Image recognition has many appliances. On iPhone X it is used to recognize its owner. Special techniques are used for 3D recognition. Samsung uses infrared and iris scans, which is another technology.
For a Dutch university, we made image recognition software for animals and people, where a series of photo’s need to be analyzed. These photos are taken in the wild and are triggered by motion detection. This appliance has its own unique challenges since most pictures only have parts of an animal or human. We don’t need to recognize the people, we only need to know (for sure) that people are in this picture to avoid privacy issues while processing. Animals do need to be recognized and matched what type of animal it is, with great detail (so not just deer, but the kind of deer as well).
We have been very successful with that especially in comparison with a similar project done by an American University where we achieved 100% recognition of humans, even if only a foot or distant unclear image was provided.
De appliances for image recognition are huge. Think about airports, police but also finding lost pets or missing persons from public camera’s. This software is complex for both processing and the ability to scale. It uses artificial intelligence algorithms to create a recognizable object from a photo and compares it to a trained database with all animals we know.
Do you have a nice or great idea for our software? Take action and contact us, because if you can think of it, we can make it.