Semantic Segmentation is a set of techniques used to separate a single image into multiple sections/images.
Semantic Segmentation uses image-processing methods that may be integrated into a client's software environment.
a) Adjustable and customizable algorithms
b) Hierarchical segmentation – from large-scale partitions to low-level segments
c) Capable to be optimized for specialized types of images (stereo images, depth images, thermal images)
d) Fast performance – up to 10 fps on a mobile device
e) Real-time learning – segments are adjusted using feedback from an object detector or a human expert
f) Multimodal segmentation performed on an array of different images (for example RGB-depth image segmentation)
a) Hierarchical segmentation for robotic applications
b) Medical images segmentation
c) Preliminary stage for object recognition
a) Object Recognition
b) Localization and Navigation