Abstract:A simple and effective method for obtaining the uniformity information of films is proposed. The method is based on the accurate measurement method for nano film thickness we given before. It firstly measures the spectrum of the film in different positions on a substrate containing a transition layer and then obtains the uniformity information of the film according to the interference peak difference. Compared with the traditional method, this method reduces the error resulting from the direct measurement of film thickness. Therefore, it can obtain the uniformity information of the film quickly. The method is convenient and simple in operation and is of important value to the improvement of coating processes.

Abstract:When taking Ar gas as the working mixture and taking CH4 as the reactive gas, diamond-like carbon (DLC) films are successfully prepared on Ge substrates by using the low voltage reactive ion plating (RLVIP) technology. The microstructures, optical properties and hardness of the prepared DLC films are characterized by means of Raman spectroscopy, infrared spectroscopy and nano-indentation test technology. The result shows that the prepared DLC films (ID/IG=0.918) have a higher sp$^3$ content, a hardness up to 28.6 GPa and a elastic modulus of 199.5 GPa. The single layer film has a peak transparency of 63.6 % and a mean transparency of 62 % in the waveband from 8 μm to 11.5 μm.

Abstract:A non-uniformity correction method for resistor arrays is given and a flood non-uniformity measurement method in the mapping ratio 1∶1 is analyzed according to the functional diagram of a non-uniformity measurement model. To improve the edge effect and non-ideal convergence due to image deterioration, a flood non-uniformity measurement method based on point spread function estimation is proposed. The simulation result shows that compared with the former method, this method can reduce the edge effect and has a better convergence when a greater smooth factor is existent.

Abstract:A remote sensing information extraction method based on smart ecological cells is put forward. In the method, the crop phenological characteristics, spectral evolution characteristics and special correlation feedback are combined and the cellular automata, fuzzy reasoning and BP neural network intelligent computation are fused. By taking the multi-temporal Landsat--7 ETM images as the experimental data, the monitoring of tobacco acreage planting area based on smart ecological cells is carried out. The monitoring result shows that this method can effectively improve the extraction accuracy of remote sensing information and is of value to practical applications.

Abstract:Sea-sky-line detection can be implemented by effectively suppressing the background and enhancing the target in an image according to the features of sea-sky environment. An improved method for wiping off crass points is proposed on the basis of the linear fitting algorithm. The method is simulated with the Matlab software. The simulation result shows that this improved method can effectively detect the positions of the sea-sky-line and ships against the sea-sky background. It provides a good basis for target recognition and image registration.

Abstract:According to the basic principle of hypothesis testing, an algorithm for detecting the region of interest (ROI) of a small infrared target in an image is proposed. In the method, the ROI of a small infrared target in an image is firstly extracted by using the minimum error probability criterion and then target extraction and analysis are carried out in those regions of interest. The experimental result shows that this method overcomes the disadvantages such as large computation and difficult analysis in traditional methods and is very suitable for the high performance detection of small infrared targets in infrared images.

Abstract:A fast and non-invasive method for determining the mixture percentages of brake fluid based on visible and near infrared spectroscopy (Vis--NIRS) was proposed. Both a partial least square regression (PLSR) model and a least-square support vector machine (LS--SVM) model were established according to the spectra obtained in the whole wavelength range. With those two models, good prediction results were obtained. The determination coefficients of their calibration and prediction sets (r2c and r2p) were greater than 0.98. The successive projection algorithm (SPA) was used to select the effective variables. Finally, eight variables of 439 nm, 443 nm, 459 nm, 519 nm, 570 nm, 717 nm, 896 nm and 902 nm were selected as the optimal variables to be input into the PLSR and LS-SVM models. The r2c and r2p of both two models were greater than 0.97 which was adequate for practical application. It was concluded that Vis-NIRS could be used to fastly and non-invasively determine the mixture percentages of brake fluid.