Search

Paper is non-uniform; thus, it is difficult to ascertain whether the slight differences in near-infrared absorption spectra are caused by the difference in paper type or the paper’s non-uniformity. We used near-infrared hyperspectral imaging to measure the spectra of many different points quickly for eight different paper samples. Next, we analysed the data using machine learning and constructed a model to ascertain whether the differences in the spectra are due to the difference in paper type or the paper’s non-uniformity. Even papers with very similar spectral shapes were distinguished with high accuracy with this method. As the number of spectra used to construct the classification model increased, the accuracy of the classification model increased. With 6840 spectra per sample, the accuracy exceeded 85%, and with 68,400 spectra per sample, the accuracy for identification was 87.8%.

We preliminarily examined whether it is possible to visualize synthetic materials applied to the surface of a base material by using near-infrared hyperspectral imaging at wavelengths of 1.0–2.35 µm. We focused on the applicability of this method for visualizing adhesives and varnishes, assessing the effect of underlying colorants on measurements, and selecting the optimal method for data analysis. Experimental samples were adhesive on a painted ceramic plate and varnish on printed paper. We performed correlation analysis, partial least squares regression analysis, linear discriminant analysis, decision tree, KNN and principal component scores 1 and 2 to evaluate the measurement data. The conventional method (visible/near-infrared spectroscopic/fluorescence photographs at wavelengths of 400–1000 nm.) could not visualize the adhesive on the painted ceramic plate or the varnish on the printed paper, whereas the new method could do so. Therefore, this method has the potential to be useful for inspecting the distribution of synthetic materials on the surface of a base material.

In the present study, we successfully obtained nonstaining blood flow images of a developing fish egg embryo using optical interference caused by the Doppler shift. The spectral distribution of light reflected by moving objects such as the heart and red cells was found to be different from that of the incident light because of the Doppler effect. Interference between different frequency components was observed in an interferogram through heterodyne interaction using an imaging-type two-dimensional Fourier spectroscopic system, and information on the intensities of the spectral components was obtained by Fourier transformation. Beat signals with specific frequencies due to the heart beating and blood flow of the fish egg embryo were detected. When the signals were plotted in two dimensions, the heart part and vessel flows were clearly visualized without staining. In addition, near-infrared (NIR) images were produced using absorbance spectra of the molecular vibrations of O-H and C-H groups included in water, hydrocarbons, and aliphatic compounds. Obtaining nonstaining blood flow images using heterodyne optical interference and images of molecular distribution using molecular vibrational information simultaneously manifests an exciting advance in NIR imaging.

By combining a bolometer detector with an imaging-type interferometer, an inexpensive, easy-to-handle wide-field mid-infrared hyperspectral imaging apparatus was produced. We measured the distributions of four types of thin adhesive layers on an aluminium plate and analysed the results using correlation coefficients to visualise the distribution of various adhesives that cannot be discerned by the naked eye or conventional methods such as visible/near-infrared spectroscopic/fluorescent photography. The measurement wavelength range, obtained spectrum’s wavenumber resolution, and measurement time was 8–14 μm, about 9 cm−1, and about 30 s, respectively. Using conventional methods, adhesives could not be distinguished from the others. By using this method, we found that adhesives could be precisely distinguished by setting an appropriate threshold value for the correlation coefficient. Thus, our approach can accurately measure the spatial distribution of different types of adhesive that cannot be discriminated by conventional methods.

We built a hyperspectral imaging apparatus using middle-infrared light of 8–14 μm, which has a strong ability to identify organic materials, and attempted visualization of the distribution of organic materials that could not be identified by a naked eye. For this purpose, we utilized a low-cost bolometer camera (Nippon Avionics co., ltd. C100V, Japan) for its easy availability rather than an expensive mercury cadmium telluride (MCT) array sensor. To compensate for the low sensitivity of this bolometer, we adopted a Fourier-type spectroscopic system (Aoi Electronics co. ltd., Japan) using an imaging interferometer devised by the Kagawa University, Japan; this interferometer has higher light-utilization efficiency than Michelson interferometers, which are used in popular interferometry techniques. In this study, 4 types of adhesives, 9 types of varnishes and more than 50 types of inks were put on Al plates of size 10 cm × 10 cm and were used as samples. Glossy paper for printing photos with an inkjet printer was also used as a sample. A 300 °C black body of size 15 cm × 15 cm was used as a light source. Spectra of 320 × 240 points were measured at a wavelength resolution of approximately 9 cm−1. The mirror was scanned only once. The measurement time was approximately 30 s. Hyperspectral images of adhesives, varnishes and inks on Al plate and paper were successfully measured. Spectra over a 5 × 5-pixel neighborhoods were averaged, and the averaged spectra were compared with those measured by a commercially available Fourier transform infrared (FTIR) spectroscopy. The averaged and measured spectra had absorption peaks at the same wavelengths. Furthermore, by analyzing the measured spectra, the distribution of substances invisible to the naked eye was visualized. Our results show that if low-absorbance organic materials are put on a high-reflectance surface such as an Al plate, the middle-infrared hyperspectral imaging could be measured using a bolometer. Additionally, hyperspectral imaging of high-reflectance paper, such as glossy paper, could also be measured. Because a bolometer camera is much cheaper than an MCT array, hyperspectral imaging with such a camera has many potential applications. Moreover, an imaging interferometer, with its high efficiency of light utilization, is very suitable for the purpose.

A simulation environment of plant operation, which consists of an operator model, a human-machine interface (HMI) model and a plant model was developd. When the operator model recognizes an abnormal situation, the model generates a countermeasure program in its short-term memory based on both the plant state information acquired from the plant model through the HMI model and the knowledge bases installed into the operator model. A comparison between the operator model's behavior and the human operator's behavior under the same malfunction was used to extract the characteristics of the human operator and to obtain information on human errors.

Cathodoluminescence (CL) is the light that is emitted from a material irradiated by an electron beam. The present study was undertaken to show the applicability to biological studies of a scanning electron microscope (SEM) equipped with a high-sensitive cathodoluminescence detection system. For this purpose, we injected inorganic fluorescent powders (P43) suspended in phosphate buffered saline into rat blood circulation, fixed the animals with glutaraldehyde within a day, and observed the hepatic tissues with a SEM. Our instrument enabled the simultaneous collection of both secondary electron (SE) and CL images of these tissues. Backscattered electron (BSE) images of the same portion were also able to be obtained with this microscope. SE and BSE images clearly showed the three-dimensional structure of the hepatic tissues including hepatocytes, Kupffer cells, Ito cells, and sinusoidal epithelial cells, while CL images visualized cathodoluminescence signals emitted from P43 as bright spots. We observed non-coated tissues under a low-vacuum condition and metal-coated tissues under a high-vacuum condition, and found that the high-vacuum observation of metal-coated tissues provided high quality CL images of P43 in the Kupffer cells. The superimposition of the CL images onto the corresponding SE or BSE images revealed that bright spots in the CL images were produced by the fluorescent powders uptaken by Kupffer cells. These findings indicate that the detection of CL as well as SE or BSE signals by SEM all provide us with useful information on the distribution of fluorescent tracers in tissues and cells in three-dimensional images.