Region Of Interest, a region of an image on which algorithms are executed on.
Region Of Interest (ROI) is a selected subset of a data set identified for a particular purpose and is commonly used in many application areas. The data set could be any of the following:
- 1D data set: a time or frequency interval on a waveform
- 2D data set: the boundaries of an object on an image
- 3D data set: the contours or surfaces outlining an object (a.k.a. the Volume of Interest (VOI)) in a volume
- 4D data set: the outline of an object at or during a particular time interval in a time-volume
Tag Usage:
An ROI on an image can be defined in different formats, like a bounding-box, a polygon or a set of pixels. The tag is commonly used in question related to computer-vision or image-segmentation. An ROI can be defined to:
- Restrict the region of the data set that an algorithm is applied to and avoid unnecessary calculations. (e.g., text-recognition in an ocr)
- Identify/Annotate a part of data set. (e.g., locate moving objects in video-tracking)
- Extract additional information from the data set. (e.g., measure size of a tumor on a medical image)
An Example:
ROI locates the boundaries of a registration plate (it's Persian), this area will be used for ANPR: