PlanetBrain is able to localize and identify many different visual categories in images and videos. This skill is often utilized by our engines to preprocess images or videos to segment certain objects, e.g. to extract text lines or objects like cars, but can also be used as a stand alone solution. PlanetBrain finds even different, sometimes overlapping objects in the same image. It can handle a variety of different but typical difficulties as there are: only parts of the objects are visble, objects are blurry or unsharp, compression artefacts, overlapping objects, shadows and a lot more. PlanetBrain not only provides the area and type of objects found in the image, but also its own confidence for the classification.