Login Register

Auto-login on future visits

Forgot your password?

Newsletter March 2015

Wish to run a classifier out-of-Matlab in your C++ application?


OK, you can build a classifier in Matlab. But how do you bring it into your machine and thoroughly test it? How do you get it into real-time production?
perClass is designed exactly for this! In this example of wood defect detection we will show how easy it is to bring a classifier into your C++ application. Do you work with OpenCV, LabView, Halcon, Cognex or .Net? You're covered.
Watch this 4 minute video to find out more.

perClass course 13-17 April 2015


training course on industrial machine learning

Want to learn how to build the best classifier for your problem?
Join us in the 13-17 April 2015 course in Delft, The Netherlands.
In this intensive 5-day course, you gain a proven methodology to solve your pattern recognition challenges. Lectures and hands-on exercises on industrial problems such as:

  • Image-based defect detection
  • Texture classification
  • Hyperspectral object sorting
  • Classification of database records
You're the expert in your field! Bring your data and learn how to build robust classifiers for your domain. Read testimonials

Learn more about the industrial pattern recognition course register now to the machine learning course for industry
perClass software article in Vision Systems Design magazine

perClass featured in Vision Systems Design article

From the blog

Happy to share that perClass is prominently featured in the February issue of Vision Systems Design magazine. In the article the editor in chief Andy Wilson investigates the current state of the art in machine vision field. Andy writes "Perhaps one of the most comprehensive software toolkits for developing and testing different types of classifiers is PR Sys Design's perClass software."
Interested in the full article? You can find it here.

visualizing confusion matrix in perClass machine learning toolbox

perClass new release


perClass 4.5 release brings:

  • Fast RBF neural network classifier scalable to large data sets
  • Better missing value handling and imputation (class mean and median statistics, imputation from external data)
  • Improved image view (easier label painting and extraction of data from figures)

Read more in the release notes.

perClass used for real-time audio classification

From the blog

Framework for real time audio classification with perClass

Joanneum Research has adopted perClass in their Chronos-Trainer framework to design and deploy real-time audio classifiers. The framework was presented at the 6th Congress of Alps-Adria Acoustics Association in Graz, Austria.
The key element is that the application does not need to be recompiled to update or entirely change the classification logic. This makes researchers more efficiently focusing their effort on design of better algorithms that are instanteneously available for testing in real-world scenarios.
Read more in the blog.