The acquisition of objects in video data is one of the core tasks of modern image evaluation tasks. In this case, the recognition of objects ("Is something there, where is something?") And the identification of objects ("What is there?").
Various features such as motion, color or background models can be used for detection. The selection of the characteristics depends on the specification of the task.
The identification is more complex, since the individual categories to be recognized usually have to be trained. That From representative data material a kind of template is first learned, with the help of which later object types can be classified.
Detection of boats and ships of marine platforms
Optical vegetable weighing machine, the fruit is automatically detected by the scale
Detection of persons and body regions using Manifold Learning Detection of micro-organisms to study drinking water quality