Scene analysis

© Photo Fotolia, Fraunhofer IPA

The group “Scene Analysis” focuses on the optical recognition, analysis and interpretation of dynamic objects and people as they interact with their environment. Besides human motion in general, the group also studies specific movements of certain parts of the body, such as arms and hands. Depending on the application, different data is collected in varying qualities from an overall scene. Diverse sensor technologies are implemented in dependence on their technical capabilities (e.g. range, resolution, angle of aperture). The recorded data is then put into a broader context. Data is merged intelligently to enable a scene to be interpreted as required by the respective application.

The uses and applications are wide-ranging. For example, person detection is used to recognize falls and emergencies, aid diagnoses or identify emotions on the basis of typical movements of the human body. Hand and arm movement recognition - also in conjunction with the analysis of interaction with objects - is particularly suitable for assembly assistance, touch-free control or the intelligent classification of handling processes.

This technology has a broad spectrum of users, ranging from private users, carers of patients and the aged, through people working in hospitals, in workshops where people with disabilities are employed and in production facilities, right up to workers performing manual and machine-aided assembly tasks and bio laboratory processes.

 

Ambient Assisted Living (AAL) and Smart Home

The scientists at Fraunhofer IPA develop assistive systems for everyday life to give elderly people comfort and security. The technical solutions can be integrated very discreetly into the home environment.

 

Motion analysis for quality assurance

The scientists at Fraunhofer IPA develop optical motion analyses of complete production environments including all stationary and moving objects and people within it, thus paving the way for the full realtime digitization of running processes.

 

Data fusion in the work environment

The scientists at Fraunhofer IPA develop functional and reliable solutions for locating and detecting workers. To recognize features comprehensively and reliably, different optical systems are combined.

 

Classifying hand and arm movements

The scientists at Fraunhofer IPA develop methods to automatically recognize typical hand and arm movements.

 

Human safety - recognizing falls and emergencies

The scientists at Fraunhofer IPA are working on a system which detects and processes emergency situations such as falls or immobility instantly and reliably.

 

Diagnostics support

The development of technical aids to support diagnostics aims to reduce the risk of diagnosis errors being made, as well as speed up the diagnosis process.