ZEISS is an internationally leading technology enterprise operating in the fields of optics and optoelectronics. Thanks to our strong ties to our global customer community, we are delighted to partner with ZEISS to make this great solution become even more widespread – a clear win-win situation.” Through this partnership with ZEISS, we continue this tradition with DeepFocus 1, utilizing the best technology available for real-time extended depth of focus. Mark Curtis, Managing Director at Vision Engineering, agrees: “We at Vision Engineering have been offering our customers state-of-the-art solutions for their measuring and inspection tasks for a long time. With its extensive experience, our partner Vision Engineering embeds this in an attractive package that I am sure will convince a lot of new customers.” Robert Zarnetta, Head of Business Unit Industrial Microscopy Solutions at ZEISS, is happy about the collaboration: “With the unique MALS™ Technology in the Visioner 1 digital microscope, we offer customers a way to greatly increase their productivity and efficiency during their quality assurance and optical inspection.
DEEPFOCUS CAREERS SOFTWARE
Another part of the DeepFocus 1 package from ZEISS’ side is the powerful ZEN core imaging software suite.ĭr. The result is a more efficient failure analysis and quality control process with fast and accurate results. This allows optical inspection for height differences of up to 69mm with an up to 100 times greater depth of focus than that of a conventional microscope, which not only makes the inspection task easier, but also much quicker, saving significant amounts of time. The key to this is the namesake system with an array of micro-mirrors that can generate “virtual” lenses with distinctly different curvatures and thus focus planes, by changing the orientation of each individual micro-mirror in an orchestrated way.
DEEPFOCUS CAREERS SERIES
However, thanks to its unique MALS™ Technology, ZEISS Visioner 1 enables the user to see the sample completely in focus in real-time, without the need to Z-stack and post-process a series of images. Extensive refocusing to obtain an extended depth of field, or post-processing, is a solution to this problem, but can be time-consuming and complex. As a result, only a small area of the sample is in focus, which can lead to features being overlooked and the inspection being incomplete as a result. With the collection of sufficient training data, our deep learning focusing model provides a significantly faster alternative to conventional focusing methods.Conventional inspection systems are challenged by a shallow depth of field, especially at high magnification.
Furthermore, the rare cases where our algorithm does not find the focal plane can be detected, and a fine-focus algorithm can be applied to correct the result. The model was able to determine the in-focus position with high reliability, and was also significantly faster than conventional methods that rely on classical computer vision.
The CNN model was tested on bare semiconductor sample using the projected shape of the F-stop. The ground truth focal plane was determined using a parabolic autofocus algorithm with the Tenengrad scoring metric. A training dataset was acquired from a semiconductor sample at different surface locations on the sample and at different distances from focus. The difference of these two images is processed through a regression CNN model, which was trained to learn a direct mapping between the amount of defocus aberration and the distance from the focal plane. As an alternative, we developed a deep learning model that predicts in one shot the distance offset to the focal plane from any initial position using an input of only two images taken a set distance apart. Conventional microscopy focusing methods perform a time consuming sweep through the Z-axis in order to estimate the focal plane.
0 kommentar(er)
