Improving the segmentation of scanning probe microscope images using convolutional neural networks

Farley, Steff; Hodgkinson, Jo E. A.; Gordon, Oliver; Turner, Joanna; Soltoggio, Andrea; Moriarty, Philip; Hunsicker, Eugénie

doi:10.1088/2632-2153/abc81c

Cited by 17 publications

(19 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In [38], because manual segmentation of nanostructured surface images is time-consuming and requires high professionalism. Therefore, it is necessary to develop an automated method for segmenting nanoparticles based on U-Net.…”

Section: Other Applicationsmentioning

confidence: 99%

A State-of-the-art Survey of U-Net in Microscopic Image Analysis: from Simple Usage to Structure Mortification

Wu¹,

Liu²,

Li³

et al. 2022

Preprint

View full text Add to dashboard Cite

Image analysis technology is used to solve the inadvertences of artificial traditional methods in disease, wastewater treatment, environmental change monitoring analysis and convolutional neural networks (CNN) play an important role in microscopic image analysis. An important step in detection, tracking, monitoring, feature extraction, modeling and analysis is image segmentation, in which U-Net has increasingly applied in microscopic image segmentation. This paper comprehensively reviews the development history of U-Net, and analyzes various research results of various segmentation methods since the emergence of U-Net and conducts a comprehensive review of related papers. First, This paper has summarizes the improved methods of U-Net and then listed the existing significances of image segmentation techniques and their improvements that has introduced over the years. Finally, focusing on the different improvement strategies of U-Net in different papers, the related work of each application target is reviewed according to detailed technical categories to facilitate future research. Researchers can clearly see the dynamics of transmission of technological development and keep up with future trends in this interdisciplinary field.

show abstract

Section: Other Applicationsmentioning

confidence: 99%

A State-of-the-art Survey of U-Net in Microscopic Image Analysis: from Simple Usage to Structure Mortification

Wu¹,

Liu²,

Li³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Furthermore, noise and other artifacts in the image can be mistaken as an object and uneven levelling makes it impossible to threshold features at a particular height. It was recently shown that the CNN U-Net, introduced above, could be used to classify different types of agglomerations of nanoparticles in AFM images [ 141 ]. U-Net was chosen because it works well on relatively small training sets (in this case 428 images, which was also a manageable size for the supervised learning applied here), and the fact that it provides precise localization, as required for segmenting such features.…”

Section: Reviewmentioning

confidence: 99%

The role of convolutional neural networks in scanning probe microscopy: a review

Azuri¹,

Rosenhek‐Goldian²,

Regev‐Rudzki³

et al. 2021

Beilstein J. Nanotechnol.

View full text Add to dashboard Cite

Progress in computing capabilities has enhanced science in many ways. In recent years, various branches of machine learning have been the key facilitators in forging new paths, ranging from categorizing big data to instrumental control, from materials design through image analysis. Deep learning has the ability to identify abstract characteristics embedded within a data set, subsequently using that association to categorize, identify, and isolate subsets of the data. Scanning probe microscopy measures multimodal surface properties, combining morphology with electronic, mechanical, and other characteristics. In this review, we focus on a subset of deep learning algorithms, that is, convolutional neural networks, and how it is transforming the acquisition and analysis of scanning probe data.

show abstract

“…While noise removal is also particularly vital, a denoising autoencoder trained on the simulated data performed extremely well at this task and may be applicable elsewhere to smooth features and remove processing artifacts. Additional improvements could be made with alternative network structures, fewer target structures, and/or a machine learning approach to binarisation, 4,38 defect segmentation, and anomaly detection. 35,36 The method we have developed for automated identification of nanostructured patterns is an effective first stage of file search, capable of isolating files and locations of interest.…”

Section: ■ Conclusionmentioning

confidence: 99%

“…Because automating this process is not trivial, preprocessing is typically done manually . However, unlike other automated methods , which maximize aesthetic quality for a large visual variety of images, we are not interested in aesthetic quality, and only need a broadly reasonable preprocessing method. By optimizing the routine for the target structures, we can also assume that images unable to be processed are of low quality and/or not the structures being searched for and so can be discarded.…”

Section: Preprocessing and Filtering Real Datamentioning

confidence: 99%

“…There has been a recent flurry of studies applying machine learning to scanning probe microscopy (SPM), ranging from distinguishing otherwise indistinguishable data, segmenting and analyzing surface features and defects, − and assessing probe quality. , While these studies will undoubtedly allow us to collect more, higher quality data in the future, we are only beginning to find ways of better using , existing data left unanalyzed on old hard drives, CDs, and even floppy discs . Indeed, while one of the greatest advantages of atomic force microscopy (AFM) is its ability to produce huge numbers of scans on a variety of distinct surfaces, the need to manually select data to analyze leads to much of it being under used at best.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Automated Searching and Identification of Self-Organized Nanostructures

et al. 2020

Self Cite

View full text Add to dashboard Cite

Currently, researchers spend significant time manually searching through large volumes of data produced during scanning probe imaging to identify specific patterns and motifs formed via self-assembly and self-organization. Here, we use a combination of Monte Carlo simulations, general statistics, and machine learning to automatically distinguish several spatially correlated patterns in a mixed, highly varied data set of real AFM images of self-organized nanoparticles. We do this regardless of feature-scale and without the need for manually labeled training data. Provided that the structures of interest can be simulated, the strategy and protocols we describe can be easily adapted to other self-organized systems and data sets.

show abstract

Improving the segmentation of scanning probe microscope images using convolutional neural networks

Cited by 17 publications

References 20 publications

A State-of-the-art Survey of U-Net in Microscopic Image Analysis: from Simple Usage to Structure Mortification

A State-of-the-art Survey of U-Net in Microscopic Image Analysis: from Simple Usage to Structure Mortification

The role of convolutional neural networks in scanning probe microscopy: a review

Automated Searching and Identification of Self-Organized Nanostructures

Contact Info

Product

Resources

About