Artificial Intelligence

NOTAVIS / Product range / Artificial Intelligence

Rely on artificial intelligence

Artificial intelligence (AI) includes several technologies which make it possible to equip machines with human-like intelligence.

Computer vision and robotics have a huge potential for artificial intelligence. For example, robots can be programmed to define their own objectives and persuing them consistently.

We offer solutions in the field of 3D image processing in combination with robotics. These include 3D vision-guided robots and the development of customized 3D applications. Solutions of our partner HD Vision Systems for optical measurement and reconstruction of objects and surfaces complete our offer in the area of artificial intelligence.

What are the advantages of artificial intelligence?

With the help of AI, computers can solve tasks which require intelligence.
AI can be used for processing speech (speech recognition).
AI enables the development of intelligent data base systems and user management systems.
With the help of AI, programming languages for computer vision (machine vision) and robotics can be developed.
In computer vision, images (e.g. grey value analysis), objects and events (e.g. meaning of a scene) in camera images can be detected more effectively thanks to artificial intelligence.

Applications of artificial intelligence for industrial image processing

Detection and picking

The bin picking system from NOTAVIS also recognizes complex 3D objects! In our system we combine 3D vision with deep learning. The robot can be programmed to take a product from a storage box und put it in a package. We automate your bin picking!

Quality inspection

The NOTAVIS technology detects geometrical deviations, surface scratches and the completeness of components. Data-based approaches of artificial Intelligence (AI) and machine learning are used for the inspection.

Scanning and measuring

With the NOTAVIS system of 3D image processing you perform a 3D scan in just about 1.5 seconds. 3D applications can be developed and customized quickly and reliably. Automate your incoming goods inspections and your measurement!

Technologies

Light field camera - light-field-based sensory

When light arrives on an object, it is reflected from its surface in different directions of space, thus generating a light field. In addition to the 2D image information, so-called light field cameras record the direction of the reflected light rays.

The LumiScanX of our partner HD Vision offers a light-field-based sensory for robot vision. LumiScanX also recognizes shiny 3D objects without error. Thanks to artificial intelligence and machine learning the system can be quickly adapted to new objects.

Stereo vision cameras

Stereo vision cameras take images simultaneously through two lenses arranged next to one another. The result are two stereoscopic frames. Viewing theses images in a stereoscope creates a spatial impression.

Bin picking is one of the most difficult tasks in robotics. For robots it’s particularly difficult to pick objects they don’t know yet. Intelligent stereo vision cameras can solve this task by means of ai-based software. The bin picking system from NOTAVIS also recognizes complex 3D objects.

Further application possibilities of stereo vision cameras include artificial intelligence, self-driving vehicles, drones a smart city applications.

Technologies

Light field camera – light-field-based sensory

Stereo vision cameras

Further application possibilities of stereo vision cameras include artificial intelligence, self-driving vehicles, drones a smart city applications.

Point Cloud

A point cloud describes many points in a linear space. Due to the unorganised spatial structure the whole of these points is also called “cloud”. Every point stores the spatial coordinates as well as colour and light intensity of a measuring point.
Georeferenced point clouds show the points in an earth-related coordinate system.

The points of a point cloud are captured primarily by optical scanning nowadays:

Triangulation via laser technology calculates spatial coordinates by the angle of the reflected laser beam.
Structured-light 3D scanners use image processing software to calculate the spatial coordinates of individual points. In this process patterns (e.g. lines) are projected onto the object.

Applications include:

Visualization of scanned objects and surfaces on a computer
Modelling: Restoring of objects using 3D rendering. The image processing software connects individual points to polygons, forming a surface.
Planning production steps
Automated processing of scanned objects, e.g. automated quality inspection
geometrical calculations, e.g. body scans or measurement of objects
Identifying people (face detection) or objects (e.g. bin picking)
Medicine (forensic, criminalistics): e.g. creation of surface models, documentation of injuries