Eisaku Oho

A new non-linear pseudo-Laplacian filter for enhancement of secondary electron images

A simple, yet effective, non‐linear pseudo‐Laplacian filter has been newly developed to enhance secondary electron (SE) images. This filter is a combination of the second derivative along the direction of the local gradient and a non‐linear weight factor. The filter can successfully enhance SE images without the undesirable effects of noise which are often seen in conventional Laplacian filtered images. Hence, the processed results with high image quality can make original SE images easier to interpret. The effectiveness is especially useful in low‐voltage scanning electron microscopy, because SE images taken at low voltages are not so likely to have good image sharpness.

A method using an on-line digital computer for the improvement of resolution of backscattered electron images

In principle, the resolution of backscattered electron (BSE) images can be little improved, even though an infinitely small beam size is achieved by various improvements in the intrinsic instrument. In order to circumvent this problem, a method is proposed which utilizes an on‐line digital computer for the image recording and processing. The major image‐processing tools are reduction, expansion, super‐imposition with matching of the images, and high‐emphasis filtering in the Fourier domain. By using various combinations of these techniques, the resolution of BSE images has been significantly improved. The validity of these improved images has been confirmed. In the case of a BSE image with too wide a dynamic range, both the present method and digital homomorphic filtering provide successful results.

Caustic patterns combined with second and third order astigmatism in high resolution scanning electron microscopes

Abstract Caustic patterns exhibiting unwanted second and third order astigmatism, which are observed for high partial coherence of the illuminating beam in a high resolution scanning electron microscope using a field emission gun, are discussed. The electron optical aberration theory was developed for a caustic formation by a microgrid specimen taking into account the deflection angle depending on different hole sizes. It was found that the caustic patterns exhibiting second and third order astigmatism appear at all defocusing planes, and the experimental results measured on the basis of caustic formation theory are in close agreement with the predicted data of the instrument. It is concluded that third order astigmatism compensation with an eight pole stigmator is essential for the high resolution scanning electron microscope.

A contamination reducing method by ion beam bombardment of the specimen in high resolution electron microscopy

Abstract The theory of contamination and a contamination reducing method are discussed on the basis of time dependent micrograph series and their tilted images for determining contamination. For a high current density of an electron probe in the field emission scanning electron microscope, it is observed that contaminated cones are formed in proportion to the exposure time of an electron beam. From the measurement of the contamination layer thickness and its area, the contamination rate and time dependent shape are formulated, mainly depending on the cross-section and current density together with the average lifetime of adsorption molecules. It is found that the contamination rate and stray contamination of outgassing molecules forming part of the specimen are effectively reduced by a pre-bombardment of argon ions on the surfaces of specimens. The contamination rate is reduced to a small extent (5%) using the present method.

Digital image-processing technology useful for scanning electron microscopy and its practical applications

Publisher Summary This chapter discusses several important subjects that are closely related to the scanning electron microscopes (SEM) field. For a number of years, SEM has provided outstanding high-resolution images with very great depths of field in biophysics, material science, and so forth. In the early years, several digital image-processing techniques, as well as many analog techniques, were introduced to the SEM field. Because the SEM image is essentially the electric signal, it is suitable for utilizing these image-processing techniques. Analog image-processing techniques were mostly employed for SEM signal enhancement in the early stages because digital techniques were in the developmental stage and the cost of using them was extremely high. Many SEM users utilize a conventional recording system consisting of a video monitor with a resolution of about 2000 lines and a high-performance camera. The conventional system is believed to be satisfactory in image quality for the average SEM user. However, serious deterioration in information obtained by this method has been confirmed by comparing it with an on-line digital-recording system that is closer to the ideal for SEM images.

