Object Based Image Analysis (OBIA) technology examines pixels, not as individual cells, but in a group context. It works by imitating the human mind, using a combination of pixel colour, shape, texture and size. It then draws upon information relating to relationships with other pixels to in-turn classify these pixels as objects. This results in the ability to use different types of remotely sensed datasets to quickly and efficiently extract pixel based information to build classification maps.
The classification of these units is carried out using a rule-based approach allowing multiple inputs to be made; for example, allows differences in seasonal variations, soil surface wetness and the proportion of non-photosynthetic components within similar vegetation types can be differentiated.
OBIA can be used for:
- Tree crown delineation studies
- Tree counting
- Species identification and health mapping
- Quantity and quality validation of land parcels