Machine Vision (Connectionist Image Recognition)  Project Overview
 
       Machine Vision Icon
Machine  
Vision  
Systems
• Bill worked closely with Motorola’s “semiconductor visual inspection” groups to
better understand the needs and state-of-the-art of machine vision.  Cyclone
hardware was evaluated for machine vision tasks.  Bill’s R&D in machine vision
led to the development of connectionist architectures that achieved simultaneous
scale, translation and rotation invariance.    

• Keywords: machine vision, retina, scale invariance, translation invariance, rotation invariance,
  convergent & divergent pathways.
   


Significance of this Accomplishment
 
  Significance
• For Motorola, automated machine vision inspection is crucial for quality control.

• Vision is the human sense through which we acquire the most information.
Bill’s early R&D helped him create PSIQ, which can achieve advanced machine vision.  

• Applications for really effective machine vision are worth many billions of dollars.


Skills Transferable to Your Projects:
 
  Juggler
Leadership / Management
• Bill initiated and led the interaction between his R&D group and Motorola’s “semiconductor
visual inspection” groups. 

Theory / R&D
• Bill originated the theory for several connectionist architectures that achieved simultaneous
scale, translation and rotation invariance.     

Software
• Bill specified the connectionist simulator’s overall structure and functionality.  Likewise, for he
specific machine vision architectures such as “Optimizing Maps (OMs).  Dr. Sidney Garrison, a
member of Bill’s R&D group made significant individual contributions to OMs.

Hardware
• Bill’s extensive experience (designing video display chips, boards and systems) provided an
quick transition to R&D in machine vision hardware.   Bill’s group created Cyclone hardware,
which accelerated our early versions of Optimizing Maps.


Patents
Patent Ribbon Patent   --Abandoned “Optimizing Maps (OMs)  
(Not issued, because “prior art” had surfaced and Bill had left Motorola).

Current Owner: Motorola, Inc.
Licensing availability: (None).
Contact Info: motorolaventures@motorola.com 


Future Opportunities
 
  Futue Clock		 By pushing beyond the current state-of-the-art,
there are these opportunities:

•  Machine vision is currently a hot R&D topic.

•  Companies like Numenta Inc. are promoting their HTM (Hierarchical Temporal Memory)

•  Lockheed Martin received $4.9 Million from DARPA to use Numenta’s HTM for machine vision.

•  Numenta’s HTM can be considered to be a small subset of PSIQ.


Links to related material
  Links of Chain •  See reference [9] in this related article.  

Back Arrow