The precision of training samples has important influence on the accuracy of Supervised Classification of Remote Sensing Images. It is easy to omit the ground feature which covers a small amount by visual selection directly on the original image. This leads to a large number of wrong points and leakage points which influences the Classification Accuracy of Remote Sensing Images. Based on what was said above, we first use K-means clustering algorithm to initialize the samples in this paper. According to the clustering results, the training samples selected by visual labeling and density-based k-means algorithm participated in the SVM classification model. The results show that selecting training samples by the k-means algorithm based on density can reduce the influence of visual labeling on the subjective factors of training sample selection. Meanwhile,through the optimization of the radius weighting factor, we have found optimal combination of weighting factors and then the classification accuracy of remote sensing images can be increased from 81.70% to 88.89%.
Microchannel plates (MCPs) are two-dimensional arrays of microscopic channel electron multipliers as the key component of the image intensifier. In this paper, the plasma etching process of large-size microchannel plates based on silicon is developed. Firstly, the etching of micro groove structure is researched for measure easily. The influences of the process parameters on the etching rate and sidewall verticality, such as oxygen flow, the pressure and the inductively coupled plasma (ICP) power, are studied. The results show the etching rate becomes larger and larger with the increasing of the pressure and ICP power. And the oxygen flow contributes to etching. But the sidewall verticality is worse with increasing oxygen flow. Though parameter experiment, the optimized parameter is got: the pressure 10mTorr, the ICP power 600W, the oxygen flow 6sccm. Using this parameter to etch micro-channel array, the microchannel array of largescale is got with pore 10μm, pitch 5μm, aspect ratio 20:1.
Microchannel plates（MCPs）are the key component of the image intensifier. Compared with the traditional MCPs, the Si MCPs which are fabricated by micro-nanofabrication technologies have a high gain, low noise and high resolution etc. In this paper, the lithography process is studied in the process of fabricating periodic micropore array with 10 um pores and 5 um pitch on Si. The effects of exposure time, reversal bake temperature and development time on the lithography quality are focused. By doing a series of experiments the better result is got: the photoresist film is obtained at a low speed 500/15(rpm/s) and a high speed 4500/50(rpm/s); the soft bake time is 10min at 100℃; the exposure time is 10s; the reversal bake time is 80s at 115℃; the development time is 55s. By microscope observation and measurement, the pattern is complete and the size of the pattern is accure, it meets the requirement of lithography process for fabricating Si-MCP.