PEKAT VISION Detector with Classifier modules screen

 

The PEKAT VISION Detector module stands out in identifying and optionally classifying a wide range of objects and defects under diverse conditions.

It’s widely used for detecting defective parts, verifying the presence or absence of components, and sorting products.

Finds and Classifies Objects & Defects

Cross icon in PEKAT VISION colors
Defect Detection
Reliably detects annotated defects & objects
Cross icon in PEKAT VISION colors
Easy Annotation
Intuitive, mouse-based annotation
Cross icon in PEKAT VISION colors
Classification
Optionaly classifies detected objects and defects
Cross icon in PEKAT VISION colors
Precise Location
Provides exact positioning of detected objects and defects
Deep learning Detector with Classifier module

Deep learning Detector & Classifier

In many inspection tasks, the Detector module is combined with the Classifier module to achieve highly precise results. The Detector locates the object or defect within an image, while the Classifier then assigns it to the appropriate class based on pre-defined categories.

This combined approach is especially effective for inspections requiring both localization and precise categorization, providing a versatile solution across diverse applications. When detailed classification isn’t needed, the Detector can be used on its own to identify any objects or defects of interest.

For either approach, all objects and defects must be annotated prior to training, ensuring that the modules are finely tuned to the specific inspection requirements.

Inspection with deep learning Detector and Classifier modules - real examples

Deep learning Classifier Module

The Classifier module can also be used independently and is well-suited for tasks where the entire image requires classification or for inspecting objects that consistently appear in the same position across images, as shown in the examples below.

Inspection with deep learning Detector with Classifier real examples

How the Detector & Classifier Modules Work

The Detector and Classifier modules leverage deep-learning neural networks to tackle complex inspection tasks by learning from annotated images. The Detector module identifies objects or defects within an image, using labeled data to accurately locate areas of interest. When combined with the Classifier module, these detected objects or defects are further classified into predefined categories, based on the annotations provided.

The Classifier module can also operate as a standalone solution, where it classifies entire images or static objects (those that remain in the same position across images) into specific classes.

Together, these modules provide a robust solution for detailed inspection, categorization, and quality assurance.

 

Deep learning Detector and Classifier modules - how they work

Watch the Detector with Classifier Video Tutorial

Watch our step-by-step video tutorial for the PEKAT VISION deep-learning Detector module. This tutorial guides you through the entire process, from selecting the image source and annotating images to determining evaluation criteria and adjusting the confidence threshold. You’ll gain all the knowledge you need to start your inspection with confidence.

Combine Modules for Advanced Vision Tasks

The individual PEKAT VISION modules can be seamlessly combined and integrated with scripting code to tackle a wide range of vision tasks in manufacturing. All in the user-friendly PEKAT VISION software interface.

Monitor showing easy customization of quality inspection

Get In Touch

Do you need more information? Let us know, we will answer any question you may have.
PEKAT VISION is now part of Datalogic Group