Manufacturing – “Smart Factory”
Through rapid developments in IoT – dynamic and ever-evolving environments which generate high-volume streams of heterogeneous data – the whole manufacturing industry is currently overthrowing traditional production, quality control and surveillance practices.
In such environments The DeepEyes algorithm on standard cameras (as data capturing device, sensor) can be used to make working environments smarter, more precise and more secure.
Various application areas are possible:
The DeepEyes technology can give robots the ability to “detect” humans who enter their operation space on a very fine-tuned basis. This ensures human safety and at the same time reduces interruption times in the working process.
Montoring if individuals are not in a physical state to operate accident-prone machines, can be recognized (a-normal movements, gestures etc.)
Monitoring if individuals wear security apparel.
With DeepEyes technology both manual and chemical processes can be monitored in order to ensure that quality standards are met. With two 1080p cameras and our software we can detect anomalies of 0.01 mm in 3D at 1 m distance.
- Automated monitoring of gestures and movements of workers in meat/fish processing thus ensuring that quality standards are being met.
- Process surveillance in the production of synthetics or food (brewing, baking etc. )
The DeepEyes technology DeepPose can identify body movements and reflect these on the screen with the use of a standard video camera. This represents a huge advance in game development and also allows to enhance existing game offerings.
The scope of the DeepEyes technology is much wider than the usual Kinect and Leap Motion technology, as it does not require additional expensive equipment. A simple webcam or smartphone is all it takes.
DeepPose uses standard hardware, is very precise in detecting motion and distance, easy to use and there is no need for an internet Connection.
The DeepEyes algorithm is undemanding in operation. Processing a video in 720p / 30fps only consumes 10% of CPU (i5 2520m) and 50 MB of RAM. Optionally, the system can be configured to accompany the human hand, its fingers, or to recognize the posture of a person.