Fei Comapny

Other

Pavel Potocek

Faysal Boughorbel

Maurice Peemen

Remco Schoenmakers

Berend Helmerus Lich

Cornelis Sander Kooijman

Tomás Vystavel

Bert Henning Freitag

Ondrej Machek

Eric Gerardus Theodoor Bosch

Hendrik Nicolaas Slingerland

Mathijs Petrus Wilhelmus Van Den Boogaard

Libor Strakos

Lukás Hübner

Xiaodong Zhuge

Milos Hovorka

Emine Korkmaz

Peter Christiaan Tiemeijer

Gerard Nicolaas Anne Van Veen

Holger Kohr

Erik Rene Kieft

Brad Larson

Martinus Petrus Maria Bierhoff

Matthias Langhorst

Hendrik Jan De Vos

Franciscus Martinus Henricus Maria Van Laarhoven

Remco Geurts

Lukás Drybcák

Remco Johannes Petrus Geurts

Maurice Peeman

Ingo Gestmann

Drybcák Lukás

Mart Bierhoff

Frank Van Laarhoven

Van Gerard Veen

Cees Kooijman

Ben Lich

Eric Bosch

Tomographic images are obtained by processing a tilt series of 2D images by aligning and combining images withing a group of neighbor images. The tilt series generally includes sparsely sampled images. Images of the tilt series at tilt angles associated with the sparsely sample images are selected as reference frames, grouped with neighbor images, and the group of images aligned. The aligned images are combined to produce replacement frames and a replacement frame tilt series that can be used for tomographic reconstruction.

Sparse image reconstruction from neighboring tomography tilt images

A method of imaging a sample includes acquiring one or more first images of a region of the sample at a first imaging condition with a charged particle microscope system. The one or more first images are applied to an input of a trained machine learning model to obtain a predicted image indicating atom structure probability in the region of the sample. An enhanced image indicating atom locations in the region of the sample based on the atom structure probability in the predicted image is caused to be displayed in response to obtaining the predicted image.

Live-assisted image acquisition method and system with charged particle microscopy

Methods for drift corrected, fast, low dose, adaptive sample imaging with a charged particle microscopy system include scanning a surface region of a sample with a charged particle beam to obtain a first image of the surface region with a first detector modality, and then determining a scan strategy for the surface region. The scan strategy comprises a charged particle beam path, a first beam dwell time associated with at least one region of interest in the first image, the first beam dwell time being sufficient to obtain statistically significant data from a second detector modality, and at least a second beam dwell time associated with other regions of the first image, wherein the first beam dwell time is different than the second beam dwell time. The surface region of the sample is then scanned with the determined scan strategy to obtain data from the first and second detector.

Dual speed acquisition for drift corrected, fast, low dose, adaptive compositional charged particle imaging

A method comprises: using a Scanning Electron Microscope (SEM) to acquire an image of a specimen; identifying one or more objects of interest within the SEM image; generating a scan mask indicating a first set of one or more regions corresponding to the identified one or more objects of interest; and based on the scan mask, providing instructions to the SEM to acquire one or more Electron Backscatter Diffraction (EBSD) images from the first set of one or more regions of the specimen, wherein the method is performed by at least one device including a hardware processor.

Adaptive specimen image acquisition using an artificial neural network

Techniques for training an artificial neural network (ANN) using simulated specimen images are described. Simulated specimen images are generated based on data models. The data models describe characteristics of a crystalline material and characteristics of one or more defect types. The data models do not include any image data. Simulated specimen images are input as training data into a training algorithm to generate an artificial neural network (ANN) for identifying defects in crystalline materials. After the ANN is trained, the ANN analyzes captured specimen images to identify defects shown therein.

Training an artificial neural network using simulated specimen images

Techniques for adapting an adaptive specimen image acquisition system using an artificial neural network (ANN) are disclosed. An adaptive specimen image acquisition system is configurable to scan a specimen to produce images of varying qualities. An adaptive specimen image acquisition system first scans a specimen to produce a low-quality image. An ANN identifies objects of interest within the specimen image. A scan mask indicates regions of the image corresponding to the objects of interest. The adaptive specimen image acquisition system scans only the regions of the image corresponding to the objects of interest, as indicated by the scan mask, to produce a high-quality image. The low-quality image and the high-quality image are merged in a final image. The final image shows the objects of interest at a higher quality, and the rest of the specimen at a lower quality.

