Deep learning for automated analyses
Deep learning is used to train an artificial intelligence that performs image segmentation and analysis of 2D and 3D images in an automated manner, quickly and with reliable results.
With our microscopy software ZEISS ZEN core, you can acquire images, analyze samples, and integrate all your data into a unified ecosystem for connected microscopy. AI-based algorithms can be applied effortlessly, with no programming required, throughout the entire workflow. It covers the complete microscope portfolio, operating both light and electron microscopes from ZEISS for a connected research environment. Capture images and correlate results across different microscopes to unlock new possibilities in industrial microscopy, boosting performance and productivity through integrated AI tools. For image processing and analysis, it even goes beyond ZEISS systems and offers smart tool kits that can be used in combination with any third-party microscope. AI-based software, such as ZEISS ZEN core, is the key to automated, fast, reliable, scalable and, above all, reproducible results.
For industrial image processing, 2D and 3D images are first divided into different areas (image segmentation), which are then analyzed and provide important information. Industrial image processing is used for defect detection, structure and material analysis and also provides other data that is essential for quality control and assurance. If image processing is carried out manually in industrial applications, a great deal of experience and time is required for segmentation and analysis of the images. The solution: Artificial intelligence. Scalable, reliable and reproducible results can be achieved with automated image analysis. ZEISS ZEN core offers you exactly these advantages - and much more.
In an industrial environment, image analysis is usually carried out in the form of image segmentation of microscopic images. To do this, the image is divided into several areas that are separated from each other. These areas and the boundaries between the different areas are analyzed, e.g. to identify defects and ensure product quality. Depending on the object, however, special cases must be taken into account. Connected areas are only relevant above a certain size and shape, for example, or it is important how far apart the areas are. This makes manual analysis very complex and time-consuming. Image processing software is used here, replacing manual analysis with an automated process. This reduces user influence and minimizes the time required to obtain information. Image analysis can work in connectivity with gray value analysis (thresholding) or via AI-based models.
ZEN core offers various tools to support image analysis in the best possible way. Here you will find an excerpt from our software portfolio. All toolkits can be used independently of the imaging system.
Package contents:
AI-enabled - pre-trained models can be executed to evaluate data. The models cannot be created or modified - the AI Toolkit is required for these tasks; a subscription to APEER ML is required for models based on deep learning.
Toolkit for 2D image analysis by creating automatic measurement programs, including advanced processing functions
Complete AI application package, including integrated training interfaces:
SQL-based image database