CAJ | 학술논문

从红外图像的特性出发,分析了红外图像的直方图、噪声、分辨率及对比度等本质属性,在此基础上,针对红外图像信噪比高、对比度低的缺点,以高斯混合模型的参数设想和高斯分布的特定规律来模拟红外图像的像素分布和动态区间,通过对分割到高斯混合模型后的图像信息进行相应的变换,来实现对红外图像的自适应均衡和对比度增强处理。实验效果表明,图像亮度和对比度增强明显,并很好地保留和增强了图像细节,整幅图像的视觉效果得以显著改善,达到了图像增强的预期目的。
종홍외도상적특성출발,분석료홍외도상적직방도、조성、분변솔급대비도등본질속성,재차기출상,침대홍외도상신조비고、대비도저적결점,이고사혼합모형적삼수설상화고사분포적특정규률래모의홍외도상적상소분포화동태구간,통과대분할도고사혼합모형후적도상신식진행상응적변환,래실현대홍외도상적자괄응균형화대비도증강처리。실험효과표명,도상량도화대비도증강명현,병흔호지보류화증강료도상세절,정폭도상적시각효과득이현저개선,체도료도상증강적예기목적。
The human eye is not sensitive to small variations around dense data but is more sensitive to widely scattered fluctuations. Thus, in order to dense data with low standard deviation should be ard deviation should be compacted. This retained. In order to achieve this, we use increase the contrast while retaining image details, dispersed, where as scattered data with high stand- operation should be the GMM to partiti done so that the gray-level distribution is on the distribution of the input image into a mixture of different Gaussian components, then transforming the algorithm according to the purpose of image enhancement