Self Learning Eyebot - Easy to use Machine Vision Products for Quality Inspection

Sightech Vision Systems Home Page


The power of Neuro-RAM technology and the PC-Eyebot:

The PC-Eyebot utilizes a massive feature-based process that processes up to 12 million features per second in order to perform high-quality inspections. Each feature may contain 100’s of pixels of information. 

These features come in many types, and the one that should be used is determined by the needs of the application. The type of feature actually determines the PC-Eyebot's "View" of the product being inspected thereby controlling what kind of information will be learned.

We have feature types that range from shape inspection to very fine color descrimination. One very unique and very powerful means of color inspection is provided by a feature type that we call "Coloration". This mode learns not only colors, but the transitions from one color to another.  There are many billions of possible colors and even more possible color transitions. This allows for very sensitive product inspection.

Another new feature that has found demanding uses is "Texture". This feature learns grayscale information, but with no convolution preprocessing involved. The relational information is scaled and extracted within the feature execution. It is very useful for extracting subtle infomation with less-than-perfect lighting circumstances. It is also good for fabric and filter inspection.

We are constantly enhancing our product's ability to learn, and are always looking for new demanding applications. We have already discovered that our self-learning technology has already enabled many applications that could not be approached with tradition machine vision techniques.


PC-Eyebot is a revolutionary device that allows you to take advantage of the state-of-the-art techniques of machine vision.  No longer will you have to worry about the headaches of programming a machine vision system.  It produces results faster than traditional machine vision approaches and it can learn to recognize parts in a matter of seconds just by looking at them.


  • Self-learning machine vision system—reduces setup time to a matter of minutes.

  • Cost effective—can be used on a wider range of applications.

  • Compact size allows PC-Eyebot to be installed almost anywhere...

  • Can inspect up to 1,800 parts per minute (PPM).

  • Does not need a strobe or trigger input for many applications.

  • Offers modern low-noise high resolution digital cameras with or without trigger inputs.

  • Can be used to detect defects, to identify one set of parts from another, and identify objects within cluttered backgrounds, foreign materials, produce quality, etc.

  •  Offers 3 state output decisions: PASS, FAIL, and MARGINAL.  Each state can be assigned to an individual relay.

  • Can recognize or inspect for defects parts with full X/Y movement and full rotation.

  • Copious I/O: LAN, RS-232, USB, 1394, Barcode Scanner, TTL digital inputs, etc.

  • 40GB internal hard-disk data store – can save hundreds of project files for different parts, etc.

  • Up to 8 Relays to control notification and/or ejection mechanisms.

  • Offers many learning modes: Shape, Spectrum, Color combined with shape, Texture, and Coloration.

  • Offers many fixture modes: None,  Horizontal, Vertical, Full (XY), and Radial.

If you have a Multi-Session PC-Eyebot, you can setup more than one Processing Window – each Processing Window may have individual custom settings and may even control separate relay outputs.  This feature offers a simple and powerful method of defect classification. 



Copyright © 2005 Sightech Corporation — All rights reserved
Privacy Statement