DIGITAL IMAGE PROCESSING TECHNOLOGY FOR SCANNING ELECTRON MICROSCOPY

Publisher Summary Scanning electron microscopy (SEM) has provided outstanding high-resolution images with very great depth of field in biophysics, material science, etc. Many SEM users still utilize a conventional recording system consisting of a video monitor with a resolution of approximately 2000 lines and a high-performance camera. The conventional system is satisfactory in image quality for the average SEM user. The scan coils of the SEM generally are used to perform a fast scan in the x direction and slow scan in the y direction. The former produces a continuous signal, while the latter gives what is called a “sampled signal.” The SEM is operated under various conditions, including electron beam size, incident current, number of scanning lines per frame, and magnification. SEM images taken with underscanning are contaminated by the aliasing error (artifact) to a greater or lesser extent. The fine structures of the specimen are not accurately converted into an analog SEM image. However, except for some particular specimens and conditions, it may not disturb the observation of the specimen experimentally.

Feature evaluation of complex hysteresis smoothing and its practical applications to noisy SEM images.

Quality of a scanning electron microscopy (SEM) image is strongly influenced by noise. This is a fundamental drawback of the SEM instrument. Complex hysteresis smoothing (CHS) has been previously developed for noise removal of SEM images. This noise removal is performed by monitoring and processing properly the amplitude of the SEM signal. As it stands now, CHS may not be so utilized, though it has several advantages for SEM. For example, the resolution of image processed by CHS is basically equal to that of the original image. In order to find wide application of the CHS method in microscopy, the feature of CHS, which has not been so clarified until now is evaluated correctly. As the application of the result obtained by the feature evaluation, cursor width (CW), which is the sole processing parameter of CHS, is determined more properly using standard deviation of noise Nσ. In addition, disadvantage that CHS cannot remove the noise with excessively large amplitude is improved by a certain postprocessing. CHS is successfully applicable to SEM images with various noise amplitudes.

A cone formation theory of contamination in high resolution transmission electron microscopy

Abstract The formation theory of the contamination cone in high resolution transmission electron microscopy is discussed on the basis of a time dependent series of micrographs and their tilted images for evaluating the contamination process. For a high current density of an electron probe focused on the specimen, it was observed that two contaminated cones covering both upper and lower surfaces of a carbon substrate are formed as a function of the exposure time of the electron beam. From the measurement of the cross-section and height of contaminated cones, the contamination rate together with the volume rate, and time dependent shapes of various cones were analysed. The analysis was dependent on the diffusion cross-section and current density together with an average lifetime of adsorption molecules.

Support system for fine focusing and astigmatism correction using an auditory signal in scanning electron microscopy

The current study describes a new support system for fine focusing and near-perfect astigmatism correction for scanning electron microscopy (SEM). The signal-to-noise ratio of a series of SEM images obtained from fast scan rates (TV scan) was adopted as a new metric for evaluating focus. Measured signal-to-noise ratio values were converted to an acoustic signal (sound wave frequency) using digital image processing techniques, enabling the SEM user to evaluate image focus using the auditory modality. Accurate focusing and correcting astigmatism in general-purpose SEM is traditionally time-consuming and difficult. The proposed system may substantially reduce the required operation time for fine focusing. Moreover, the system is relatively immune to noise, successfully supporting focus and astigmatism correction with very noisy SEM images. Our proposed focus support system may be helpful for general-purpose SEM observation of a variety of specimens under a wide range of operating conditions.

Special raster scanning for reduction of charging effects in scanning electron microscopy

A special raster scanning (SRS) method for reduction of charging effects is developed for the field of SEM. Both a conventional fast scan (horizontal direction) and an unusual scan (vertical direction) are adopted for acquiring raw data consisting of many sub-images. These data are converted to a proper SEM image using digital image processing techniques. About sharpness of the image and reduction of charging effects, the SRS is compared with the conventional fast scan (with frame-averaging) and the conventional slow scan. Experimental results show the effectiveness of SRS images. By a successful combination of the proposed scanning method and low accelerating voltage (LV)-SEMs, it is expected that higher-quality SEM images can be more easily acquired by the considerable reduction of charging effects, while maintaining the resolution.