Synopsis

​

The core of all machine vision systems is the solution of problems based on the input images. The solution for each problem is unique to its own application. The applications in general can be divided into four main categories, namely, defect detection, classification, measurement, and location. The solution is obtained by processing the images digitally. The processing methods can be divided into image pre-processing (such as enhancement, noise suppression, distortion correction etc.), segmentation to separate the feature of interest from its background, extraction of useful image properties such as area, shape factor, bounding box, convex hull etc., and classification. 

​

Traditionally, and as practised till today, processing of digital images is done by laboriously writing codes using common languages such as C#, Python, etc. combined with image processing libraries such as Halcon, Matrox, OpenCV etc. Windows-based applications for machine vision application are usually developed by vision software engineers with in-depth knowledge of both the programming languages and image processing basics, that is until now. Today, thanks to AI tools such as ChatGPT, it is possible to develop advanced Windows applications for solving machine vision problems without the need to write a single line of code.

​

In the training, the participants will learn through guided hands-on activities on how to develop Windows applications to solve simple to complex machine vision tasks, ranging from defect detection by template matching to classification. This workshop will not dwell on the theory of image processing or fundamentals of programming, but will dive straight into developing Windows applications that can be implemented immediately, but without having to write a single line of computer code.