be mut be noted change detection can be carried out at various levels of detail or sophistication, depending on the interest of the user.
This learning path will not only introduce you to the different kinds of changes associated with change detection but the methods/techniques of change detection.
In Remote sensing and image analysis, scales are fundamental and crucial elements, hence this path introduces Scales of processes and observations as the subsequent concept for our discussion.
Another element worth of consideration under this learning path is the term --Observable.
In change detection, different kinds of change are likely to be distinguished and some of these changes are introduced in this path. Firstly we will take a look at Gradual Changes --sometimes also known as Linear changes and that is changes involved in climate change investigations, urbanization, deforestation, etc.
Again in change detection, we also have Sudden changes--mostly related to disasters like floods, fires, earthquakes etc.
A change also distinguishable under change detection is Periodic change, which this learning path introduces to you.
Under this learning path, we now drift on to have a look at the techniques associated with change detection as the next line of action.
Here, one of the change techniques is Algebra technique, which includes all kinds of algorithms that are based on combinations of values of a pixel in subsequent images, such as image differencing, vegetation index differencing, imaging ratioing, image regression and so forth.
Classification-based change detection-- where the techniques in the category all involve some kinds of classifications of separate or combined images. Some of these techniques categories include:*post-classification comparison; *spectral-temporal analysis; *hybrid change detection; *Unsupervised change detection etc.
Another method of change detection is the Visual analysis, where the human eye remains the instrument available for detecting and interpreting any changes.
Under this learning path, another method of change detection is the combination of maps and images thus the GIS-based-- where GIS overlays on image data can provide a means for better interpretation and for detection of changes, e.g. new buildings, changes in parcel boundaries or forest limits.
The final element to be discussed under this learning path is the Object-based change detection -- where detection is mainly on object-oriented analysis.