Euresys Deep Learning Bundle, a suite of powerful image analysis libraries

Euresys has been on a journey for a number of years, producing some of the most powerful image analysis libraries and software tools, recognized throughout the machine vision industry.

The Deep Learning Bundle is the latest development that completes a range of image processing libraries in regular use with an extensive customer base, mostly machine manufacturers who supply the semiconductor and electronics industries.

The Deep Learning libraries consist of software tools responding to three specific functions for which Deep Learning provides a solution: classification, segmentation and localization. They are part of the Open eVision Deep Learning Bundle, which can be used with the free Deep Learning Studio training and evaluation application.

BUT WHAT IS DEEP LEARNING?

Neural Networks are computing systems inspired by the biological neural networks that constitute the human brain. Convolutional Neural Networks (CNN) are a class of deep, feed-forward artificial neural networks, most commonly applied to analyzing images. Deep Learning uses large CNNs to solve complex problems difficult or impossible to solve with so-called conventional computer vision algorithms. Deep Learning algorithms may be easier to use as they typically learn by example. They do not require the user to figure out how to classify or inspect parts. Instead, in an initial training phase, they learn just by being shown many images of the parts to be inspected. After successful training, they can be used to classify parts, or detect and segment defects.

The Deep Learning Studio is an application that assists the user during the creation of the dataset as well as the training and testing of the Deep Learning tools of Open eVision. It is free and does not require any license. It allows users to test the Deep Learning libraries using their own images. It can be downloaded from the Euresys website. Sample images, manuals and example programs are included and no programming is required.