Summary

PATTERN SORTING TOOL

The Datalogic Pattern Sorting Tool (PST) has state-of-the-art patented algorithm which is able to find thousands of different patterns in an extremely effective way. The Pattern Sorting Tool guarantees maximum performances and consistency in any situation, even with objects poor of texture in cluttered fields of view. This tool is a tremendous breakthrough in the machine vision industry as no algorithm is able to offer a reliable and robust recognition over such wide pattern databases today.

Thanks to Its capability to manage big databases of patterns, the Pattern Sorting Tool (PST) is able to distinguish among thousands of different items manufactured or simply conveyed on the same line. The algorithm delivers robust recognition in any situation (e.g. 360°pattern rotations, perspective distortions, different scales and light variations)..

The Pattern Sorting Tool (PST) is not just a tool for item sorting. The algorithm can be used for many different purposes and applications. For instance, it can detect the presence of stickers or logos as well as identify items by means of the artworks printed on the package.

PRODUCT HIGHLIGHTS

Large pattern database management

The PST is able to handle databases with thousands of different patterns. With the new Pattern Database Manager (PDM) software application, users can easily create new databases or edit existing ones.

Recognition of patterns with random orientation and different sizes

The Pattern Sorting Tool is able to find a trained pattern no matter its position and orientation. Moreover, the algorithm can detect a reference pattern even when its dimensions are not fixed.

Robustness against out-of-plane rotations

The Pattern Sorting Tool is able to cope with perspective distortions (i.e. out-ofplane pattern rotations). This capability is absolutely needed when the objects to inspect have variable and inconsistent positioning (e.g. boxes on a conveyor belt) or have irregular shapes (non-planar objects).

Partially occluded pattern detection

Partially occluded patterns do not represent an issue for the PST. By leveraging its ability to extract and match several pattern features simultaneously, the algorithm is able to identify patterns even when partially damaged or occluded.

Robustness against light variations

The Pattern Sorting Tool extracts and matches features with a minimum dependency on lighting. This guarantees an extremely reliable pattern detection even when the surrounding lighting is variable and inconsistent.

SUITABLE APPLICATIONS

• Large pattern database management
• Recognition of patterns with random orientations and different sizes
• Robustness against out-of-plane rotations
• Partially occluded pattern detection
• Robustness against lighting variations