Adaptive specimen image acquisition

The invention relates to a method implemented by a data processing apparatus, comprising the steps of receiving an image; providing a set-point for a desired image quality parameter of said image; and processing said image using an image analysis technique for determining a current image quality parameter of said image. In the method, the current image quality parameter is compared with said desired set-point. Based on said comparison, a modified image is generated by using an image modification technique. The generating comprises the steps of improving said image in terms of said image quality parameter in case said current image quality parameter is lower than said set-point; and deteriorating said image in terms of said image quality parameter in case said current image quality parameter exceeds said set-point. The modified image is then output.

Method implemented by a data processing apparatus, and charged particle beam device for inspecting a specimen using such a method

Disclosed herein are CPM support systems, as well as related apparatuses, methods, computing devices, and computer-readable media. For example, in some embodiments, a charged particle microscope computational support apparatus may include: first logic to, for each angle of a plurality of angles, receive an associated image of a specimen at the angle, and generate an associated scan mask based on one or more regions-of-interest in the associated image; second logic to, for each angle of the plurality of angles, generate an associated data set of the specimen by processing data from a scan, in accordance with the associated scan mask, by a charged particle microscope of the specimen at the angle; and third logic to provide, for each angle of the plurality of angles, the associated data set of the specimen to reconstruction logic to generate a three-dimensional reconstruction of the specimen.

Charged particle microscope scan masking for three-dimensional reconstruction

The present invention relates to a method of training a network for reconstructing and/or segmenting microscopic images comprising the step of training the network in the cloud. Further, for training the network in the cloud training data comprising microscopic images can be uploaded into the cloud and a network is trained by the microscopic images. Moreover, for training the network the network can be benchmarked after the reconstructing and/or segmenting of the microscopic images. Wherein for benchmarking the network the quality of the image(s) having undergone the reconstructing and/or segmenting by the network can be compared with the quality of the image(s) having undergone reconstructing and/or segmenting by already known algorithm and/or a second network.

Method, device and system for remote deep learning for microscopic image reconstruction and segmentation

Methods and systems for generating high resolution reconstructions of 3D samples imaged using slice and view processes where the electron interaction depth of the imaging beam is greater than slice thicknesses. Data obtained via such slice and view processes is enhanced with a depth blur reducing algorithm, that is configured to reduce depth blur caused by portions of the first data and second data that are resultant from electron interactions outside the first layer and second layer, respectively, to create enhanced first data and second enhanced data. A high-resolution 3D reconstruction of the sample is then generated using the enhanced first data and the enhanced second data. In some embodiments, the depth blur reducing algorithm may be selected from a set of such algorithms that have been individually configured for certain microscope conditions, sample conditions, or a combination thereof.

Depth reconstruction for 3D images of samples in a charged particle system

Methods and systems for creating TEM lamella using image restoration algorithms for low keV FIB images are disclosed. An example method includes irradiating a sample with an ion beam at low keV settings, generating a low keV ion beam image of the sample based on emissions resultant from irradiation by the ion beam, and then applying an image restoration model to the low keV ion beam image of the sample to generate a restored image. The sample is then localized within the restored image, and a low keV milling of the sample is performed with the ion beam based on the localized sample within the restored image.

Low keV ion beam image restoration by machine learning for object localization

Methods and systems include acquiring instances of data relating to multiple layers of a sample obtained via slice and view imaging where the electron interaction depth of the charged particle beam during each irradiation of the sample is larger than the thickness of the first layer and/or the thickness of the second layer. A simulated model is then accessed that identifies a plurality of yield values that identify expected portions/ratios of detected emissions that are expected to be generated by material in corresponding layers/depths of the sample. The yield values are used to segregate the instances of data into component portions based on the particular layer of the sample in which the structures expected to have generated the associated emissions are located. The component portions are then used to create reconstructions of individual layers and/or 3D reconstructions of the sample with reduced depth blur.

Slice depth reconstruction of charged particle images using model simulation for improved generation of 3D sample images

Methods and systems for neural network based image restoration are disclosed herein. An example method at least includes acquiring a plurality of training image pairs of a sample, where each training image of each of the plurality of training image pairs are images of a same location of a sample, and where each image of the plurality of training image pairs are acquired using same acquisition parameters, updating an artificial neural network based on the plurality of training image pairs, and denoising a plurality of sample images using the updated artificial neural network, where the plurality of sample images are acquired using the same acquisition parameters as used to acquire the plurality of training image pairs.

Acquisition strategy for neural network based image restoration

Growing community of inventors

Eindhoven, Netherlands

Pavel Potocek