General Inquiries: 1-800-331-7301   Product Catalog

News Blog

Solve Challenging Applications using Insight ViDi Deep Learning

Written by: Zack Serna, Automation Specialist

Introduction
In-Sight ViDi™ is a software designed specifically to run the ViDi deep learning tools for tough to solve applications that in the past weren’t possible.  Applications are deployed on the In-Sight D900 smart camera without the need for a PC, making deep learning technology accessible to non-programmers. The software uses the familiar and easy-to-use In-Sight software platform which simplifies application development and factory integration for those with and without vision experience.

Highlights

  • Example based training
  • Decodes challenging OCR applications in minutes
  • Performs fast and accurate assembly verification
  • Analyzes complex defect detection tasks
  • In-Sight spreadsheet guides application development
  • Deployed on powerful, customizable vision systems
  • Web-Based HMI

What is Deep Learning, and Why Does it Matter?
Deep Learning uses example based training from an image set to create a neural network to determine good parts from bad parts.  This can be handy for applications where the distinction between a good and bad part is minimal, when there are part anomalies, or poor part contrast.  This is a fast and reliable solution for tough applications where traditional machine vision might not have been a perfect fit.  

How Insight ViDi Deep Learning Works
A set of good and bad images are collected and run through the ViDi deep learning software using a GPU enabled PC and one of the 3 app tools described below.  The user validates the model created by ViDi using the images collected to help define good and bad parts.  Once the model is complete ViDi creates a neural network which is used in the software to determine a good part from a bad part.
 
ViDi Main App Tools
ViDi Read
The In-Sight ViDi Read tool deciphers badly deformed, skewed, and poorly etched codes using optical character recognition (OCR).  It works right out of the box, dramatically reducing development time, thanks to the deep learning pre-trained font library. Simply define the region of interest and set the character size. In situations where new characters are introduced, without vision expertise, this robust tool can be retrained to read application-specific codes that traditional OCR tools are not able to decode.

 ViDi Check
The In-Sight ViDi Check tool uses artificial intelligence to reliably detect complex features and objects and verifies parts and kits are assembled correctly based on their location within a pre-defined layout.  The tool can be trained to create an extensive library of components, which can be located in the image even if they appear at different angles or vary in size.
 
ViDi Detect
The In-Sight ViDi Detect Tool learns from images of good parts in order to identify defective parts. ViDi Detect is ideal for finding anomalies on complex parts and surfaces, even in situations where defects can be unpredictable in their appearance.  It learns from images of good parts in order to identify defective parts. This allows the tool to detect a wide range of defects that do not need to be pre-defined at the time of training.
 
Deployed on powerful and customizable vision systems
In-Sight ViDi uses an In-Sight D900 (video link) smart camera powered by In-Sight ViDi software (spec link) designed specifically to run deep learning applications. The D900 can be deployed without a PC or the need for an external controller. This highly-modular, IP67-rated deep learning vision system includes field-changeable lighting, lenses, filters, and covers that can be customized to match your exact application requirements. It also includes an embedded inference engine that is specifically designed to solve complex deep learning applications at production line speeds.