Standard Software Packages

We're proud of the speed with which WAY-2C can be taught to identify even the most complex colored objects. However, we know that, no matter how capable the vision system, real automation tasks often require much more time and effort to get up and running than originally anticipated. Then there's the issue of retraining the system to accommodate light source aging or replacement, product updates etc. without losing expensive production time.

Drawing on our experience in developing color-based recognition systems, we have put together a set of standard packages which can dramatically reduce this time and effort. These packages, often utilized in our custom turn-key systems, are also available separately. Each contains automated script generation, operator interface, and other modules that utilize customer maintained databases to facilitate such operations as:

• camera alignment
  
• training setup and execution
  
• inspection setup and execution
  
• goodness of fit testing
  
• switching from product to product
  
• operator interfacing
  
• digital I/O interfacing
  
• system maintenance
  

Assembly Color Inspection System	This is a general purpose WAY-2C front end. It automates the generation of scripts for inspection of assemblies when configuration changes are relatively infrequent and many identical items are inspected between product switchovers. The Assembly Inspection System is specially designed for applications in which the potential exists for most components to be incorrectly installed in more than one place. It is based on our original Fuse Color Inspection System which has been successfully used for many years in automotive fuse block inspections. Added functionality and updated documentation reflect its usefulness in other areas such as electronic assemblies, wiring harnesses, sample cards and kit inspection.
Basic Color Sorting System	This is a powerful, easy-to-use menu based front end designed for high speed identification and sorting of singulated objects. It incorporates our experience with successful applications in such areas as sorting plastics for recycling and lumber color matching. The many features simplify such things as routine training, retraining, testing and product switching.
Sample Card Inspection System	Can be setup for a new card in just minutes and switched between different card designs in seconds. A typical single-camera manual feed sample card inspection system can inspect over 3600 30-swatch cards per hour. Customers report inspection reliability generally much better than human inspectors.
Page Color Sort System	Many companies are in the process of upgrading from paper records to electronic records. As part of this process they are having existing paper records scanned, often OCR'd, and converted to PDF files. During this process companies who used color coding to differentiate different classes of documents are faced with the possibility of losing the benefits of this color coding. The Page Color Sorting System (PCSS) is designed to avoid this loss. Starting with a PDF file containing JPG images of scanned documents PCSS can perform two separate functions. The first, bookmarking, generates a new PDF file in which color based bookmarks have been added to the original. The second function, sorting, extracts copies of the individual JPG images and places them into separate folders, each folder containing pages of a single color.
In addition to the above standard packages we have a library of literally hundreds of scripts and other software modules for all aspects of color-based assembly inspection and sorting. These include such things as multi-lane and web based sorting, handling as many as 190 discrete outputs, area, length and curvature measurements. There is a good chance that some of these may be applicable to your project.

Hyperspectral Data Selection and Analysis (Contact us for details.)

With the increasing availability of multispectral and hyperspectral imagery it is important to know which combination of spectral bands is most useful for differentiating a specific set of classes of interest in a scene. We have developed, and been awarded US Patent 8,918,347 for, an automated method of rapidly determining the optimum band combination to differentiate any particular set of target classes. This method, which we refer to as relevance spectroscopy has been demonstrated by experimental studies on suites of hyperspectral images collected by NASA, military, and other sources. The resulting optimum image combination is suitable for analysis by traditional methods and/or using WAY-2C.