Image recognition
Image recognition ( English image analysis ) is a branch of pattern recognition and image processing . In image recognition, one tries to segment objects in an image. A symbolic description is assigned to these, but connections between the objects are not searched for, as is usual in pattern analysis .
application areas
- Automatically label objects in images in large image databases ( tag , e.g. on flickr.com or Google's Cloud Vision API).
- Recognize images or image sections of media products and assign them to the corresponding product. Information on the products, for example content information, genre, assignment, criticism, reviews or price comparisons, can then be requested.
- face recognition
- Registration mark recognition in industry. Electronic components and registration marks are measured in order to achieve the most accurate assembly possible.
- Sorting and error detection of bulk goods, boards, photo prints, ...
- Object-based image analysis OBIA as a method of geography
- Automated detection of "abnormal behavior" of people in the video data streams of public surveillance cameras for crime prevention, see the controversial EU project INDECT
Applications
With the TapTapSee app, blind people can take pictures of their surroundings. A voice then says what can be seen in the picture. For example, blind people can find out the color of a piece of clothing. The app is free, so you can try out this technology yourself.
literature
- H. Niemann: Pattern Analysis and Understanding. Springer Series in Information Sciences, Berlin 1990
- A. Pinz: Understanding images. Computer science textbooks, Springer-Verlag
Web links
- KreissLInformatik: Image recognition and understanding (private side)
- OpenCV: Free software package for image recognition
Individual evidence
- ↑ A free app shows the blind the world chip.de.