http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
NIR 초분광 영상의 효과적인 플라스틱 분류를 위한 Dark-Threshold 기반 전처리 기법
김희강(Heekang Kim),김성호(Sungho Kim) 대한전자공학회 2016 대한전자공학회 학술대회 Vol.2016 No.6
Recently, various classification plastics methods have been studied, as the importance of recycling increases. In case of plastic, optical methods such as Raman spectroscopy, Fourier transform infrared (FT-IR) or Near infrared (NIR) sensors were applied for plastic classification. However there are some problem using these sensors. There are some noise such as salt noise and random noise. Therefore, this paper presents spectral analysis by comparing preprocessing methods. The classification results demonstrate the superiority of the proposed preprocessing in the experiments.
초분광 영상의 조명 불변성을 위한 Per-Norm 기반 분광 정규화 기법
김희강(Heekang Kim),김성호(Sungho Kim) 대한전자공학회 2015 대한전자공학회 학술대회 Vol.2015 No.6
When analyzing hyperspectral images, the same material shows different spectral information, because there are variatious caused by shadow shading and highlight. This paper introduces more effective spectral normalization than the traditional spectral normalization. The propsed Percent-Normalization(Per-Norm) can noral ize spectral profile by removing min value and making the max value to equal 100 percent.