Deep Learning, Machine learning for image processing
This Deep Learning add-on is used together with Zebra Aurora Vision Studio. It's a new breakthrough in machine vision applications. It is a set of ready-made tools which are trained with Good and Bad samples, and which then detect defects or features automatically. Internally it uses large neural network structures, designed and optimized by our research team for use in industrial inspection systems. For the user, however, they are provided as simple filters with very few parameters, and with easy-to-use graphical tools for convenient execution of the training process.
Deep Learning Add-on truly embodies the main principles of Zebra Aurora Vision:
- Intuitive - can be used even by users with no programming skills
- Powerful - unleashing advanced capabilities of neural networks
- Adaptable - deep learning models can be retrained to include new features.
Quality Inspection of textile, scratch detection using Deeplearning
Deeplearning is ideal for scratch detection on textiles, wood, metal and other materials. The products are natural products and scratches don't have a fixed shape. Therefore this is very difficult to detect using classical machine vision. With deeplearning, these defects can be detected easily.
Inspection of syringes using deeplearning.
Products, like syringes, that are packed in transparant foil are hard to inspect with regular vision algorithms. Deeplearning can tackle these problems. You can inspect for example the print quality on the product or if all parts are present.
Classification of objects, like nuts and bottle caps
Deeplearning can also be used to easily classify objects. You can detect if for example a bottle cap is positioned ok. Or you can even differentiate between different products. With classical machine vision, this is also possible, however with deeplearning you have a robust algoritm developped in minutes while with classical machine vision it can take days up to weeks, depending on the complexity of the different object classifications.
Segmentation and locate point using deeplearning
With deeplearning, products that have not a fixed form, can be segmented easily, even when they touch eachother. Also the location can be detected easily. Below to examples with segmenation of Nuts and point location of bees.